Back to white screen on opentendo build

Discuss hardware-related topics, such as development cartridges, CopyNES, PowerPak, EPROMs, or whatever.

Moderator: Moderators

vinnieb
Posts: 19
Joined: Mon Jan 18, 2021 10:58 am

Back to white screen on opentendo build

Post by vinnieb »

Hi All,

Was really hoping I could crack this one but hitting a dead end. With some help here a few years back I got my opentendo build up and running, more or less. Finally getting around to try and finish it up, but I seem to have gone backwards.

The initial problems I had seemed around getting an OK clock signal going, which with some variation in parts the board fired up at least.

Coming back to the project now, I had no luck at power on, but I saw the spec of a resistor and capacitor around that area changed, so I thought right lets get back on spec.

After alot of fiddling, I have what I think is a good clock signal, and what I think might be OK on M2 and CPUCLK. RST is also staying high.

From the clock crystal itself:
DS1Z_QuickPrint3.png
Then from M2 on the CPU:
DS1Z_QuickPrint5.png
And CPUCLK:
DS1Z_QuickPrint6.png
RST staying high:
DS1Z_QuickPrint4.png
I'm not sure, if M2 should look like that, or perhaps the problem is already there.

But assuming that's OK, I was then wondering if its something with the CIC workaround. The "How to fix a NES mainboard without 10NES lockout chip?" had alot of info on this, so I thought I'd switch from redherrings NullCIC which I had before, to the design on there (as running out of ideas really..)

Looking at the 1Mohm resistor going to 74HCU04, I get this at power on:
DS1Z_QuickPrint7.png
I thought that might repeat when I touch RST to ground, but it doesnt seem to, I don't know if that's related. The screen output does glitch to black and then back to this white after doing that though.

Here's current state of the board:
board_view.jpg
And output:
screen_view.jpg
And that's where I'm a bit stuck. I've got various spare parts if maybe I've fried something along the way.. only thing I don't have is any other game cartridge.


Any thoughts on where the problem might lie / other bits to test would be much appreciated!!
Joe
Posts: 689
Joined: Mon Apr 01, 2013 11:17 pm

Re: Back to white screen on opentendo build

Post by Joe »

I'd check A0 on the CPU. It might only toggle briefly after reset and get stuck if the CPU is crashing.
lidnariq
Site Admin
Posts: 11645
Joined: Sun Apr 13, 2008 11:12 am

Re: Back to white screen on opentendo build

Post by lidnariq »

All of your oscilloscope signals and PPU photo imply that the CPU and PPU and crystal are working correctly. As Joe says, the next thing to check is CPU A0, which will get stuck if the CPU executes a specific set of bad instructions.

If it does stop oscillating, the next place to check is connection to the cartridge - something's caused it to execute one of those bad instructions.
vinnieb
Posts: 19
Joined: Mon Jan 18, 2021 10:58 am

Re: Back to white screen on opentendo build

Post by vinnieb »

Thanks Joe and lidnariq, that's pretty encouraging that the early part of the circuitry is checking out OK.

So I've taken a look at A0, initially I guess not with the scope setup that well as seemed to be getting loads of sort of repeat but shifted / ringing pulses, unless I played around with the trigger level alot and even then still a few:
DS1Z_QuickPrint11.png
But looks a bit cleaner with a different trigger mode, assuming thats a better choice
DS1Z_QuickPrint10.png
Gave it a reset too and goes back to a similar kind of pattern. I guess that means there's some sort of code execution going on?
lidnariq
Site Admin
Posts: 11645
Joined: Sun Apr 13, 2008 11:12 am

Re: Back to white screen on opentendo build

Post by lidnariq »

That repetition implies it's getting stuck in an infinite loop - the repetition period is 198px=7.9us, roughly 13cy. That's slower than the normal 7-cycle bit $2002 / bpl self power-up wait, but they could have done something slower. If you can 'scope both A0 and /PPUCE (pin13) that would be diagnostic.

Do you have another PPU to test? That's my best guess right now.
vinnieb
Posts: 19
Joined: Mon Jan 18, 2021 10:58 am

Re: Back to white screen on opentendo build

Post by vinnieb »

Interesting! OK, so I tried to probe both of those from the PPU chip - looks like A0 would be pin12 there so easy to take side by side.

Not sure if either i've not got my setup right, or /PPUCE isnt very eventful, just seems staying high? That's with A0 probe pulse driving things, tho just measuing /PPUCE and nothing else I get the same result:
DS1Z_QuickPrint12.png
I do have a few other PPU chips, as was ordering from aliexpress thought might just get a few as expected a few duds. The one currently in there was working at some point but perfectly possible its got damaged since. But, I popped a different PPU in and get the same result. Though - I don't know if that other one is a good PPU either. I've got another couple I could try if its worth it
lidnariq
Site Admin
Posts: 11645
Joined: Sun Apr 13, 2008 11:12 am

Re: Back to white screen on opentendo build

Post by lidnariq »

Ok, that's helpful, it means the infinite loop is not waiting on the PPU.

Unfortunately, that puts us back in a short list of other options: the cartridge connection is bad, or the CPU's RAM is bad
vinnieb
Posts: 19
Joined: Mon Jan 18, 2021 10:58 am

Re: Back to white screen on opentendo build

Post by vinnieb »

Hm.. both sound feasible. I 3D printed a cartridge holder slot, and it was a bit tight initially, wonder if that's damaged any of the connections. I guess for that ideally would have an old cartridge that I could test continuity back to the board from?

