Super Game Boy Console

I was watching the Andy West’s video of Joe Ostrander’s Game Boy Console project on the Element 14 Youtube channel. It was inspired by and earlier project video that Andy West had done awhile ago. In the new video Andy builds one of Joe’s v3.0 boards, and gives some details I will likely be missing here.

The latest version from Joe was v3.1 it uses a Raspberry Pi Pico, a Super Game Boy Cartridge with a few other parts to build a Super Game Boy Console in a tiny Retroflag NESPi case. With v3.1 Joe used the Super Game Boy Cartridge Port instead of the SMD Cartridge Port Andy had for the v3.0.

Here is the Video:

I liked this project because it included some things I could practice with and some new things.

Andy’s original project included the DMG Link Port for Multiplayer games, but Joes project did not. I looked over the schematics for the SGB and the DMG as well as Joe’s version and it seemed simple to add that feature.

The Super Game Boy can play the original Game Boy DMG games. The Super Game Boy didn’t have the Game Link Port, but the CPU was the same except it doesn’t have the Nintendo name scrolling down the screen at power on (unless the cartridge isn’t recognized properly you never see that screen at all).

I was able to contact Joe and he sent me a copy of the Easy EDA Project files. He had it listed at v3.1 at the time with the Super Game Boy Cartridge port. I modified Joe’s main pcb to add in the parts required for the Link port. This was the first time I have worked with Easy EDA. It took me a little bit to get used to working with it. I have used Eagle in the past for other projects, but for Hobby work it is very limited in board size. Easy EDA can send an order to JLCPCB for production, or your can have it generate Gerber files to send to your favorite PCB Producer. I made a Footprint for the DMG Link Port is a recreation of Gekkio’s KiCad DMG Link Port KiCad footprint in Easy EDA, so that I could make the DMG Link Port adapter board.

https://github.com/Gekkio/gb-hardware

There is a single Pi Pico that handles controller input and generates the VGA Video output. It also uses a VGA to HDMI Adapter to take the VGA output from the Pico to HDMI including the Audio.

It has options to either use a NES Controller port and Controller, or to take a Wii jack and use a compatible controller for that, be it one of the Wii Classic type controllers or one intended for the NES Mini such as I have used. There is different firmware for the Pico depending on which controller type you go with. Joe has both versions ready to go for you in the .u2f format. I went with the Wii port, and the cheapest way for me to get it quickly was to buy a Wii controller extension cable and salvage it out of there. I figured it was likely I could salvage and reuse it, and while 3 of the pins were cut short, it did turn out to be the regular through hole port and works just fine in my case. There are different Controller Port Mounts and as I said Firmware depending on which controller option you go with. The firmware is in the .uf2 format that is very easy to install to the Pico. You first download the .uf2 that you need on your computer, then get a usb cable for the Pico, hold the button on the Pico as you plug it into the USB port on your computer. You can release the button then, if you did it right, the Pico will show up as a drive on your computer. From there copy and paste the proper .uf2 file into that drive. The Pico will upload the code and reboot itself automatically. The files are gb_vga__pcbv3_nes_controller.uf2 for the NES Controller Port and gb_vga__pcbv3__nes_classic_or_wii.uf2 for the Wii Controller port (filenames prior to 1/15/23).

Update: 1/15/23 It appears Joe released an update for the code on the Pico, the new version only supports the Wii type controller. The Home button now brings up the OSD Controls. I’ll have to check it out later.

The Controller Ports also require different Adapter PCBs and wiring as well depending which you go with. The PCBs are labeled on both the Main PCB and the Breakout Boards. The pictures of my prototype board are not the final design for v3.21 on the final one (which I show a render picture of at the end of this post) you will find the “<– NES” and the “<– Wii” markings on the silk screen. Be sure to use 3V3 for the Wii Breakout. I have updated Joe’s Wii Breakout board to list 3V3 instead of VCC. For the NES use 5V and Gnd just as the board is labeled to accept.

There are numerous parts. Much of it is surface mount, minus the Pico and the reused SGB Cartridge port.

BOM:

Name	Designator	Footprint	Quantity	Manufacturer Part	Manufacturer	Link
27pF	C1,C2	0603 CAPACITOR	2	C1608C0G1H270J080AA	TDK Corporation	https://www.digikey.com/en/products/detail/tdk-corporation/C1608C0G1H270J080AA/513761
1uF Tantalum Polarized Capacitor	C3,C4	1411 CAP-SMD_L3.5-W2.8-R-RD	2	T491B105K050AT	KEMET	https://www.digikey.com/en/products/detail/kemet/T491B105K050AT/3724722
100pF	C5,C6,C7,C8	0603 CAPACITOR	4	06035A101JAT2A	KYOCERA AVX	https://www.digikey.com/en/products/detail/kyocera-avx/06035A101JAT2A/563285
MMBD7000HS-7-F	D1,D2,D3,D4,D5	SOT-23-3_L2.9-W1.3-P1.90-LS2.4-BR	5	MMBD7000HS-7-F	Diodes Incorporated	https://www.digikey.com/en/products/detail/diodes-incorporated/MMBD7000HS-7-F/2179586
1M	R1	RESISTOR_R0603	1	WR06X105 JTL	Walsin Technology Corporation	https://www.digikey.com/en/products/detail/walsin-technology-corporation/WR06X105-JTL/13239221
10K	R2,R3	RESISTOR_R0603	2	RNCP0603FTD10K0	Stackpole Electronics Inc	https://www.digikey.com/en/products/detail/stackpole-electronics-inc/RNCP0603FTD10K0/2240139
1K	R4,R6,R8	RESISTOR_R0603	3	RMCF0603FG1K00	Stackpole Electronics Inc	https://www.digikey.com/en/products/detail/stackpole-electronics-inc/RMCF0603FG1K00/1713699
500R	R5,R7,R9	RESISTOR_R0603	3	RT0603BRC07500RL	YAGEO	https://www.digikey.com/en/products/detail/yageo/RT0603BRC07500RL/7708292
47R	R10,R11	RESISTOR_R0603	2	RC0603JR-0747RL	YAGEO	https://www.digikey.com/en/products/detail/yageo/RC0603JR-0747RL/726793
180K	R12	RESISTOR_R0603	1	RC0603JR-07180KL	YAGEO	https://www.digikey.com/en/products/detail/yageo/RC0603JR-07180KL/726723
510R	R13,R14	RESISTOR_R0603	2	RC0603JR-07510RL	YAGEO	https://www.digikey.com/en/products/detail/yageo/RC0603JR-07510RL/726801
220R	R15,R16,R17,R18,R19(Power/Reset Board)	RESISTOR_R0603	5	RMCF0603FT220R	Stackpole Electronics Inc	https://www.digikey.com/en/products/detail/stackpole-electronics-inc/RMCF0603FT220R/1760809
ABLS-4.194304MHZ-D-4-Y-T	X2	CRYSTAL-SMD_L11.4-W4.7-LS12.7	1	ABLS-4.194304MHZ-D-4-Y-T	ABRACON	https://www.digikey.com/en/products/detail/abracon-llc/ABLS-4-194304MHZ-D-4-Y-T/9997833
SN74LVC245ANSR	U4	SOIC-20_L13.3-W5.5-P1.27-LS7.8-BL	1	SN74LVC245ANSR	TI	https://www.digikey.com/en/products/detail/texas-instruments/SN74LVC245ANSR/499514
Raspberry Pi Pico	U5	RASPBERRY-PI-PICO	1	SC0915	Raspberry Pi	https://www.digikey.com/en/products/detail/raspberry-pi/SC0915/13624793
Super Gameboy Cartridge Header	U6	SUPER GAMEBOY CARTRIDGE HEADER	1		Nintendo	From Super Game Boy
DMG-CPU	U1	DMG-CPU-B	1		Nintendo	From Super Game Boy
LH5264 RAM	U2,U3	DMG LH5264	2		Nintendo	From Super Game Boy
GANA VGA to HDMI Audio Video Converter			1			https://a.co/d/2GTq5d7
Retroflag NESPi Case Plus			1			https://a.co/d/1K0sy7a
Super Gameboy Cartridge (to source the CPU, RAM and Cartridge Port)			1		Nintendo	Search Ebay
Wii/NES Classic Port 90 degree (Optional this or the Original NES Port)			1			https://www.aliexpress.us/item/3256804436132647.html
Original NES Controller Port (Optional this or the Wii Port)			1
Pin strip / header (for the Pico if it didn’t come with them)			1
Wire 24awg Silicone Insulated Wire						https://www.amazon.com/gp/product/B07TJXRGXM/?th=1
Wire 30awg Insulated Wire						https://www.amazon.com/Treedix-Copper-Wrapping-Jumper-Optional/dp/B08D6D1TKH/

Super Game Boy Console Main Board PCB	sgb_cart_motherboard_v3_21
Power and Reset Board PCB	power_reset_v3_21
Wii Controller Breakout Board PCB (Optional this or the NES Breakout)	wii_jack_w_holes_v3_21
NES Controller Breakout Board PCB (Optional)	nes_breakout_v3
DMG Game Link Multiplayer Breakout PCB (Optional)	dmg_link_v3_21
USB Power with Data PCB (Optional, to Program the Pico in the case)	pcb_v3_usb_w_data

BOM

I used the USB Port from Retroflag instead of the optional “USB Power with Data PCB”, I don’t expect to need to reprogram the Pico that often. I ordered in Joe’s Wii controller PCB. I also ordered his Power and Reset button board. I ordered the version of the Main PCB that I had added in the Link Port parts. For the Link port I ordered in a DMG-07 Four Player adapter to salvage off one of the Link Ports as Joe suggested.

I have since created a somewhat more refined version of the Main PCB and some other PCB changes. Keep in mind the BOM list above is for the v3.21 board not the prototype I have pictures of here. Some component designations have changed and D1 is now a Diode Array on the “finished” board. I also created the DMG Link breakout board. I used the protoboard version as seen in the pictures, but the Mount for the DMG Link board is designed for the PCB.

The first part was to do the main PCB modifications. I ordered the required boards, I also found out you can have them panelize the boards for you. I now have a crazy number of Wii Controller Breakout boards for the same price I would have paid for 5. That opens up some real possibilities for when I have a small board I would like more of.

The next part of the project was sourcing all of the other parts. Joe’s BOM at the time had a mistake on it the 22pf capacitor listed was a 0201 imperial part. The BOM above is listing a 0603 imperial sized part. That is an easy mistake to make as 0201 imperial is 0603 metric.. The parts listed above are the specific part numbers I ordered except for the wire. If you don’t care to do the Link Port, then you don’t need R15-R18, D1-D5, or C5-C8, or the DMG Link PCB, and the Link Port itself.

Once I had all the required parts it was time to prep the various parts. For the Retroflag case, I had to desolder the USB power board from the Power Reset button board so I could reuse it. I then had to desolder the Power and Reset Buttons as well as the LED and the LED Light pipe so I could reuse them. I then installed the Power and Reset Buttons and attached the USB power board to the new PCB. Just a note I missed seeing the pad for the Resistor for the LED on the board, I revised the board so that was listed on the silk screen and installed a 510 Ohm resistor on it ( I expect a 1k resistor would do well too, if you don’t want it so bright).

The case requires a bit of modification to fit the new parts. The opening where the two USB ports needs enlarged depending on which Controller port type you go with. The top needs modified by cutting out the opening for the Cartridge to get in. That requires taking the front center post out of the top as well. Be careful to not remove too much. I did have to slightly trim some of the top posts (but not all) about the thickness of the PCB to get it to fit nicely. The back center post was removed on mine as well using the PCB as it was on the prototype. I revised the PCB so that v3.21 should let that back center post remain now. To see how much the top posts needed shaved down (about the thickness of the PCB) just check the fit of the PCB laying in the top of the case, it basically needs to sit level and about flush with it. For the Pi Pico to clear the top, I had to remove the tabs and a little bit more where the fan is to snap in. I think if the Pico is mounted with short posts it would have more clearance. I decided to mount it with some header sockets so I can remove it if I want/need to in the future. For the bottom there there is some material that needs removed from under where the green mounting part goes in for the Controller port. That part is where the Controller Port is mounted as well as the hole in it is the one hole that holds the VGA to HDMI board in place. Below you can see the modifications to the case. I didn’t think to take “before pictures”. You can also see the new Power Reset board, USB Power board as well as the Wii board mounted into the bottom of the case. The Link Port will be added to the lower right cavity later.

