Is this going to come to market within the community sometime down the road? If a run of the boards is done I am definitely in for one!

Yeah, I hope a run of these boards are made. Only problem is finding a case to mount the board in. Has any progress been made on that?

It was designed around a premade case (you would just have to cut the cartridge slot.

I decided to test the RGB output of the Super 8 this weekend. I built another PCB and installed the Duck Hunt PPU and a CPU from a VS motherboard. I tested it using my PSOne LCD screen since it has RGB input. The picture was good quality but one problem, it had no sprites! I thought my PPU was bad, but it turned out to be a bad CPU. I swapped it with another and it was working great. So a bad $25 chip is much better than a bad $200 chip.
It's good to know there aren't any issues with the RGB.
While testing the RGB output on the PSOne screen I realized this combo could make a sweet portable.

I tried out the Zapper on the composite Super 8 connected to a CRT TV and it worked with no issues.

I tried the RGB Super 8 with the Jrok video converter but it had a black screen. One guy on Youtube said you need a LM1881 sync stripper on the sync line to get it to work, so I will try that.

I think the PPU clock trace may be too long on the v1.0 PCB. I will test this by cutting the clock trace and connecting these points directly with some kynar wire. Either way I'm making it shorter on v1.1.

No progress on the case yet, I did a test fit and the PCB fits nicely and the screw holes line up in the AG-85. I'm more concerned with testing and getting the PCB design finalized at the moment.

I want to build a run of 25 or so Super 8 systems eventually. I want to incorporate every improvement I can think of into the v1.1 PCB so I'm going to take my time finalizing it. I can't give a timeline, depends on how much time I have to work on it.
I may get some cases custom cut, but I might just offer the boards.
Fully assembled systems ready to use will be pricey, but I could sell the assembled PCB without CPU and PPU for cheaper. I could also do custom builds. I don't know, it's a ways off at the moment.

So did you find out what is wrong with SMB3, too?

Not yet. I'd like to get that figured out before I make any new boards.

Did you test other games that use WRAM? Games that use MMC3? Mmmmm...that should be the only 2 things that can change. It'd be really easy to find out what's wrong if you could log the system data/address bus for a few frames.

Picture? I never found one in this thread.

http://www.polycase.com/ag-85...

I've seen people use that same case for Superguns. Not my first choice from an aesthetic perspective, but at least there's a case for this Super 8.

Any projected price for the board? Or perhaps a fully populated board?

Your boards look really nice, but they are only slightly smaller than the original NES motherboard.

Could you lead me to a source for the 72-pin connectors? I'm sick of the flaky NES connectors, as both the new "death grip" connectors and the original ZIF connectors have issues. I'm kinda thinking about getting two strait 72-pin connectors (one for the board and one for the cart) and modding my Toaster into a top loader, or possibly mounting the second 72-pin connector directly into the loading bay so that the cartridge hangs out horizontally. IMO, the audio/video quality of the original AV Toasters is sufficient and the original hardware is well made and very reliable; it's just that the ZIF socket design is the Achilles heel of the system so if I can do away with the original loader design that would be great.

What issues? I bought one and after I broke it in and softened the connectors (and never pushed the games down), that same connector has been working 100% ever since. It is my single best operating NES and my other one can't compare with it's original connector. I'd like to get a couple more death grip connectors, know anyone on eBay who for sure sells them?

Supposedly this works:
http://www.digikey.com/product-de...

Tested the RGB Super 8 with the Jrok video converter board today and used a LM1881N sync separator. The s-video looked excellent, but the component video out didn't display properly. Only the top half of the screen had an image, the bottom half was garbage. It seems to be an issue with the video converter.

I made a slightly less crappy Youtube video showing the RGB video converted to s-video.




Super Mario Brothers 3 DOES work with the Super 8. My cart had a dirty connector, and that was the only problem. If only I checked that earlier...

I'm going to try to get one game that uses each major mapper for testing. I've tested games using UNROM, MMC1, MMC3, MMC6, and AOROM and found no issues.

I'm still having some issues with the composite video output version. One of the CRT TVs I tried it on was in black and white. The LCD TVs work well, just not CRT. Since a CRT TV is the only way to use the Zapper or ROB I want to get this working.

I tried a shorter PPU clock line and it had no effect.
I used a 2N3906 transistor for the video amp and 2N3904 transistors in the clock circuit.

Good stuff!

1) resistors in video amplifier - try 100 and 150 Ohm, instead of 220 Ohm and 2.2kOhm
2) if line of PPU is too long, you can compensate it by removing(or decreasing) 15pF capacitor
3) check ground layout around pin21 to avoid jailbars
4) any general low power PNP silicon transistor will do the job
5) try to use clock after buffer transistor for PPU as well

P.S. How much for the "board only"?

I have already tried the 100 and 150 ohm resistors in the video amp circuit as well as a smaller 33uf ceramic capacitor. It does look better but there wasn't much change over the original schematic. My PPU clock line is actually shorter than the original NES front loader, might not be a problem. I think my grounding and/or video signal lines may be the issue.
On v1.0, the composite video signal crosses over the data lines leading to the PPU before it's amped by the transistor. I think replacing this PCB trace with a jumper wire would eliminate this interference since there would be a ground layer between them. I didn't want the PCB to need jumper wires, but in this case it may be worth it.

I'll do some more experimenting with the grounding since this is critical for least interference.