For CPU RAM, i do have a spare chip I think, I just regret not using a socket for it first time around now if I need to desolder that + replace, but can fit one 2nd time around.
vinnieb
Posts: 19
Joined: Mon Jan 18, 2021 10:58 am

Re: Back to white screen on opentendo build

Post by vinnieb »

Out of interest, the TP points mentioned here https://wiki.console5.com/wiki/Nintendo_NES-001 and the waveforms, do they follow in order of fundamentals, i.e. if no TP1 then nothing else worth checking, work through them sequentially? Just that TP1 to 3 seem OK, I did try TP4 from ROMSEL but I don't really see that shape (this is with M2 taken from the same U3 chip)
DS1Z_QuickPrint13.png
I've got a DIP socket ready so I think I'll take out the CPU RAM this weekend and replace that.
Killbox91
Posts: 59
Joined: Sun May 12, 2024 9:25 am

Re: Back to white screen on opentendo build

Post by Killbox91 »

vinnieb wrote: Fri Feb 07, 2025 7:32 am Out of interest, the TP points mentioned here https://wiki.console5.com/wiki/Nintendo_NES-001 and the waveforms, do they follow in order of fundamentals, i.e. if no TP1 then nothing else worth checking, work through them sequentially? Just that TP1 to 3 seem OK, I did try TP4 from ROMSEL but I don't really see that shape (this is with M2 taken from the same U3 chip)

DS1Z_QuickPrint13.png

I've got a DIP socket ready so I think I'll take out the CPU RAM this weekend and replace that.
UA6538 can you try it ?
vinnieb
Posts: 19
Joined: Mon Jan 18, 2021 10:58 am

Re: Back to white screen on opentendo build

Post by vinnieb »

I've only got UA6527P and UA6528P, 3 of each.

It was working with the 6528P at one point, though a few years ago - I can try looking for a cheap 6538 if its worth a shot?
lidnariq
Site Admin
Posts: 11645
Joined: Sun Apr 13, 2008 11:12 am

Re: Back to white screen on opentendo build

Post by lidnariq »

vinnieb wrote: Fri Feb 07, 2025 7:32 am Out of interest, the TP points mentioned here https://wiki.console5.com/wiki/Nintendo_NES-001 and the waveforms, do they follow in order of fundamentals, i.e. if no TP1 then nothing else worth checking, work through them sequentially? Just that TP1 to 3 seem OK, I did try TP4 from ROMSEL but I don't really see that shape (this is with M2 taken from the same U3 chip)
They're not really in order, I think.

It'd be really nice if console5 also said what those signals were, not just the waveforms...

TP1: 21.5MHz/26.6MHz clock after buffer, on CPU
TP2: 4MHz CIC clock, on driver
TP3: M2 from CPU, on 74'139
TP4: /ROMSEL from 74'139 -- this should be low when M2 is high (TP3) and A15 is high (74'139 pin 13). A crash may cause A15 to stay low, so it's not indicative on its own.
TP5: Composite video from the PPU
TP6: /PPUSEL
TP7: Player 1 joypad reload strobe
TP8: Player 1 joypad clock
TP9: PPU A8-A13
TP10, TP11: CPU audio outputs
TP12: audio amplifier output
TP13: CPU R/W, on the PPU
TP14: PPU /WR, for writing to video RAM
TP15: CPU A2, on the PPU
TP16: Player 1 joypad reload strobe, on the CPU
TP17: CPU A13, on the 74'139
TP18: Internal CPU /RAMCE, on the 74'139

Not the specific signals I'd choose, but hopefully these descriptions will let you know which you don't need to check
vinnieb
Posts: 19
Joined: Mon Jan 18, 2021 10:58 am

Re: Back to white screen on opentendo build

Post by vinnieb »

Ah great, that sure gives some useful context. Looks like there's more diagnostics to do yet sadly.

Unfortunately doesn't look like it was the CPU RAM (I'm assuming this is the one marked SRAM (WRAM) right under the CPU), placed a new unused one (SRAM 16k) from when I bought spares and getting almost exactly the same response. Only slight difference is the picture displayed is now a sort of navy blue image but still the same edge pattern around it as the white/grey one before.

So that leaves, cartridge connector? Or cartridge itself? Cartridge connector was a new one and not really seen much use.

I don't suppose there's anything I can spoof, or measure, from something other than a cartridge to give some sort of input/output check there? Plenty of Arduino type devices here if that might do it
lidnariq
Site Admin
Posts: 11645
Joined: Sun Apr 13, 2008 11:12 am

Re: Back to white screen on opentendo build

Post by lidnariq »

Theoretically it should be possible to replace the RAM with a ROM that would do some simple self-tests, especially since you have it socketed. Unfortunately, I don't think anyone's already written this; even my "cartless" demo assumes that it's RAM so hold state.

That said, do check what A15 is doing as you release the reset button. It should at least go high and low a few times.
vinnieb
Posts: 19
Joined: Mon Jan 18, 2021 10:58 am

Re: Back to white screen on opentendo build

Post by vinnieb »

Hopefully I'm mimicking it right, I don't have a proper power/reset panel setup, so just touching RST to ground briefly for the reset.

Doing that, and looking at A15 (pin 19 on CPU?), it goes to zero during the reset, but then after this constantly:
DS1Z_QuickPrint15.png
Post Reply