Beyond the case prep, and pulling those parts from the original Power Reset board there is the Super Game Boy Cartridge parts to deal with.

With the Super Game Boy, we pull the board. We need to salvage four parts. The CPU, the two Ram Chips (they are identical) and the Cartridge Port. Below you can see the board after those parts are removed. I did clean the flux from the board and kept it if I find a future use for it. To remove the chips I used the technique that Andy West showed with Low Temp Solder alloy. I watched him do it a couple times, I then watched some other people using it to remove chips. For me I wasn’t having all that great a luck with the CPU, so to make the process go easier I used my hot air station to make it easier that helped me a lot. It seemed that in Andy’s video it must have also taken him awhile to get the CPU off, I felt better using the hot air as the heat was applied for less time than it was taking with the iron alone. I am sure I would have gotten it with just the iron, but it is also nice to use the hot air station for more than heat shrink now and then. That low temp alloy is weird stuff, the flux smells odd too. It is the first I used it. Be sure to clean the low temp solder from the chips with some solder wick after removing them. It needs removed as it is not good to have it on the final board it doesn’t work as well as solder.

For the Cartridge Port, I used some money from Christmas to purchase one of the cheaper desoldering guns. I also used the new desoldering gun to remove the switches from the Power Reset board and VGA and Audio sockets from the VGA to HDMI adapter. I could have done it with my desoldering iron, it is shown in one of my C64 refurbishing posts. I also used it to pull the Link Port from the DMG-07 4 Player adapter.

Super Game Boy after removing the required ICs and Socket to be reused.

Below is the VGA to HDMI Adapter after it was prepped for installation. The VGA Port and the 3.5mm Audio port had to be removed. Even with the desoldering gun, I ended up pulling some of the through holes out of the VGA port, nothing that was a problem though in this case. It has 5V (Large Red), Ground (Large Green) for power that go over to the Power Reset board. Then the Dark Blue (Left Channel) and Brown (Right Channel) wires go to the pad for Left and Right audio over on the Main PCB, Ground is all joined so that doesn’t need a wire in this case. Then we have the Red, Green and Blue for the R, G, B pins then the Orange ( V Sync) and Yellow ( H Sync) and I did wire the Black wire for Ground there. The small wires are 30awg Wrapping Wire.

Below is the main work done on the unit. This was before adding in the Link Port.

Above you should be able to make out most if, not all the wiring there. I did put 2.5m screws into the Link Port holes to give it stability. You can possibly make out the green 3d printed washers under the screw heads. I didn’t want to scratch the pcb, and there is one trace right there by the left screw. The revised board has that trace relocated now away from the hole.

There is not too much going on on the top side of the PCB. The picture above was before I added the screws to secure the cartridge port. It is just the port and the Pico. I am using header sockets and pins so the Pico is removable if needed, but that shouldn’t be a big deal to solder it in with the normal pin headers. If you go to put it on header sockets, you may want to test fit it in the case to make sure it will fit. For me it got pretty close to having to cut out all that area where the fan is intended to be mounted.

To add the link port I modeled a part to sit down in the remaining cavity in the bottom corner of the case. I put the 3d printed link port mounting plate in for a test fit and I could not get it back out. Luckily my chip puller was able to get in under it and pop it out without damaging the case. The part is made so that you put the Game Link port in place and then press the 3d printed part in place, it clamps the port in by the little ears sticking back on it a bit. The main thing is to get the opening made in the right place at the right size for the port. I made a template of the port shape, then drilled a hole where it would be inside the area to help me get an idea on placement and got lucky enough I had it right in the end. I used a combination of small drill bits and an Xacto Knife.

I used two of the remaining shorter smaller diameter black screws from Retroflag to screw down the port PCB (or protoboard in my case). I didn’t expect the green printed part to fit in so securely, if it works loose I can just add a dab of glue and it shouldn’t be able to move. I do have all 4 sides fairly snug and the Link Port is put out into the case plastic so it can’t lift out on that side. I used Design Spark Mechanical to make that adapter mount, it was quite easy, far easier than getting something that dimension specific done in Tinker Cad. I have to spend more time using it. I probably could have made the Ampi case in it far more easily if I had stuck with it.

The image above is a view of the DMG Link Port Mount model. The first test print worked perfectly, which is uncommon. It was a snug press fit, the port fit right into the cavity for it, the screw posts printed solidly and the holes in it are the right size for the screws. You have to put the Link PCB with the jack on it while the mount plate it loose yet, just work them in together and you should have a good fit for both. It may look like the screw posts are back a bit far, but that is because my piece of proto board is a little shorter than the PCB I designed. I made the mount to match the PCB design, not the piece of proto board I am using.

Above is are renders of the DMG Link Port Breakout Board.

The unit is complete now and tested. I used it awhile without the Link Port as I was still sorting out some things on how I wanted to install it. I also had tested the Link function and it wasn’t working due to me not knowing that there is a twist in the cable, so two of the pins had needed switched with how I was connecting it up.

I have the revised v3.21 board shown above as it is now. Like I mentioned some component designations have been changed. D2 R18 etc. I have relocated some pads. I put the Wii pads together by bringing down the 3V3. The Work Ram is lower on the new revision. I think it may now clear the center front post in the lower case body, but with it where it is on my prototype it just gets in the way a little bit. I have added a notch in the board that should allow keeping the back center hole in the case instead of cutting the post off.

Joe has pulled my Fork into his project and corrected the Left/Right Audio being flipped as of 1/20/23.

https://github.com/joeostrander/consolized-game-boy

If I do more updates they will be available in my Fork of Joe’s project and hopefully get pulled into his:

https://github.com/Markeno76/consolized-game-boy

There are a few things I may do yet. Time will tell. I’ll have to fix the schematic listed here for the Left/Right being flipped.

Full Schematics:

Follow up post :

Super Game Boy Console : Part 2 Revisions

AMPi4 – Custom Amiga Inspired Pi4 Case with Mechanical Keyboard

3d Files are here: https://github.com/Markeno76/AMPi4

Some may ask why make a Wedge case with Mechanical Keyboard with Full Size Ports for the Pi4b when there is the Pi400. I do like the Pi400 it certainly has it’s place. The AMPi4 was made for a bit different use case. I wanted a complete case with a mechanical keyboard, robust full size ports, a bit of weight to it, as well as SSD/NVME for storage and the OS.

If you look at my other posts here, you will see I like working with many of the 80s era computers. I never had an Amiga, and they are quite expensive these days. I also don’t want to dedicate more space for the storage of Retro Computers. I found after I started to get more and more Retro Computers around, that I am less and less likely to go to the effort of connecting one up and using it.

First I looked at the Pi400, which I thought may be a good option for Amiga emulation, and possibly more of the retro computers. The keyboard didn’t feel right, it was to small and too light so it easily to slides around the desk. The power adapter and other connectors are easy to accidentally pull out playing games etc. I also don’t like SD Cards for the OS as I have lost a few over the years in Pis. In many cases a Pi SD Card can quickly and easily be recreated or restore from a backup copy. For this setup, it takes a bit of work to get the OS just right, also it is to be a Living OS. I want to use it for Retro Computer Emulation, but I also want to use it for Linux with the GUI as a more modern option for basic usage and possibly some development work for various Microcontrollers and such.

I did also try Amiga Forever on my Windows pc, but it just doesn’t feel right on a wide screen LCD and Tower pc.

There are a number of Keyboard Wedge cases people have created for emulation systems. There are some for the Mister, which also I believe had a Pi option. I had seen some cases that take the Pi400 and wrap it into a larger case. I just did not find a case that was exactly what I was looking for. They were generally missing something I wanted in the system. They generally were a bit complex in design, with many parts and well they did have some style to them.

I expected I could come up with something that had all the features I wanted and be simple to print and build. It is a rather basic design. Mostly made to be easy to print, but certainly I am not that creative or very experienced in 3d modeling.

With that I looked at what I want in this setup:
1. Amiga Wedge inspired case. The initial case looked more like the Ti99/4a in profile until I narrowed the electronics compartment at the back.

2. Mechanical Keyboard

3. Full size HDMI Ports

4. Ethernet Port

5. USB Ports accessible, and more of them.

6. SATA SSD or NVME Drive

7. Internal Power Supply. This adds to it being simple to connect and adds weight to the system.

8. DB9 Ports for Joysticks (This doesn’t work quite right, I may yet get that sorted.. For now I have let that alone.)

Below is a view of the Second Revision of the case design that I printed. Some photos may be of the First Revision. There is a newer revision of the case with mostly small modification to adjust for some “mostly workable” but inconvenient items. I do plan to print that newest version of the case at some point. This part of the case is 300mm wide, so not every printer can print it. My S8 printer is 310×310, it seems that it is not possible to get the full 310mm though I got about 307mm with a correction to the alignment. That alignment issue had caused a mess with the First Revision print on the left side as it was trying to push the bed further than it actually could be moved.

Looking at the back panel, the far left hole is for the momentary Safe Shutdown Button that I put standard on just about any Pi project. It is the same type of setup as my Bartop Arcade machine. Next is the IEC AC Power port. After that is the Fuse. There is No Power Switch, I kind of would like one, but it was getting crowded. Then the 2 full size HDMI ports with the 30mm Fan beside them. RCA ports for the Composite Video and Audio output. Then a place for a standard Keystone jack, which I installed a RJ45 jack into for ethernet. Then a single USB Port. The far right side there just the place for the DB9 Joystick ports and the USB3 Hub ports and mounting posts for the screws. The Pi4 standoff screw points there in the middle of the case and holes for mounting the Meanwell 5Volt power supply.

To design the case I had to work out the parts I was going to use inside. I needed to make it big enough and get placement for all the parts. I created placeholders by measuring various components and making simple placeholders for them. Yes it would have been better to build the case in something other than Tinkercad. I would rather use something else, but that is all I had enough experience with at the time. The Keyboard Tray was started in Designspark Mechanical. I found I just wasn’t familiar enough with it, and it was slowing me down making it likely I would give up before finishing the project. I then went over to Tinkercad to design most of the case.

I already had two of the type of keyboard I chose, so I knew the how it was built. I was thinking initially to only design the electronics enclosure and attach the whole keyboard to the top of it. I decided to recreate the keyboard case “tray” instead. The keyboard tray/case is tapered a bit, making it problematic, the “feet” and “cord” openings were not optimal either.

Below is the original tray and my test print of the recreated tray. The Keyboard tray is two parts as it is well over 300mm wide maximum my largest printer can print. You can see I am at that point using the black keyboard instead of the white one. They are physically identical, just different plastic and paint colors.

My first print print of the keyboard tray showed me the post positions were off and unusable. I adjusted for that and ended up with something closer. Even at being closer, the far left two posts are too far left by a little bit. Two others there are more slightly off. The smaller right side part was just about perfect for all posts, it was a second print as well. The posts were positioned based off of measurements, which wasn’t the easiest to get perfect. I think I did measure starting at the right side, so any variance was likely to get worse going further to the left side of the keyboard. I have since revised the position of those posts.

There is an issue in that the holes are to small and the posts tend to sheer off at the base. I carefully drilled the screw holes a bit larger after printing. I would revise the hole side on the 3d model, except it is in Tinkercad, and that’s really not practical in Tinkercad. That is why I wanted to use something else to make these models. If I had worked out using DesignSpark Mechanical or some other 3d design package it would have been an option to resize thoses holes after the fact. There is also variation in 3d printers, so you may find hole sizes are a bit different anyway, but likely not going to be large enough When it is together it is very solid. Not all of the posts are screw posts, some are just supports that don’t need holes at all.

The keyboard is mounted with brass inserts. These inserts are actually in the Top not bottom, but the screws come up from the bottom. They are very unlikely to ever pull out as they are pulled in tighter by the screws from the other side. Those are the four in the top of the keyboard that rest on those posts/pillars. The inserts in the keyboard tray here are actually salvaged out of old broken laptop cases. The 4 here in the upper area where the posts are did take longer screws I believe. I have a selection of Metric m2.5 screws of different length. There are also 4 more along the bottom edge of the keyboard as can be seen as holes without the inserts in them in the first keyboard photo. Those four lower ones are short screws.

Above is the original black keyboard when it was being installed.