I could sell a v1.0 PCB for $30.

glad to see progress

Keep in mind capacitance of PCB material. About video amplifier - you should retrace that part of PCB. Feel free to PM me for more info

V1.1 is in progress and I have been working on the composite video amp. I would like to put the 2N3906 transistor right by PPU pin 21, but there isn't enough room for a through hole transistor. I was thinking of using a surface mount transistor here, then the signal crossing the 0-7 data lines and controller data lines would be amped and less prone to interference.

I'm haven't heard about capacitance from the PCB itself before, but it makes sense since a capacitor is basically 2 metal plates seperated by an insulator.

I don't have much room to work with on the PCB for the composite amp. I want to add provisions for an on board sync seperator for RGB converters, but I don't think there will be room.

I think my grounding scheme is pretty good.
PPU pin 20, the 68pf cap to pin 24, and the 30pf cap to pin 13 are connected to the ground plane.
PPU pin 17, 2N3906 ground, THS7314 ground, and its 0.1uf capacitor are connected to the video out ground. There is a jumper connecting this to the ground plane (at the video out connector.) I will try removing the jumper to see what effect it has.
I'm wondering if PPU pin 17 should go directly to video out, leaving the other grounds connected to the ground plane.

On the Famicom schematic, it also has a 330pf capacitor from PPU pin 13 to ground near the 74LS139. I added provisions for this, not sure if its actually needed.

That, and I hope to add tap points for possible wiring of a NES expansion port for the Famicom Disc System or one of those new ethernet adapters. This has proved difficult because the Famicom and NES have the 74HC368 chips wired very differently. I may label these tap points based on the NES expansion connector. I'm trying to add lots of tap points to make mods easier, like CPU overclocking (if you really want to do that.)

I'll be working on the Super 8 Design more this weekend. I want to get v1.1 manufactured ASAP, but I want it to be as good as I can make it first. I am low budget, and can't afford to keep making prototype boards. This next one has to have everything right.

Hopefully another update Saturday if I make progress.

I like the idea of a tap for overclocking the CPU. I've heard some NES CPUs can run up to 2x speed without errors. The user could add a simple switch to the circuit and a daughter board with a high speed 555 timer to set the CPU clock. I have heard it is very difficult to pull this mod off on the NES because the trace from the clock circuit passes through both the PPU and CPU pins, so you have to desolder the CPU and resolder it just to lift out one leg off the board for the OC circuit, either that or cut traces and rewire everything.

This is the most exciting NES thing I've seen for awhile!  I've always dreamed of having a 100% compatible, reliable clone.  I would love to be in on this.  Thanks for keeping us posted on this project.

Low budget, did you get my PM about the NES boards I have for you?

I have measured the PPU clock line in the NES front loader and it is 4-3/4" long. In my v1.0 PCB, the length is only 3-1/2". I have tried 10pf and 33pf capacitors but there was no change from what I could tell.

I disconnected video ground (PPU pin 17) from the ground plane but there was no change in picture quality. I decided to leave all the PPU and video ground pins connected to the ground plane when using the composite PPU.

I tried replacing the 47pf caps with 56pf caps and that had no effect.

I tried tapping the PPU clock off the CPU clock and that did have an improvement in stability.
When testing Super C, the game actually froze soon after a game was started and required a reset. After tapping the PPU clock off the CPU clock, the game could be played. I also had Super Mario 3 freeze while I was playing, but this fixed that game as well.

Basically, what's going on with the composite PPU Super 8 is some games have graphics issues. As in the first video I posted, the composite PPU Super 8 worked great with Batman and Mega Man 2. Only 10% or so of the games I tested have these glitches which sometimes make the picture roll and appear out of sync.

I'm 90% sure the issues I'm having with the composite PPU are from the clock circuit. I think I'll need to make some changes or come up with a new clock circuit. Great.


The Super 8 with the RGB PPU works great and I haven't had any issues. All games tested played without glitches (except Mike Tyson's Punch Out.) I hooked it up to my CRT TV using component video from the Jrok converter and it had excellent picture. It's easily as good as a well done NES RGB mod.

I have lots of work left to do.

Not sure about double, but CPU can run with 27MHz clock

[quote]I tried tapping the PPU clock off the CPU clock and that did have an improvement in stability.[/quote] Bad shape of signal. Try to "shape" it via HC14

I overclocked a NES one time using a 32mhz oscillator. The CPU divides the incoming clock signal by 12 so it was a little fast at 2.6mhz. The sound was much higher pitch than normal, there were graphics glitches, and some games didn't work at all. I wouldn't really recommend overclocking, but if you do I would use a slower oscillator than what I used.

To overclock the NES, all I did was cut the CPU clock trace close to the IC, solder a short piece of 30ga wire to the CPU pin and the other to the output of the oscillator. You don't want to mess with the PPU timing, as everything will get all jacked up.

Still working on making all games display correctly with the composite PPU.
I ordered an oscilloscope, it should be here next week. I think that would help a lot and decided it would be worth it.
On the Super 8 PCB, there are inverters on board, but just not near the clock circuit. I would prefer not to have an additional IC on board just for this purpose, as there is very little room to work with in that area of the PCB.

I made progress on the case while I'm waiting for that.
I cut the cartridge slot and holes for the controller ports. The cart slot has the chamfered edges so the game can only be inserted the right way.
The controller ports are quite long, so the only place I could mount them is near the cartridge slot on the top.

The RGB PPU with heatsink in a socket has plenty of clearance.

It's looking like this case will work out pretty well.