Above you can see the various internal components test fit into the chassis. The USB Cable was modified already. The first revision the DB9 ports didn’t fit, the fan had an integrated grill that printed awfully and blocked to much of the fan.

Above most of the internal wiring is completed. The DB9 Joysticks ports are not yet wired to the GPIO and the Fan isn’t wired in yet. It is quite tight in there. The wires on the GPIO are going to the two LEDs on the top cover. Power and Activity lights.

Above is the second test print of the corner of the case as the DB9 ports didn’t fit and the USB Hub wasn’t properly positioned. The little black cable is the cable I made up for connecting the USB Hub. Due to the limited space those cables are made up with usb connectors without the shell covering them. The bottom cable there is the short keyboard usb cable, it plugs into the bottom of the keyboard where that square opening is in the keyboard tray.

The 5V case fan is powered that little buck converter there between the power supply and HDMI port. The fan was louder than I liked, so I used the buck converter to lower the voltage a bit to make it quieter. You can see the 2.5″ SSD Drive is in the case. It also still has the 2gb Pi4 in it in this picture.

The DB9 Joystick ports were very problematic. The first option for using them didn’t work at all, well it made my Power and Activity LEDs not work. The other options I tired required rewiring them to other GPIO pins. They were unreliable. It would have incorrect button presses as the timing must have been off for some reason. I was initially only going with Atari 2600 or the Commodore 64 3 button hybrid, but when those didn’t work I went with another option that was to have supported the Sega Genesis 6 Button controllers. I am not sure if the issues can be sorted. I have went to building and using a USB DaemonBite Retro Controller adapter setup for the Sega Genesis 6 button controllers. It even seemed to have faulty timing issues with the Genesis Controllers. I found a fix for that which seems to be working properly now. The DeamonBite adapter can be seen in the picture below. It is the more or less triangle shaped blue box to the right side with two DB9 ports on it. It is actually a simple build with an Arduino board controlling it. The case was available on Thingiverse. I want to go back to get those DB9 ports working off the GPIO. I have thought of putting an internally connected DeamonBite controller and wiring it to the case’s ports. That would loose me my USB2 port in the back though.

The system runs Raspbian and Retropie. I have it setup to boot into Retropie, but I can exit that and go into the GUI and use Raspbian. The main intention of this system is to run Retro Computer games. I worked to set it up primarily for Commodore 64, Atari 8bit, Ti 99/4a and Amiga emulation.

It was setup with at 2.5″ SSD Drive, and I ended up switching to an NVME drive later on. The keyboard was the cheapest version of the Redragon K552 87 key keyboard which was the Black version with Red Switches. I don’t like Red Switches. Once I was confident that the project was going to work out, I purchased the more expensive White version with the Brown Switches that I prefer.

The newest Revision 3 main case has various changes. Mostly small alignment changes. The Pi is moved to the left because the 3.5mm jack couldn’t fit without the outer plastic cover on it. I may not have moved it enough, I am not sure as the space is very limited and I have not reprinted it yet. Some other small adjustments to the back ports. I also changed the “vents” below the keyboard in the main frame as with the current shape they didn’t print very well.

I later found out that the key switches on the keyboard are removable. They aren’t quite “standard”, the contact pins on the bottom are smaller than standard switches. They probably won’t make contact in another keyboard. Standard switches won’t fit into the keyboard without filing the contacts down to be narrower like the ones that come in it. It is possible to do, I replaced several of the switches in my son’s keyboard as W A S D etc were worn and not working well. I will see if the Key Switches that Adafruit sells hold up better than the factory switches did.

STL 3D Model Files are here:

https://github.com/Markeno76/AMPi4

Main Parts List:

MEAN WELL 323282 RS-25-5 AC to DC Power Supply, Single Output, 5V, 5 Amp, 25W, 1.5″ : AmazonSmile: MEAN WELL 323282 RS-25-5 AC to DC Power Supply, Single Output, 5V, 5 Amp, 25W, 1.5″ : Electronics
Redragon K552 Mechanical Gaming Keyboard RGB LED Rainbow Backlit Wired Keyboard: https://smile.amazon.com/gp/product/B016MAK38U/
USB Hub 3.0 – Rekidm 4 Port Aluminum Desk USB Hub 3.0 Clamp Design: AmazonSmile: USB Hub 3.0 – Rekidm 4 Port Aluminum Desk USB Hub 3.0 Clamp Design for Desktop, Computer, PC, Table Edge with Durable Adjustable Clip, Space-Saving Mountable USB Hub Fast Speed Transfer : Electronics
HDMI Coupler 4K*2K HDMI Extender Female to Female HDMI Adapter Connector Coupler with Panel Mount Holes: AmazonSmile: HDMI Coupler 4K*2K HDMI Extender Female to Female HDMI Adapter Connector Coupler with Panel Mount Holes HDMI Jioner Kit (5 Packs) : Electronics
Micro HDMI to HDMI Cable, Ultra Thin HDMI Cable 35cm/1.15Ft, Super Flexible Slim: AmazonSmile: SMALLRIG Micro HDMI to HDMI Cable, Ultra Thin HDMI Cable 35cm/1.15Ft, Super Flexible Slim High Speed 4K 60Hz HDR HDMI 2.0, Compatible with GoPro Hero 7/6 / 5, for Sony A6600 / A6400-3042 : Electronics
IEC 320 C14 Panel Mount Plug Adapter Power Connector Socket: AmazonSmile: 10 Pack AC 250V 10A IEC 320 C14 Panel Mount Plug Adapter Power Connector Socket Black Screw Mount 3 Pins Inlet Power Plug Socket by MXR : Electronics
Fuse Holder 250 VAC 10 A 125 VAC 15A: BOJACK 10 Values 30 Pcs 5X20mm Black Screw Cap Fuse Holder 250 VAC 10 A 125 VAC 15 A with 10A/15A Fast Blow Glass Fuse kit – – AmazonSmile
30mm 5V Fan (The optional buck converter was used to lower the 5V to make the fan quieter)
16MM Momentary Push Button
Pi4b
USB 3 to NVME (or USB 3 to SATA adapter depending on the drive used)
NVME (or 2.5″ SATA SSD Drive)
2 DB9 Male Adapters
Dupont 2.54mm cables or connectors and wire.
2x3mm LEDs (and 220ohm Resistor)
RCA Jacks
RJ45 Keystone Network jack
USB A Single Panel Mount Port
USB A Male solder connectors: Amazon.com: Maxmoral 10PCS USB 2.0 Connector A Type Male 4-Pin Plug with Black Plastic Cover DIY Connector : Electronics
Provided the prints are successful, the 4 printed parts take most of a 1kg spool of 3d printer filament.

There were various other supplies, M3 Brass Inserts for the top cover, some M2.5 Brass Inserts for the keyboard mounting. Various M3 and M2.5 screws of different lengths.

Lenovo Tiny PC Mount

I have a Lenovo Tiny PC M600 Series that I wasn’t sure what to do with. It is a Pentium based CPU so not the best to use for very much. I have my Desktop PC on the other side of the room behind me when I am at my electronics bench. I’ve been wanting to put something over on the bench for awhile. I had various thoughts of using a Pi4 or something, but they are at such a premium right now I feel very doubtful I’ll get my hands on another one. They are not worth what they are currently selling for, if the Pentium Tiny doesn’t work out, I could buy an i5 Tiny for the same as the Pi4 these days.

Lenovo has made VESA Mount adapters for the Tiny, but I decided to check for a 3d printed mount design. I found a couple, but most were not what I was looking for. I picked one out that was close, but after taking measurements for some reason it was too narrow. All Lenovo Tiny’s I have had over the years have had the same dimensions.. So I find that odd.

I started over and came up with a one piece design that I liked, but it was a big waste of support material. I then came up with a two part design to save material and not need any supports. The first test print warped rather badly due to part of it coming loose from the bed part way through. It was good enough to test my dimensions, which needed some minor adjustments. I made a few other changes and printed it again, this time it stayed to the bed properly. The Tiny also fit it perfectly with the changes. I did make some one piece designs that take minimal support material as well. Since we had another model Lenovo tiny in the office, I decided to test fit that model. It does not fit due to the placement of the Feed on the case, I had added some material to stiffen the corners and feet for the second tiny are placed closer to the case edges. I figure I will modify the designs to accept either model before I call it finished.

Today I installed the M600 on my workbench with the Two Piece mount.

The mount files can be downloaded here:

https://www.printables.com/model/301500-lenovo-tiny-mount-m600/files

GBS Control

I came across a video on the GBS Control upgrade project for the GBS8200 boards. I had been interested in the GBS8200 back when I was doing my Commodore 128 RGBi/CGA to RGB video adapter. I had went with a SCART to HDMI adapter, as at the time that looked to be a better solution. It seems that device puts in a lot of delay though. It is useful to get “something” out of old RGB devices onto a HDMI compatible display, but when doing something like playing games the delay introduced is bad. I had not previously seen anything on the GBS Control project. The reason I had not used it at the time, was the SCART was to HDMI, where the GBS8200 outputs to VGA. The GBS8200 was a bit questionable too as it wasn’t super well liked in some circles at least.

GBS Control Project
https://github.com/ramapcsx2/gbs-control

Voultar’s Build Video
https://www.youtube.com/watch?v=1AVXhiTlmgo&t=2382s

GBS Control Case I used

https://www.thingiverse.com/thing:4946190

There is now the GBS Control project, which involves wiring in an ESP board to take over control of the board. This greatly reduces the delay introduced in the output. It accepts input from 15kHz (or 31kHz) RGB/HV (VGA), RGBS (RGB Sync often wired with SCART or 4 RCA jacks) and YUB (Component) and outputs to regular 31kHz VGA. There are DAC nonbuffered based VGA to HDMI adapters that are available very cheaply as well that are said to not introduce additional lag.

It has various features beyond near no lag being introduced. It has Scan Line emulation, various resolution outputs, downscaling video from higher resolution to lower resolution output. The controls are all onboard the ESP web interface, which you can use by using the default broadcast it does, or link it to your wifi network. It also lets adjusting position, size and various other things with the output.

In my case, I went with a compact case that was based on the build instructions I followed from a video by Voultar that introduced me to the project. With the case I used I don’t have the RGBS input accessible which is wired to a SCART port. It does seem though RGBS input is an option into the 15pin VGA style port with Sync on the HSync pin. HV means Horizontal and Vertical Sync, where RGBS just uses a single Combined Sync or CSync signal. The case turned out great, it is a near perfect fit.

I made a few changes compared to Voultar’s instructions. First, I did not add the resistor to the Sync line for the RGBS, as with it that affects the RGBHV input from what I read, so if I did do it. I would want to wire up an input option to switch between it being in place and disabled. On going over what that is about, I believe it is to be there when devices use TTL Sync a 5V Sync Signal instead of the 1V max Sync that RGBHV generally uses. The next change I made was when soldering in the ESP board I used angled header pins for the two ground connections, that provides a much sturdier connection between the two boards than a piece of copper wire or solder only. In most cases either just solder, or at least using a bit of wire will likely be just fine though. I like that if it gets dropped the pin headers will be a lot sturdier. I also connected the power for the frequency generator to a different capacitor, Voultar had misspoken what the polarity was in his video while wiring up the power, that let me wondering if he was wiring it correctly or not. I decided to then go with how the project instructions at gitub showed it wired in. Voultar’s wiring was a little shorter for the power, but it was also a little harder to get in at it with the soldering iron. For all of the power wiring I used heavier wire, and for the various other low power wiring I used solid core 30awg wrapping wire. I found that I had issues with the solder points on the two chips with the adjacent pins getting bridged, and had used some desoldering braid to get rid of the bridges. More flux would have likely helped me avoid bridging the pins. Voultar has certainly done that kind of thing far more than I have. Thankfully there are only 4 solder points to various chip legs. The capacitors were easier to stack than I expected and they turned out very well.

Mostly I want it to use the GBS Control with my RGBI to RGBS or RGB/HV converter project I posted previously for the Commodore 128 80 column RGBI (CGA) mode.

CGA/RGBI to RGBS Adapter: CGA/RGBI to RGBS Adapter Post

I have more recently made an update to that projects with a PCB and 3d Printable Case design.
Part 2: CGA/RGBI 2 RGBS PCB Design, Schematics, BOM

Part 3: CGA/RGBI 2 RGBS Board Assembly and Testing

Part 4: CGA/RGBI 2 RGBS Adapter Case and Release Files

I want to make a 5V output port for the VGA to HDMI DAC adapter at some point. I may do an audio passthrough option going in the front and out the back to connect into the adapter as well. I had thought of a power switch, but I use power strips for my equipment and generally turn them off when not in use. I have thought of a power LED as well, but still not sure on that one either. There was a version of the case that included putting the VGA to HDMI DAC adapter in the case. That was setup to remove the VGA output though, which I didn’t like. I have a number of VGA monitors I like to use from time to time.

For my first test of the unit, I connected it to the VGA output of a laptop. It passed through the VGA signal to the VGA output just fine. I then connected my Commodore 128 and my Digital RGBI to Analog RGBS / RGB/HV adapter. I built that adapter with the GBS8200 board in mind, so while it is setup it as RGBS + Audio through to the SCART connector I have configuration jumpers on the board. I changed the jumpers to NOT put 5V to the one pin which was to be the SCART mode detection. I also changed the Audio to not go to the HD15/VGA type connector and instead go to the RCA jack on the side. I changed the jumper to RGB/HV mode where it bypasses the Sync combiner circuitry on my adapter and outputs H Sync and V Sync. When connecting it up it worked great, I just adjusted some settings in the web interface to see how that worked out. The Scan Lines look pretty nice on it. I did not have any 80 Column mode programs prepared to use on the Commodore 128 to check it out too much at this time though.

This GBS Control is a baseline model build, more or less, for a bit more complex build check out the newer post linked below when I built a new unit and upgraded the one above to the new specification.

I have built another GBS Control and updated this GBS Control with some more connector options and features.

You can view that post here: https://hobbytronics.home.blog/2023/12/18/a-new-improved-gbs-control-and-upgrades/

TI 99/4a Recap

I figured I would make a post on the TI 99/4a I picked up awhile back. It was one of my first computers. I didn’t think of it at the time as a “real” computer though. I don’t even remember when I got it, but it was older and not current at the time, probably the early 90s. I wanted to get one and tinker with it like I have been with the other old systems from the 80s.

Really I am not sure I would recommend it yet. The system is so limited, I hate to invest in it. With a Commodore 64, short of a drive solution to get to and save data the Commodore 64 is complete. With the Ti if I want to do anything interesting, it seems I have to get a 32k ram expansion for it. That or stick to Basic and original cartridges. Many of the cartridges made now look like they require the 32k Expansion to operate as well.

Anyways I did get it, and I picked up some cartridges and even a tape player that I will look at in the future.

I did check the system out when I got it and took it apart. Inside it looked pretty good. The keyboard doesn’t work properly, it is one of the membrane types.. I am afraid also if I want to do much I will have to find a donor keyboard for it.

I did buy capacitors for it, there weren’t many.

Capacitors for the Mainboard. A few spares as well.

Initially I wasn’t going to get out my desoldering iron, but I quickly changed my mind before even getting the first capacitor out. I use an ECG J-045-DS desoldering iron. It is a rather cheap device, but it has done very well for me. I do keep thinking about and looking at those fancier units like you see so much of on Youtube. The ECG is a quarter of the cost of even the cheapest of those though and has done well for me so far. I don’t get it out much, mostly it takes awhile to get heated up and does take two hands to operate.

First was the power supply, with 6 capacitors. I cleaned the power switch with some Tuner Contact cleaner while I had it open, initially I wasn’t sure if it was making nice clean contact. I also replaced the LED as it was so dim that it was hard to see. The original LED had longer legs than typical, so I extended them with a piece of solid copper wire and then put heat shrink to insulate the legs. The original LED also had some heat shrink on the negative leg. I tested the power supply and it worked properly afterward, and the LED was much brighter, but not too bright. I also tested with the 99/4a board attached and it all worked fine. The voltages tested good as well.

I did add the zip tie to the largest capacitor after the picture was taken, although it was quite secure already.

Have to crack open the board. Two holding clip things and three bolts to remove this shield.

Then was the mainboard. That has 9 total. There is a single 10uF 35V, there are five 22uF 25V, and finally three 100uF 16V. To replace them I did get the five axial 22uF caps, but the 100uF I could only get in radial from the supplier I choose. The 10uF I couldn’t get an axial cap either so I went with a Tantalum instead, I could have picked up a radial 10uF though instead. When I did the Timex Sinclair 1000 I ended up getting some Tantalums for it as well.

Here is the board before. The Heatsink compound on the video chip there is replaced as well.

The worst thing of switching out from the radial capacitors was making sure I had the new ones installed correctly. Even the Tantalum is a polarized capacitor. The Tantalum marks the Positive, where Axial Electrolytic Capacitors mark the Negative. Then Radial Electrolytic Capacitors have a stripe with an arrow on it that points toward the Negative. So I did a lot of triple checking before powering up the board. I also checked my pictures of before and after to make sure I didn’t flip anything. That is why it is good to take pictures of these things. I then cleaned up the flux from the board and hooked the bare board up to test.

It didn’t power up properly after putting the capacitors in. That was an awful feeling, this is why doing such things are questionable. I checked and rechecked the board, I looked at each capacitor, checked if any were warm and possibly in backwards. I checked all solder points. I pushed in the socketed chips. Then I checked over the board again. I found a few flecks of loose solder, and one of them looked like it may have been shorting two points on the board. I checked again to see if there was any more on it then reconnected the board to the power supply and monitor. This time it came up normally.

I put it back together and checked it, well the keyboard is still awful, but it is working. I can play some Parsec just fine, well the joysticks are still awful. So it seems I have some more work to do with this one.

I pretest the new capacitors before installing them to make sure they are in spec. I also test the old ones I check how close to spec they still were. In this case I think all of the original ones are likely just fine. I just use one of the cheap atmega based component testers.

Still in the end, I had to open it for the new LED and well I say that was worth getting in place. It looks so much better with the new LED. Sure doing just that would not have been nearly as complex a task though. I also replaced the thermal compound on the video chip as well and do recommend that. The RF Shield was reinstalled the same as original. There is the later QI board that apparently doesn’t include it, or part of it at least.

My next project was a Joystick Adapter to use the Atari 2600 joysticks with the 99/4a. It is a simple enough project just taking some connectors and diodes. It did give me a bit of trouble though initially. Possibly a solder bridge issue again, or simply an issue with a poor connection at the TI joystick port, or an issue with not cleaning the flux from the board properly. I have it working now, I am just waiting on some more DB9 connectors to finish it up. I did also make a cartridge with a new pcb and eprom that I may mix in with some post at some point. I am really not quite understanding the way to use the new cartridge boards properly yet. I did get the homebrew remake of Pitfall on it, but I am not quite understanding how to set them up for different size roms.

Remaking Commodore 64 Cartridges

I ran into an issue with some old Commodore 64 Cartridge Roms that I purchased. I found 4 Rom chips with a single PCB for sale and purchased them. The PCB had a IC Socket on it to swap the chips around, all four chips worked on the board. I wanted to get them all usable again as full cartridges. I purchased some 3d printed cartridge shells at some point after that. There is also the issue that with the cartridge shells you can not use an IC Socket, it makes the chip sit too high to close the shell.

I decided to look around for some PCBs, but I couldn’t find any with the proper footprint for the original ROM ICs that have the 2364 pinout and not the pinout of the now commonly used 2764 27xxx EPROMs that the Modern PCB designs use.

I had not yet worked on any PCB designs. It is a lot of work to go from scratch for something like this if you aren’t familiar with it. The main issues are recreating the physical board in the right dimensions, the edge connector, as well as the proper size and placement of the hole to fit the case shells. With those in place it would not be too bad.

The original board after removing the socket and putting one of the ROMs on it.

I did find a project on Github with Eagle design files. I tried to start working on it, but just wasn’t getting it at the time. Part of my issue is I have four ROMs and one board already. I only needed three boards. I didn’t want to order in 5-10 PCBs to only ever end up using three of them. That would have been quite a waste. I expect to that I will not be getting any more 2364 pinout ROMs in the future.

It has been a good while since I put those chips in some ESD Foam for storage. I recently purchased some premade PCBs for other projects, which you can find the GAL PLA replacement post here which is one of the boards. Another of the boards I purchased was a 2364 to 2764 adapter board. The GAL PLA was so easy to put together as well as the 2364 adapter that I really wanted to get back to this project.

I looked around and had seen some of bwak’s stuff. I found his Versa64Cart over at Github.

https://github.com/bwack/Versa64Cart

It looked like a great candidate as it was available with the Eagle files. I haven’t used Kicad and the other project I found only had files for it. I wasn’t up to learning another program just for this project. Bwak’s design is also the most complete and has all required documentation.

The only thing I really had to do with bwak’s design was add the 2364 footprint beside the 27xx footprint and wire it up properly. I was careful and used the design for the 23 to 27 rom adapter as part of my reference.

https://github.com/SukkoPera/Open2327RomAdapter

I also did some other reading of cartridge schematics and reverse engineering the original PCB I had received with the ROMs. In the end I put on the 2364 footprint, I tried to use an existing Eagle Library that had it in it, but there was some issue with the footprint in it. That lead me to making a new footprint for it in the library using the standard DIL/DIP footprint. It is the oval pads rather than the minimal round ones that bwak used for his footprints though.

Here you see the 2364DIL24 added to bwak’s schematics

I also added a jumper between EXROM and IO2 to the board. The only reason I added that was because the Original PCB I have has them wired together. I have seen no other point at which those are referenced as being wired together. These specific ROMs are set for “GAME” and “ROMH”. EXROM is not used on these cartridges. I have found no reference of IO2 or IO1 used on any standard cartridges, maybe they are used for bank switching cartridges? If that jumper was connected AND “EXROM” was tied to ground, that would Ground IO2, which is probably not good. Beyond those changes the board is the same as bwak’s 1.5 design. This will let me use my remaining boards with some 27xx Eproms or EEproms.

In the end I ordered 5 boards from JLC PCB. You can see the IC2 and JP1 footprints below on the Gerber viewer.

The PCBs arrived 7 days from ordering them. That was manufactured, packed and shipped from China to the Eastern United States. They do say two weeks estimate, it was very impressive to get them so quickly. I got them for a bit under $20 for the five boards shipped to me.

Front and back of the board after it arrived.

If you look at my board compared to bwaks’s 1.5 the ground plane is different, I think I widened the gap between traces, as well as the additional footprint for the other socket is probably blocking some areas. The old boards had no ground plane it is not a big deal. I usually like to fill it out as much as possible though, for this it is not important.

Here is the first one I assembled beside the old board.

When they arrived I tested them against the original PCB. Everything checked out, the boards all looked correct with no defects. I did have to round over the card edge connector though, I expected that. I didn’t want the sharp edge there going into the C64. It was easy take the edge down with some sandpaper on a sanding block with a few passes across it.

The next thing I needed to do was solder on one of the ROMs, the capacitor and do the solder bridges for the GAME and ROMH pads. Initially it did not work, but that was because I forgot to do the GAME solder bridge. I did that and it worked perfectly. For the next three ROMs I soldered one onto the original board and prepared two more of the new boards for the last two. They all tested out and worked properly. The next thing was to put them into the 3d printed cartridge shells. That was easy enough, I will say compared to the old PCB, which fit perfectly into the printed shells that the new boards are slightly different. They are maybe .5mm to wide to fit, the hole for the screw must also be about .5mm to high or maybe even 1mm . This meant I had to slightly shave the 3d printed shell to fit the pcb into it, not a big deal. The hole placement means they are sticking slightly out of the bottom of the cartridge, but part of that is these cartridge’s have the screw about 1mm to low making that difference between the old board and the new ones a bit more pronounced. The old PCB even is about sticking out of these shells.

Here are all fully soldered up and ready to close up.

I am working on modifying a 3d printed shell design for my remaining boards, as those are the only ones I had. I had purchased them specifically for use with these four Original ROMs. I do have an Ender 3 Pro and can print them myself now. There are a number of designs on Thingiverse and other sites. So I picked up a few designs and started tweaking them with Tinkercad. The primary design I started with is a Stumpy type, it perfectly fits that old factory PCB, but that is the tallest it will accept. That is fine with me. It looks cute and takes up a bit less space, is quicker to print and takes less material. The problem with it is that the screw hole is was not placed properly, and the diameter of the standoff is incorrect, is also lacks support to keep the cartridge from rocking back and forth along it. I worked on it to get the screw hole placement correct, as well as fix the other issues with the standoff. I had placement corrected, and the length of the pins was great, the problem being the shaft was too large.. why.. So back to working on it. I did get it downsized properly now. It will use a M3 screw, so I reworked the face to accept a threaded brass insert. If I used the brass insert I can use M2.5 screws as well. The revised case now prints well and fits the cartridge PCBs nice and securely in the proper location.

Here they are closed up with my second test print of the Stumpy shell

The short shell is going to be for the remaining boards and possibly a few other modern cartridges. The original cartridge board fits it perfectly and is the longest that will fit in it. The Versa64 board a bit shorter so they fit well. I was thinking of making an even shorter version, but then it may get to be difficult to remove from the computer. I am thinking of maybe making a post about the shell if I do something interesting with it. It does not have the removable nameplate, which it really can’t as the screw mount is on the back of it. The cool thing about the removable faceplate on the shells I ordered is that a custom plate can be made to put in them. I had also purchased a shell for my Dual C64 Diag/Dead Test Cartridge which as a customized nameplate and opening to make the switch accessible.

The last thing I did with these cartridges was print up some labels with my Brother PTouch Labeler.

I thought of doing some printed labels on my inkjet printer, but they tend to fade as well as smear the ink if they get wet/damp. I have also done reproduction labels for cartridges using either label paper with the clear packing tape over them to protect them. Those sometimes the adhesive fails but it does usually hold up, the problem is more likely the packing tape adhesive fails and it starts to come loose. I have done it with paper and spray adhesive, but then after a couple years the spray adhesive has failed and the labels started coming off. Maybe I didn’t use the right adhesive spray. I did use clear laminating tape, and that seems to hold up well and stay on the label, but due to the spray adhesive they haven’t been on that long. Also the laminating tape I used is not glossy and is slightly hazy. I have also recently used a matte finish inkjet printable vinyl (see the RAD REU Cart). I feel that should hold up well, it isn’t quite the finish I like though it is durable. I mean the vinyl won’t have the exact look of an original cartridge label.

The PTouch labels have nice gloss finish and they stick very well. I expect them to hold up well. I have used them up for years for various projects as well as for work. They look nice for what they are, but are rather limited in the “art” and styles available.

You can see they aren’t any high value Cartridges. My projects are often about learning to do something new, or get better at some things I have done in the past. I like the bonus of getting something useful in the end. I may not use these games often, but I am glad that it will be easy to do so now. You can see how close the PCB is to the cartridge’s shell. Radar Rat Race is actually the original pcb, and even it is sticking out a bit. If I cared, I could sand down the new PCBs a bit shorter to be a more proper fit, but I don’t care that much. The new modified shells I have been working on are modified to fit better as they are.

This is just another quick easy project that I did. It is so much easier to put together a project with a proper PCB. I had toyed with the idea of designing and etching the boards myself. The ones I make were a very delicate and complicated project. I really don’t want to do any more of them anytime soon. Being a double sided board makes that much more particular in alignment. Cutting the boards correctly, drilling, etching… It is a lot of work, the lack of solder mask and through holes make it far more complex to assemble (not to mention not having the cartridge edge connector plated properly). I also design the boards with more space between traces when etching the boards myself to allow for the difficulty of keeping them intact and getting them to separate properly during the etching process , as well as going with as few through holes as possible.

I certainly expect to look into getting PCBs manufactured for future projects where required. The tough thing is for one off projects they are rather wasteful. I don’t want to buy 5 of them and only ever use one. Prototypes often end up with some mistakes, so I might order 5 and not be able to use any, or use one of them and have to rework on the board making it a bit of a mess.

Just a little addition. Here is a further modification of the Stumpy Shell shown above. The PCB is nice and flush in the shell. This is a C128 Dual Diag cart I have made up with one of the spare PCBs using an EPROM. The switch insert makes it reachable from the outside and the little red reset button extension makes it accessible. The EPROM is in a socket for this cartridge so I needed to make an opening for it. I find it quite difficult to get proper measurements for making the openings. There is a bit of trial and error making some test prints. An easier way may be to export the pcb design either to a scaled image file to use for a background or export it as a 3d model, which I did for my RGBI to RGB adapter pcb when making the case for it.

Another Pi1541

I now have three working Commodore computers. Two 64s and a 128. I felt I wanted to have a second Pi1541. I didn’t have a non repairable 1541 laying around this time though. I do now have an Ender 3 Pro 3d Printer though and a little experience in working with it and 3d models. This gave me the interest in making a much smaller Pi1541 that looks a bit like the 1541ii. I figured that styling fit better with my Commodore 128 as well.

I didn’t need another Tapuino. There was talk of Steven adding .tap support to the Pi1541 back when I built my Pi1541 & Tapuino. That seems to not be making any real progress though so far, but it may yet happen. It may have and I missed it. Steven White and any other contributors have done a great job with the Pi1541 though. It is a great thing to have around.

My first Pi1541 case had so much room that I wanted to make more use of it, it has a 7″ Composite LCD than can be toggled between the Pi and an external input, an internal mono audio amp that can be toggled from the Pi (with a simple mixer to mix the stereo from the Pi) to an external input, it also has access to the USB, Network and HDMI from the Pi, as well as a bare Tapuino. It has controls on the front panel for the Pi1541 and Tapuino functions, as well as the SD Cards for both the Tapuino and Pi. I can take it and one of my Commodores and use it as a portable 7″ monitor with Audio for the Commodore 64 or 128. I can swap the SD Card and use the 7″ LCD for the Pi and run any other Pi OS, or connect the HDMI to a TV and use another Pi OS be it RetroPi or Raspbian etc…

This one was built to be more like the common Pi1541s out there. To be rather minimal in size, but do the job.

There has been a problem for me and the Pi1541 in that the Pi3B+ has gotten expensive with the release of the Pi4. I guess it is supply and demand. The Pi4 has a different architecture so it can not work as a Pi1541. I am betting there are quite a number of other projects that probably are in the same situation. Steven did come out with a Pi Zero version for the Pi1541, and as I was going for “Small”, and these are cheap I purchased one for my new Pi1541. Then I found out you have to overclock it, and it doesn’t support all of the features that the Pi3 does due to the more limited cpu and ram it has available to it.. So that idea was scrapped. There are other Pi3s though that are still more reasonably priced and support all of the same features with the Pi1541 as the Pi 3B+. So that brought me to getting a Pi 3 A+. They are slightly less powerful than the 3 B+ model, but they can still be purchased in some places for $25.00. Granted I can currently get a 3 B+ for $35 at http://www.adafruit.com which is where I purchased my Pi 3 A+ at for $25. Maybe there is less of a shortage right now. I still have a hard time spending $35 even when you can buy the more powerful Pi4 2gb for the same price (and the 1gb model for less)..

The next thing was to find a case design that I liked. I don’t like having bare boards laying around. I also don’t like simple block cases a lot of the time. I came across a case design on Thingiverse that I mostly liked.

https://www.thingiverse.com/thing:3127040

Like I said I “mostly” like the case. Mike from thegeekpub used this case for the Pi1541 he posted a video on. It is styled after the 1541ii. It is listed as a “work in progress”, and kind of is. I made quite a number of mostly small changes to the case for my needs. Some were functional changes, some were cosmetic. The case is a 4 part case in the current design, the pictures at Thingiverse don’t all reflect that.

My prototype print. The only change being the rectangle LEDs at this point.

So let us go with the cosmetic first. I wanted Rectangle LEDs like the real 1541ii. My skills with 3d modeling are limited. That was one of the points of this project though, it was to help learn a bit more in that area. I use Design Spark Mechanical for 3d modeling. I have used Tinkercad for past simple models or modifications, but that certainly wouldn’t work with this project. My first attempt on the LED openings was to take the round holes and close them and put the rectangle openings into them. I managed that and made a prototype print of that for fit etc. I learned a bit with that and what I needed to make more adjustments with. I had to make the openings a little larger to accept the LEDs, also even then a bit of filing was done to the final print. The LCD wouldn’t quite line up with the LCD opening. The later “final” print I managed to move the LEDs stacked to the left edge like I wanted to start with. They were a bit easier to fit, but still took a little filing of the openings, it is hard to print a sharp enough corner.

Here you can see the LCD alignment issues a bit.

I enlarged and relocated the LCD opening a bit the alignment wasn’t quite right for my particular LCD. It may not be quite the same for all of these types of LCDs.. I also angled the top edge, as the depth that the LCD sets back made it so that I would have made it even taller making it odd to look strait on at. The other change I made to the face was cut a slot into it where the “disk” would have went. I like how that looks better.

I did not use a Pi Hat PCB, so in the back I needed to close the opening and make a circular opening for the DIN socket I used. I did that by adding on to the narrow “center” part of the case. Closing it was fairly easy, making the proper opening then for the Panel Mount DIN socket and screws was done as well. To provide some support when inserting the socket I put a lip on the inside of the top cover. This does not protect the socket when pulling it out though. If I had put the socket in the Top of the case, it may have been stronger but would have been more difficult to work with.

First prototype on the back opening. I didn’t make the “tab” tall enough. I don’t yet have the inner “lip” to help support the socket.

The last part was the bottom of the case. With that I made a change to one of the standoffs, with the Pi3A+ it had a component on the bottom hitting and pushing it off the standoff. I didn’t want to damage my Pi.. I think a Pi3B+ doesn’t have something quite as close to that spot. I closed some of the opening in the bottom where the SD Card opening is, I didn’t want it any more exposed than required, it was only a minor change. The more important change was on the side where the Power, Audio and HDMI ports are. For one I closed the areas up a bit, raising the bottom up, the top down, putting in a bit of a panel at the one place that didn’t have one. I also had to widen them, I couldn’t get my cables in at the ports, and it wasn’t because the openings were shorter now. Maybe some cables are a bit more slim than mine. My final intention was to take some thin plastic something like a transparency and cut the USB, HDMI and audio/video jack out to use as a backer to close the rest of the opening and not block access to the ports to make it a bit cleaner. I haven’t gotten around to that. I did do something like that on my large Pi1541 for the one port in the back.

Here you can see the Pi3A+ installed in the prototype case. There is room for a Pi3B+ but the usb and network ports would not be accessible. Those little blocks on the right side that do not exist in the upper center part are areas I added some material on the bottom part to make the openings look a bit neater to me.

For the inside I did go with hand wiring everything using some protoboard. The front control panel being fairly easy to do, short of alignment of the LCD. The buttons as designed work great, it is an great design.

Here you see the back of the control board in the front. The LCD and 5 buttons are on the front side.

For the Pi1541, well I make what I call an Option B+SRQ design for the Pi1541.. This is Steven’s Option B plus another Level converter and wiring up the SRQ lines. I will have to update this post with the schematic. You can find the schematic though that I used on my Pi1541 + Bare Tapduino project. Just note that one includes a Tapduino in addition to the Pi1541 section. I can’t find an Option B+ wring diagram, it has been unfortunately rather vague on how the later additions to the Pi1541 are wired up. For my reference I did modify the Option B wiring diagram to be a Option B+ but as I didn’t make the original diagram I won’t post the modified one. The SRQ is required for the high speed emulation of the later Commodore drives with the 128 I think.

I don’t believe my schematic is the best to follow. I think a Wiring Diagram is generally easier for someone to follow. The schematic below is from my Pi1541 + Tapuino The SV2 header goes to the Pi to various pins. I am sorry it is not the best to follow.

To make the main Pi1541 board section, I used another piece of protoboard with my required IC and two level shifters. Also the filter capacitor and resistors etc. Everything connects back to this board, then this board goes off to the Pi3A+.

Here is the start of the main interface board. It had a lot to add yet.

Below is the completed electronics. The White 3d printed parts were the final parts I made. The front being the LED placement, LCD Opening and Slot changes. The middle ring with the back tab at the proper length. The interface board is mounted with the two screws that hold the middle ring in place as well.

Below are pictures of the completed Pi1541.

Final front. There are some 3d printing minor defects.

Here you can see the LCD Alignment is much better that the prototype.

The port in the back with the tab height corrected, it also has the supporting inner lip on the top cover. Still with the yellow bottom. Then the odd bonus of seeing the Pi LED in the back.

The side view with the yellow bottom. The USB and HDMI are just a bit narrow. Yes the PLA case is a bit translucent.

Might as well have the top view too. Yes that is my Pi1541 schematic in the background.

I figured I would do some size comparisons to my other 1541 drives.

Here is my 1541ii drive with it. I used the same Green and Yellow LEDs. I have seen other 1541ii with different colored LEDs though?

Here it is my 1541, and my Pi1541 + Tapuino. It is far more compact.

Does it work? Yep.

I have printed a white bottom for the drive and installed it. The difference being the color and slightly wider openings for the USB and HDMI ports. Over all this Pi1541ii design should be easy for someone to build for the most part. With a Pi1541 Hat that fits the alignment of the original case, just print the original center part without the tab added. It can accept a Pi3B+ as well as the A+ just fine. Then the only bit to custom build is the front panel board, sure a pcb could be designed for that to even make that easy.

If anyone wants the modified 3d models I should be able to get them to you. If there is interest, I would probably look into the option to post them as a modified design on Thingiverse. I have never looked into that though and do not know what is involved. I don’t think the support rim on the top cover should get in the way of the Din plugs in the back if they line up properly. The main modifications were the tigher fit on the bottom, as well as the changes on the front of the case.

You can check out the Pi1541 site for the basics and the wiring diagram for the Option B design. It is just adding the SRQ lines with another level shifter, I tend to label it as “Option B+” or “Option B + SRQ”. There are people making pi1541 hats that include the SRQ lines, but I have seen no official diagrams posts, they are just doing it the way I am not that I am the first to do it. The SRQ I believe is needed for the 1581 Commodore 128 high speed transfers. All of the required pins are labeled on the GPIO header in that diagram. There are what I call the Option B Plus Pi Hats out there for sale. That is why I don’t understand why there has been no Option B Plus wiring diagram posted.. You can also now use a Rotary Encoder to replace the UP/Down and Select buttons. I like the idea of the Rotary Encoder, but it didn’t fit the case design. For my Pi1541 + Tapuino I did use a two way toggle switch for the Up/Down controls to eliminate a button. I might have used a Rotary Encoder on it had that been an option at the time, as I could have used it to replace 3 of the buttons (Rotary Encoders have rotate left, rotate right, and have a button that you can press them in).

https://cbm-pi1541.firebaseapp.com/

Commodore 64 GAL PLA Replacement

Back in March, Adrian at Adrian’s Digital Basement posted a video (and a later follow up) on a cheap PLA replacement made with some still accessible GAL chips. When I saw it I picked up 5 of the required GAL chips. You need two chips to make a PLA replacement from them. The actual chips are GAL20V8B chips, preferably the -25LP but with an alternate programing of the one chip you can use the faster -15LP chips instead. The chips can be programmed by downloading the .jed files from the Daniel’s project website: https://www.freepascal.org/~daniel/c64pla/

While the GAL20V8B chips are no longer manufactured, they are in Daniel’s words still rather easy to get and cheap. The chips are able to be programmed in the popular TL866 programmers. I have the newer TL866ii Plus that I purchased for programming EPROMs. There are the older TL866 programmers and those work too, which is why he choose that model of chip rather than a similar still in production chip. This indicates that if someone wanted to they could likely recreate this project with that still in production chip at some point, but unless there is a shortage on these chips I doubt it will happen.

I happen to look around for parts and projects, so I had the chips laying around, but I have never pulled the trigger and ordered any manufactured PCBs from any of the providers out there. If you read my other posts you will see I have designed a couple of my own and etched them at home. That is a lot of work. These premade designs are not meant for that level of prototype work, the tolerances are too tight for me to make the way I made my other ones. Double sided and very small traces, a lot of through holes that need plated etc.. My one design being a single sided board, with wider gaps on the traces, also the lack of solder mask. The other design being wider gaps yet between traces to help with the lack of solder mask and it is difficult to manually align both sides and get them transferred to the board properly. I also designed my double sided board specifically to having to use vias between the sides.

So, how did I come about making progress after eight months? Well in looking for some parts for my TI99/4a, I found the PCBs for the GAL PLA for sale at what I felt was a reasonable price and ordered two. It turned out to be a very easy project. Since it is a manufactured board, all it really amounted to was installing the pin headers for everything. Adrian used Machined Sockets for the GALs and Machined Pin headers for the legs, in my case I used Machined Pin Headers for everything. Beyond that I did the same thing as he had.

Keep in mind what order you want to solder the headers up. If you use sockets, then you must put the underside legs on first. If you use all Pin Headers like I did, I found it easiest to put on the inner pin headers for the GALs on the top side, then the lower “legs” and finished up with the outer pin headers for the GALs. It turned out great. I certainly do NOT recommend using the larger square pin headers.. They damage sockets badly as they are far to large. The “turned pin”/”machine/machined pin” headers can be problematic as well especially in old single wipe sockets.

The Left are “square” pin headers like on Arduino boards, rpi etc. The right though is the machined pin headers

Above you can see the more popular Square Pin Headers. Those are not what should be used for the legs, you should use the Machined Pins instead. These are for the legs, the shorter pins going into the circuit board and soldered on there. The longer legs sticking down and those insert in the IC Sockets on the mainboard.

Machine Pin Sockets. I used these instead of IC Sockets for the GAL chips to insert into.

Above are the Machine Pin Sockets that I used in place of standard IC Sockets. I generally prefer Machine Pin Sockets to the single or double wipe standard ones. They machine type sockets are more flexible than standard sockets as they don’t go all the way into the PCB. These Machine Pin Headers though are almost identical to the Machine Pin IC Sockets, but can be made any required length. I didn’t have sockets of the size of these GAL chips, so I used the header strips. I also used these for the ATMEGA in my Pi1541, as I needed a Machine Pin Socket and the standard double wipe socket wouldn’t work on my homemade pcb.

The bottom with the round Machine Pin Headers.

A side note on the Machine Pin Headers and Machine Pin Sockets that I used. They fit together just like they are made for each other. With a little careful work and some heat shrink on the ends you can make them into inline connections for wiring. Sometimes it is useful like when I make lighting in models and such. They fit fairly well, it is a little like using those DUPONT Male and Female cables that use the “square” pins as seen further above. I think these kind of click in a little more securely, they also make a shorter connector in the end.

Unfortunately, before programing the GALs, I tried to program an EPROM for a cartridge I had assembled. A bit over half way through programming the EPROM the programmer failed, now it says it has a short of VCC problem, and fails the self test on the first VCC test.. That meant I had to order a new TL866ii Plus and wait on it to come in. I wasn’t very happy as I did not get very much use out of the first one even though I had it about a year and a half. I tried downloading the new software and putting on the lasted firmware, but the same result.. I have mostly given up on that programmer, but I certainly will keep it around incase I have an issue with the new one, or find a solution on repairing it.

The new programmer arrived the other day, and since I had the latest software and all the files already, it didn’t take long to program up the GALs. Still out of the 5 GALs two of them didn’t want to program successfully. I did program a pair then used the fifth one to start another set. I cleaned the legs on one of the two that didn’t want to program initially and it seemed to program properly that time around. I should have the chips ready to go for a second replacement if I need it in the future. I haven’t tried cleaning the other GAL and programming it again, I can’t do anything with just 1 GAL anyway at this point, but I did keep it incase I can get it working. I generally don’t do much during the week as far as pulling the Commodore apart, so I didn’t test it until today.

GAL PLA Replacement installed. I like Sharpie Oil Paint Markers for marking stuff like this. Standard Sharpie fades or rubs off over time..

Here it is installed in one of my Commodore 64s. If you read my other posts on this one where I cleaned up and repaired it, you may know I have a PLAnkton PLA in it normally. I am just testing this one. The board has a Machined Socket installed in it already which properly accepts the Machined Pin Headers, which the PLAnkton also uses. The PLA in my other Commodore 64 is also socketed, but it has a the factory single wipe socket. Installing the round Machined Pin Headers into that may damage the socket which wasn’t a great quality when it was new 40 years ago. That Commodore 64 still has an original working PLA in it, if I damage the socket it may not make good contact. If the PLA had failed, I wouldn’t be so concerned. In this case I am just testing that it works, so I know it is ready to go if needed. I doubt the PLAnkton will fail me anytime soon, but the factory PLA in the other 64 may.

I also have an ARM SID in this Commodore, I found the SID that was in this one had an issue. It may have been damaged all along, or it may have been when some traces were shorted by the unclipped legs on it.. The original SID was directly installed into the mainboard, and the legs were never trimmed at the factory. That left them sticking through quite a bit, and having the board sitting on the workbench ended up folding them over onto traces, and cut into the solder mask causing a short. The ARM SID wasn’t cheap, but I felt it was the best replacement if I wanted full function SID in the board. I am quite happy with it, but I do need to get the firmware update installed which fixes some issues.

I did not do exhaustive testing of the GAL PLA, but I did some testing here with the Dead Test and a short run of the Diag cartridge without the harness before that. I let the Dead Test run 8 full cycles without any issues.

I certainly like how easy this worked out. It looks well modified, but fits very well and is nicely built. I don’t have Super Zaxxon or anything to really stress this out that I know of. I have reinstalled my PLAnkton and closed the ole girl up. The GAL PLA is going to be packed safely in some EDS conductive foam for a rainy day.

Below is a link for Adrian’s first video on the GAL PLA. He did a follow up testing the Epyx Fast load as well.

Thanks for reading. I hope something here is helpful. I do most of these posts to kind of pull together different sources and include little things that I came across in doing these projects that may not have been mentioned in the other source material that I had when doing the project. This was a quick and easy project even though I kind of started thinking about it quite a while ago.

Due to getting these boards in and a couple others, I have finally made a customized board that I have ordered from JLCPCB. I know I see a lot more advertisements for PCBWay. Maybe I would try them at some point. JLC seemed more strait forward to get the initial order in and seemed a bit cheaper. Provided those boards work out, I expect that I will be putting together a post on that project after they come in. I have worked on a few things that I haven’t gotten around to posting. Mostly simple stuff that doesn’t take much in the way of a post and may not be of much value to post.

In the end I really don’t like waiting on things, or I may have ordered some of these PCBs myself. Then of course I waited 9 months to find some that someone else went to the trouble to order…. It really showed me how easy it is to put together a simple project like this when you have a properly made PCB compared to trying to wire up some sort of adapter from Pin Headers or IC Sockets and jumper wires. My Pi1541 was very time consuming to build. Sure there was a good bit of planning and construction, but had I ordered in the PCB, it would have been so much easier to assemble. It would have also been a bit easier to design the board when I didn’t have to take into account making VIAs by hand, dealing with lack of solder mask etc.. The Pi1541/Tapuino board did initially have a number of mistakes, and that would have kind of sucked to have ordered in and have to fix that, but it would have been possible. The board is kind of specific too, I didn’t want 4-9 laying around forever.. How many people want a Pi1541 the size of an original 1541 anyways? If I were to redo the project and it turned out possible to design it to be the proper layout and size to be a Pi Hat type but fit the original 1541, as well as maybe a break away Tapuino board section that could be stand alone.. I may have seen a use for more of the boards. That was far beyond ever doing with my hobby level board etching ability. I have since made a second Pi1541 (without a Tapuino), that one was a lot of hand wiring and would have been so much easier with a PCB or two.. Still what am I going to do with 5-10 of a board.. Most of my projects I only need one or two of.

Another Bartop Arcade Build

I ended up building a second Bartop Arcade. This one has a few changes compared to the first one. The cabinet is made with grey Melamine coated fiberboard. So it didn’t need painted. I also didn’t do the cutout of cabinet or the main assembly. The marquee top issue was addressed and cut to the proper angle. The Door was properly sized. The cabinet is assembled with Kreg Jig pocket screws and not corner blocks. For the fan the larger “speaker” pattern was used at the bottom of the door. Edge banding was used on the back of the cabinet instead of tmolding. The edge banding was used on the edges of the door and door opening, any of the board ends that weren’t covered with the Melamine coating. It makes it look very clean.

With the VESA Mounted monitor, it is cleaner inside, with more room. This build is a better fit to put a small or “Tiny” form factor PC into with Ubuntu and Retropie instead of using a RaspberryPi with Retropie. The Pi5 is out now, which is even more powerful than the Pi4 used on this build, but as of 3/24 does not yet have a Retropie Image made for it, although it is possible to do a manual install just as I did with the Pi4 on this build back when it didn’t have a build for it then. I don’t know the limits of a Pi4 or Pi5 and what it can’t run, but I expect there are plenty of (even used) Tiny PCs out there that are even more capable. It is good though that the Pi4 prices have come back to proper retail pricing, and the Pi5 is having some availability, so they are again at least viable options. With the Pi4 I don’t think ventilation is an issue, with the Pi5 it does run warmer, and with a PC you may want more ventilation that was put into this build.

here you can see the lock placement was a bit different. I also used one of the spare 30mm buttons for the safe shutdown.

The monitor was specifically purchased for this build. It is a BenQ 24″ IPS Monitor Model GW2480. This monitor supports using a VESA mount, so this time it is mounted with the VESA bracket. That made it much cleaner inside as far as not having that big block in the way in the middle. This monitor is larger than the one in the first cabinet. It “just” fits. Some 24″ monitors do not fit, this one does as it is virtually bezel-less on the sides and top. I used the same techniques for the plexiglass bezel on the monitor, there was very little black on the sides with this one though.

The back of the control panel has a piece on it that the plexiglass panel slots into so it doesn’t flex there. You can just see part of that piece of white textured material there below the edge of the monitor, that is the part the plexiglass rests into the front and it is attached to the control panel board. For the Marquee I went with the same design, as well as the same speakers. I am also using the same type of Meanwell dual voltage power supply.

There are other changes on the inside, for this build I am using a Raspberry Pi 4 not a 3b+. The basic reason being the Pi 4 is the same price and a more powerful system. When I build the unit there wasn’t yet a supported build for the Pi 4 though. In the end the only features that wouldn’t work for me was the splash screen/video.

I did run the speakers to the audio jack on the Pi this time tough. I couldn’t do the audio from the HDMI off of the BenQ and get rid of the static. Running the audio off of the Monitor audio out in the last one had eliminated the noise from the shared power supply on the speaker and Pi, but this time with this monitor that didn’t work. I had to put in a ground loop isolator. It is hard to make out much of it, but it is sitting there at in the middle of the bundle of wires. It is a little unit that has what looks like “Design” on it it is actually “Besign” Ground Loop Noise Isolator available on Amazon for $8.99 usually. I have had a number of these little units, one being in my pi1541. Since putting the audio through the monitor out didn’t fix the ground loop, I went strait to the Pi’s audio jack, as the output on the monitor I believe was much lower making the speakers not as loud.

The Pi4 does get hotter, and requires a good cooling solution. I went with a heatsink style case with the 2 fans on it. This case keeps the Pi nice and cool, the fans are not very loud. It is attached to the bottom of the cabinet with a piece of Velcro.

Here is another view, just showing the fan mounted in. You can also see the the stopper block in the middle of where the door closes. I think the unit looks much nicer inside. You can’t see it any pictures I though to take, but the controllers use a different pc board design. The ones I have used before are the kind with a header for a permanently attached usb cable, these have the usb port and use standard usb a to b cables. The order of the buttons is different. So it took me a little while to sort out how to wire these in the exact same order as my other cabinet and usb arcade stick. Having everything wired the same lets me use the usb arcade stick I had made as a third player controller. I also can have a single Retropie image for any cabinet without having to reconfigure the controllers. My Pi3 SD card is not compatible with the Pi4 though. I used alternate instructions on getting a new image setup with the Pi4. I built this back in the end of January, but they have since released a new version with “beta” support for the Pi4. The game scraper has not been working on either machine, but the update is listed to have fixed that. To set this one up, I pulled game info and images from another Retropi install that I have.

Here you can see how narrow that black border is along the sides. The larger screen goes nearly to the edge on this build. It looks like I still had the tape on the speaker grills as the silicone that I used to install them was still curing.

Here you can see a better view of the Marquee as the unit is powered off. The marquee is the same design and print that I made for the first one printed on the inkjet printer. This marquee is slightly shorter, that is why the Atari is about to get cut off in the trim pieces. I don’t think I went over the trim pieces in the last build, they are cut from some angled vinyl window trim pieces I had left over. They go back in between the mdf and aluminum pan to hold them in, and have a short lip that goes up and holds in the plexiglass.

The other changes with it are that I ended up doing a 3d printed customized volume knob for the speakers. The knob is nothing special, it was off of Thingiverse, so that I had the proper shape and fitting, I then stretched it in Tinkercad to be tall enough. It did have a bit that stuck out to denote the position, I removed most of that, as there really is not a relative marking on the cabinet. I did it in white so it would not stand out too much. I also ended up designing and printing out some speaker covers on my 3d printer. They are nothing fancy, and are held in with some clear silicone in the openings. With the first one, I used the 2″ desk insert pieces with the fabric, I didn’t have any fabric or 2″ desk grommet inserts.

The Melamine coating saved a lot of work. The draw back being it is a very thin coating and brittle. It is easy to chip out and not much can be done to fix it that I know of anyways. The fiber board in it is coarse board not the fine cardboard like consistency of the MDF, which I think may make it lighter. Tmolding is just very slightly undersized, this exposed some of the unfinished edge of the board. With the MDF black cabinet, I had paint on the edge although not as smooth or well coated as the other surfaces, so that is more hidden on it. If I had realized that, I would have painted the edge of the board before putting on the tmolding or had the edge banding put on it all around. Overall the cabinet is lighter, I think partly the coarser board rather than the MDF, also the lack of a base on the monitor as well as no blocking to build that up. The monitor itself may be lighter even though it is a bit larger. Then there is the lack of corner blocking that may very well add up overall.

It is certainly different looking than the first cabinet. I like it though.

If you are interested and have not seen the other Bartop cabinet posts I do go into more detail there. This just notes many of the differences on this build. I have also posted about an AirMouse project that I made to use as a Zapper replacement for Duck Hunt.

AirMouse for Light Gun Games

Since I built my Bartop Arcade I have been wanting to play Duck Hunt. I knew the original Light Gun could not work with the Raspberry Pi on RetroPie. I came across that you can use a mouse, which wouldn’t be fun. I knew the Arduino Pro Micro (and a couple others) could emulate a Keyboard or Mouse. There was a video about a year ago on the Element 14 Presents Youtube channel “NES Zapper on Retropie”. It is the same concept with a little different components. The Pro Micro board is a cheaper option than the board he used there. The boards based on the ATmega32U4 have a built in USB inferface on the ATmega chip, where the more common Uno Nano, and Pro Mini do not support being used in this way.

Using that as the base idea, I found various info on making an Air Mouse using a GY-521/MPU-6050 module attached to the Pro Micro board. I had found a video by “Asiq’s Theory” on Youtube with a basic Air Mouse using this setup along with code. I can no longer find the channel or video though.

I didn’t have either the GY-521/MPU-6050 board or a Arduino Pro Micro so I had to get them in to try this project out. I have plenty of Arduino Unos and Nanos around, but I hadn’t purchased any Pro Micros. The boards are similar but I usually use an Uno for prototyping then I can directly do that to a Nano without any code changes, where the Pro Micro with the different ATmega model is slightly different as well as slightly more expensive. I had been wanting some Pro Micro boards for keyboard projects in the past, but I never got into anything with them to this point. Once they arrived I took built it up on breadboard to test operation.

I ended up looking for a decent 3d model to use as a light gun model. I couldn’t find something that I could print reasonably for the amount of effort I wanted to put into this project. Really I don’t see this getting much use, mostly a little Duck Hunt now and then, maybe a few other games. I did end up finding a model I could use as a basis that had a completely different intended use. https://www.thingiverse.com/thing:2560577 I can print it as two parts now without any support. It feels pretty good to hold it. I went with some larger buttons I had in my stock with snap on caps.

I did my initial modifications included taking away parts from the model and slicing it in half. I also cut an opening for the usb cable as well as areas inside to hold the electronics. I did that work in Tinkercad. It works well enough for modifying things, but I wouldn’t want to use it to start it from scratch. I am looking for another program to use that I can get familiar enough to make items from scratch though. I first made a blue printed prototype half. It worked out fairly well, the Pro Micro fit in the area I made for it and the MPU-6050 board fit great as well. I put the 6050 forward in the gun expecting better readings of the motion, I am not sure if it makes much difference. That first half is basically a hollow shell, then I had to go on and make the other half “hold” everything. For the buttons I used my calipers to measure the button dimensions and built a placeholder model that I was able to use to give proper placement in the frame. It let me build up the back support and I slightly over sized the button “caps” so I could use them as holes to cut into the model. I also made a mock up of the MPU-6050 board so I could model in posts to hold it in place. I cut in recessed screw holes in the right side as well. For the prototype gun also had 3 opens in the shell that were from the original model that I then had to fill in so I printed them as extensions off of the right side on it. The screws I used used pointed screws to go into the left side making holes as I put them into it. That is holding that prototype together quite well so far.

After the first one was done, I decided I would make another for the second cabinet in Grey. For this one I reworked the left side with the holes filled in. I also setup to put in threaded inserts into it. This was the first time I used them, and I wasn’t sure what size to make the holes. They went in, but not very well and I ended up giving them more support and hold with glue so they hopefully won’t pull out. Otherwise the inserts are working well. They are very cheap and not well made though. For the Grey gun I tried to line up the usb cable coming in more to the center of the Pro Micro holding area. I placed it in a cut out to hold a zip tie as strain relief in the cord area. I didn’t make enough room for the “head” of the zip tie though, so when I put it in I cut out a smaller square deeper into the shell so that it would slot the whole way in and I could properly close the gun. It worked out very well. Still I have had to make my own custom cables to get enough room in there for the cable and the Pro Micro. I may rework the model a bit more to see if I can get a standard MicroUSB cable inside instead of soldering up a minimal one of my own.

The 3d model is posted here: https://www.thingiverse.com/thing:5805908

The electronics are mostly the Pro Micro and MPU-6050 board. The additions are the two momentary switches with pull down resistors wired to ground to prevent false presses. As an added feature I put in a green 5mm LED with an appropriate current limiting resistor in the tip that glows when pressing the trigger button. The device is a motion sensitive mouse using the accelerometer feature of the GY-521/MPU-6050 module. The Trigger button is the Left Mouse Button, and the other button is the Right Mouse button. So it isn’t like the old Light Gun where the game knows where the gun is pointing. It is based on relative motion, and so Duck Hunt has an onscreen cursor. It isn’t the same, but it still is fun. Those Ducks are tricky even with a cursor. To keep the gun relatively lined up, between rounds I point the gun to the upper right corner of the screen making sure the cursor has moved the whole way up there.

For the base code and initial wiring I used info from a post and YouTube video by “Asiq’s Theory”. I think it was under 50 lines of code. It was about the bare minimum to get one of these MPU-6050s working and that was very helpful to me to break down what changes I needed to make.

The exact code for one gun doesn’t work for the second one as each MPU-6050s seems to be slightly different. That is as far as “standing still” status. If that is wrong, then the cursor will constantly drift in either up or down or left or right. So a couple numbers are changed to kill the drift. I think for the one I had to do a +3 to the X (which is left/right) and the other a +4.

Asiq’s code had been based on a different orientation of the MPU module, so I adjusted for that and the code also included pauses and would freeze the cursor when pressing the button. It made it pretty good for an air mouse on a computer, but bad for tracking Ducks on the screen. This means these are not very good mice,

The Air Mouse can be pulled into about any pc and used, but as it is constantly tracking the action of pressing the buttons tends to move the cursor even if you are quite careful. You could make a more proper Air Mouse with the same components and a bit different code on the Arduino board. I have thought of making it so that if one of the buttons was held down as it was plugged in that it would make it behave differently. Still this is was to be a relatively quick project that I don’t expect to use all that heavily.

Parts:

This project uses an Arduino Pro Micro and a GY-521 Accelerometer Gyroscope Module. There are two buttons with 10k pull down resistors on them. There is also a single LED with a 1k Resistor to limit the current on it. Those bits of information are on the Code below as to what they are and what pins they goto. The resistor values aren’t critical. You could use 20k for the pull downs and possibly higher, you could use something a bit lower than 10k for the pull downs but the lower you go the more current that is drawn when pressing the button. For the LED resistor you can go as low as 330 Ohm for the LED and be safe with most any standard LED, I also commonly use 1.5k and sometimes higher for LED current limiting resistors.

Arduino Pro Micro clone: https://www.amazon.com/HiLetgo-Atmega32U4-Bootloadered-Development-Microcontroller/dp/B01MTU9GOB/

GY-521 Module: https://www.amazon.com/HiLetgo-MPU-6050-Accelerometer-Gyroscope-Converter/dp/B00LP25V1A/

12x12x7.3mm Tactile Push Button: https://www.amazon.com/TWTADE-Momentary-Tactile-Button-12x12x12mm/dp/B07CG7VTGD/

Resistors, 1k an 10k: https://www.amazon.com/EDGELEC-Resistor-Tolerance-Multiple-Resistance/dp/B07QJB31M7/

LED, hopefully you have a couple, or you can pick them up on Amazon etc too.

All links are current as of posting this in 2/4/23 none are affiliate links or anything.

Below is the code for the Arduino Pro Micro Line 78 and 79 are where the drift adjustment numbers are Currently the “+7” on Line 78 and the “-1” on 79. I am referencing “Test code” that Ashiq mentioned, again I can’t find that source to get it. I guess Ashiq’s video was deleted or channel on Youtube or something, if someone finds it let me know so I can reference it properly. That code would return numbers from what are the gx and gz values. Using those values as a starting point, I guess I had something around a constant 400 and 100 respectively out of the 6050. The point of Line 78 and 79 are to get us a result of 0 if the Air Mouse is not in motion. Changing the divide by 150 will change the speed at which the mouse will track and possibly throw off your adjustment values at the end by a number or two. You may notice the Serial.print at the end there, that returns the vx and vy values that are created on line 78 and 79. So if you open the Arduino (or other) serial monitor set to the 9600 and to the serial port the Pro Micro is showing up as and look at those numbers you can use that as your Test code to get an idea of what changes you need to make be it to the 400 or 100 or the minor fine tuning numbers at the end of those lines. The goal is to get the values to 0 when the gun is not being moved. If you make the “dead zone” area to large you can’t make fine movements, which makes aiming even more difficult. These motions are relative, and not to “scale”, so at the start of a round you point at the screen, but to get the cursor to move the way you want you likely end up pointing in a very different place by the end. You can quickly move to aim at the center again and over correct to get the cursor more inline with where you are pointing. It is hard to describe. I enjoy it for a few rounds of Duck Hunt, for longer round games it may not be as much fun. I don’t know for sure as I haven’t played other games with it to this point.

Air Mouse Code for the Pro Micro is below:

#include <Wire.h>
#include <I2Cdev.h>
#include <MPU6050.h>
#include <Mouse.h>

/*
  Name: Prototype Arduino Pro Micro AirMouse
  Author: Markeno
  Created: 3/7/2020
  Version .99
  Based on a sketch posted by "Asiq's Theory".

  Feature:  USB 2 Button AirMouse for RetroPie Gun

                     +-----+
        +------------| USB |-------------+
        |            +-----+             |
    D1  | [ ]1/TX/INT3            RAW[X] |    
    D0  | [ ]0/RX/INT2            GND[X] |    
        | [ ]GND                  RST[ ] |    
        | [ ]GND                  VCC[X] |    
    SDA | [X]2/INT1    ___         21[ ] |   A3
    SCL |~[X]3/INT0   /   \        20[ ] |   A2
    A6  | [X]4       / PRO \       19[ ] |   A1
        |~[X]5       \Micro/       18[ ] |   A0
    A7  |~[x]6        \___/   SCLK/15[ ] |     
        | [ ]7/INT6           MISO/14[ ] |     
    A8  | [ ]8                MOSI/16[ ]~|     
    A9  |~[ ]9                  SS/10[ ]~|   A10
        |                                |    
        +--------------------------------+  
		
		Based on: http://busyducks.com/ascii-art-arduinos
		
		Gy521 2=SDA 3=SCL, RAW to GY521 VCC in (didn't want to work from the PRO Micro VCC pin), GND to GND 
		4 Left Mouse Button, 10k Pulldown to Ground, press to VCC
		5 Right Mouse Button, 10k Pulldown to Ground, press to VCC
		6 Additional Wire, for an LED in the tip when the main trigger was pressed.
*/

MPU6050 mpu;
int16_t ax, ay, az, gx, gy, gz;//Variable for the Accel and Gyro Data
int vx, vy;// X and Y axis variables

const int button1 = 4; //For Left Mouse Button
const int button2 = 5; //For Right Mouse Button
const int fled1 = 6; // For Tip LED

int responseDelay = 10;
// Generally, you should use "unsigned long" for variables that hold time
// The value will quickly become too large for an int to store
unsigned long previousMillis = 0;        // will store last time Button was updated
// constants won't change:
const long interval = 100;           // interval at which to change button states (milliseconds)
// Variables will change:
int btn1State = LOW;             // ledState used to set the LED
int btn2State = LOW;             // ledState used to set the LED

void setup() {
  pinMode(button1, INPUT);
  pinMode(button2, INPUT);
  pinMode(fled1, OUTPUT);
  
  Serial.begin(9600);
  Wire.begin();
  Mouse.begin();
  mpu.initialize();
  if (!mpu.testConnection()) {
    while (1);
  }
}

void loop() {
	
  //Get the Acel and Gyro Data as 6 values
  mpu.getMotion6(&ax, &ay, &az, &gx, &gy, &gz);

  vx = ((gx - 400) / 150)+7 ; // (Grey Gun +3) (Blue Gun +4)"-400" because the X-Axis Left/Right of gyroscope give values about -400 while it's not moving. Change this value if you get something different using the TEST code, checking if there are values far from zero. "+3" added due to value around -3..
  vy = (-(gz - 100) / 150)-1 ; // (Grey Gun 0) (Blue Gun -1)"- 100" same here for the Y-Axis Up/Down. may need to "-1" "+1" etc due to some drift still.

  vx = vx / 2;
  vy = vy /2;
  
  //Create a small -1,0,1 DeadZone for the Y-Axis
  if (vy>=-1 && vy <= 1)
  {
    vy=0;
  }
 
  //Create a small -1,0,1 DeadZone for the X-Axis
  if (vx>=-1 && vx <= 1)
  {
    vx=0;
  }
  
  int buttonState1 = digitalRead(button1);
  int buttonState2 = digitalRead(button2);

  // check to see if it's time to blink the LED; that is, if the difference
  // between the current time and last time you blinked the LED is bigger than
  // the interval at which you want to blink the LED.
  unsigned long currentMillis = millis();

if (currentMillis - previousMillis >= interval) {
    // save the last time you blinked the LED
    previousMillis = currentMillis;

    if (buttonState1 != btn1State) {
      if (buttonState1 == HIGH) {
        Mouse.press(MOUSE_LEFT); //5
        digitalWrite(fled1,1);// Light LED
      }
      else
      {
        Mouse.release(MOUSE_LEFT);  
        digitalWrite(fled1,0);// Light LED Off
      }
    btn1State = buttonState1;
    }

    if (buttonState2 != btn2State) {
      if (buttonState2 == HIGH) {
        Mouse.press(MOUSE_RIGHT);
      }
      else
      {
        Mouse.release(MOUSE_RIGHT);  
      }
    btn2State = buttonState2;
    }

/*
    if (buttonState2 == HIGH) {
      btn2State = HIGH;
      Mouse.press(MOUSE_RIGHT);
      delay(100);
      btn2State = LOW;
      Mouse.release(MOUSE_RIGHT);
    }
*/

  }

  Serial.print(-vx);
  Serial.print(-vy);
  Serial.print("\n");
  if (millis() - previousMillis >= 20) {
  Mouse.move(-vx, -vy);
  }
 
}

Note: the “X” in the Pro Micro Ascii diagram denotes something is wired to that pin, below the diagram is described what is going to those various pins.

There are far better options out there. You could use a regular mouse, but that certainly wasn’t remotely the same. You could buy an Air Mouse, the feel is not the same either. There are units you can get similar to the Wii motion bar that will work with Retropi and use a Wiimote in a gun holder. Those should track much better because it is real tracking and not relative acceleration. With the Air Mouse solution it just knows it went right or left or up or down, you don’t even have to be facing the right direction. Still it is not a very expensive solution.

Addendum: MPU-6050 Test code. This is some Test code I had on file for the MPU-6050 this is NOT the Air Mouse code, that code is the Section Above here. I believe this is the code I used to get my baseline “stationary” values to put in the Air Mouse Code The 400 and 100 in the Air Mouse Code above. To use the code, again open the Arduino or other Serial Monitor set to 9600 and the Serial Port that the Pro Micro is connected to. I could be wrong and that may not be what this is. It has been sitting there a long time.

// MPU-6050 Test Code.  This is NOT the Air Mouse Code!
// (c) Michael Schoeffler 2017, http://www.mschoeffler.de

#include "Wire.h" // This library allows you to communicate with I2C devices.

const int MPU_ADDR = 0x68; // I2C address of the MPU-6050. If AD0 pin is set to HIGH, the I2C address will be 0x69.

int16_t accelerometer_x, accelerometer_y, accelerometer_z; // variables for accelerometer raw data
int16_t gyro_x, gyro_y, gyro_z; // variables for gyro raw data
int16_t temperature; // variables for temperature data

char tmp_str[7]; // temporary variable used in convert function

char* convert_int16_to_str(int16_t i) { // converts int16 to string. Moreover, resulting strings will have the same length in the debug monitor.
  sprintf(tmp_str, "%6d", i);
  return tmp_str;
}

void setup() {
  Serial.begin(9600);
  Wire.begin();
  Wire.beginTransmission(MPU_ADDR); // Begins a transmission to the I2C slave (GY-521 board)
  Wire.write(0x6B); // PWR_MGMT_1 register
  Wire.write(0); // set to zero (wakes up the MPU-6050)
  Wire.endTransmission(true);
}
void loop() {
  Wire.beginTransmission(MPU_ADDR);
  Wire.write(0x3B); // starting with register 0x3B (ACCEL_XOUT_H) [MPU-6000 and MPU-6050 Register Map and Descriptions Revision 4.2, p.40]
  Wire.endTransmission(false); // the parameter indicates that the Arduino will send a restart. As a result, the connection is kept active.
  Wire.requestFrom(MPU_ADDR, 7*2, true); // request a total of 7*2=14 registers
  
  // "Wire.read()<<8 | Wire.read();" means two registers are read and stored in the same variable
  accelerometer_x = Wire.read()<<8 | Wire.read(); // reading registers: 0x3B (ACCEL_XOUT_H) and 0x3C (ACCEL_XOUT_L)
  accelerometer_y = Wire.read()<<8 | Wire.read(); // reading registers: 0x3D (ACCEL_YOUT_H) and 0x3E (ACCEL_YOUT_L)
  accelerometer_z = Wire.read()<<8 | Wire.read(); // reading registers: 0x3F (ACCEL_ZOUT_H) and 0x40 (ACCEL_ZOUT_L)
  temperature = Wire.read()<<8 | Wire.read(); // reading registers: 0x41 (TEMP_OUT_H) and 0x42 (TEMP_OUT_L)
  gyro_x = Wire.read()<<8 | Wire.read(); // reading registers: 0x43 (GYRO_XOUT_H) and 0x44 (GYRO_XOUT_L)
  gyro_y = Wire.read()<<8 | Wire.read(); // reading registers: 0x45 (GYRO_YOUT_H) and 0x46 (GYRO_YOUT_L)
  gyro_z = Wire.read()<<8 | Wire.read(); // reading registers: 0x47 (GYRO_ZOUT_H) and 0x48 (GYRO_ZOUT_L)
  
  // print out data
  // Serial.print("aX = "); Serial.print(convert_int16_to_str(accelerometer_x));
  // Serial.print(" | aY = "); Serial.print(convert_int16_to_str(accelerometer_y));
  // Serial.print(" | aZ = "); Serial.print(convert_int16_to_str(accelerometer_z));
  // the following equation was taken from the documentation [MPU-6000/MPU-6050 Register Map and Description, p.30]
  // Serial.print(" | tmp = "); Serial.print(temperature/340.00+36.53);
  Serial.print(" | gX = "); Serial.print(convert_int16_to_str(gyro_x));
  Serial.print(" | gY = "); Serial.print(convert_int16_to_str(gyro_y));
  Serial.print(" | gZ = "); Serial.print(convert_int16_to_str(gyro_z));
  Serial.println();
  
  // delay
  delay(1000);
}