nesdev.org

I am the proud new owner of the yellowest of yellowed Hyundai Comboy consoles:)
It has a sound issue which I suspect will be easy to fix once I know what I'm looking for.
Here's a video:
http://www.youtube.com/watch?v=GZspODlCK88

Anyone know where I should start?

Cheers,
James

If it has the same PCB like a regular NES check the audio traces and resistors going from CPU Pin 1 and Pin 2.

phreak97 wrote:yellowest of yellowed Hyundai Comboy

Hydrogen peroxide and sunlight are great for de-yellowing plastic. I made an SNES and its controllers look like new this way.

Um. The triangle is out of tune with the pulse waves, and the values written to the noise channel seem to be broken in some other novel way... as though it can't ever turn it off?

It vaguely sounds like one of the bits in the triangle period is forced set? Comparing your recording to the NSF, sometimes it's flat and sometimes it's right. There's one point in the theme where the triangle is trying to play a G at 392Hz, but it instead plays something at ≈360Hz. In binary, those would be periods of b'10001111' and ... maybe b'10011111' ?

Do you have a flashcart or the ability to make one? Could you try playing around with livenes ?

lidnariq wrote:Um. The triangle is out of tune with the pulse waves, and the values written to the noise channel seem to be broken in some other novel way... as though it can't ever turn it off?

It vaguely sounds like one of the bits in the triangle period is forced set? Comparing your recording to the NSF, sometimes it's flat and sometimes it's right. There's one point in the theme where the triangle is trying to play a G at 392Hz, but it instead plays something at ≈360Hz. In binary, those would be periods of b'10001111' and ... maybe b'10011111' ?

Do you have a flashcart or the ability to make one? Could you try playing around with livenes ?

Yes! aside from not knowing it was called the triangle period did think it sounded like at least one stuck address line or something similar, your binary values would make sense, do those bits relate to a set of lines on the pcb which I can probe?

I have an eprom burner and quite a lot of carts, if there's a cart you know of which I can easily attach the livenes rom to I'll do it temporarily.

Thanks,
James.

LiveNES is a just a tiny minimalist thing that you could stuff into any cart with CHR RAM and a 4 KiB (or larger) PRG ROM. It doesn't use any banking, so it should sit ok on almost any hardware.

The ROM will need repacking to fit into a smaller ROM (it comes by default as a full 16 KiB image, with a 12 KiB hole in the middle), or duplication to fit on a larger one.


But fixing this is mostly out of the reach of a hobbyist, sadly. If my guess is correct, something bad happened to the 2A03's die, and I don't know how you'd fix it there. The first workaround that comes to mind is using an FPGA to emulate the entire APU, but ... it's probably easier to find another 2A03.

lidnariq wrote:LiveNES is a just a tiny minimalist thing that you could stuff into any cart with CHR RAM and a 4 KiB (or larger) PRG ROM. It doesn't use any banking, so it should sit ok on almost any hardware.

The ROM will need repacking to fit into a smaller ROM (it comes by default as a full 16 KiB image, with a 12 KiB hole in the middle), or duplication to fit on a larger one.


But fixing this is mostly out of the reach of a hobbyist, sadly. If my guess is correct, something bad happened to the 2A03's die, and I don't know how you'd fix it there. The first workaround that comes to mind is using an FPGA to emulate the entire APU, but ... it's probably easier to find another 2A03.

Good thing I have several scrap famicom consoles
I'll swap out the 2A03 and post back.
Thanks!

Well diagnosed! I removed the cpu and installed it into a famicom which was just showing a black screen (guessed cpu). The comboy cpu brought the famicom back to life and took the sound faults with it, so now I just need a cpu for the comboy.
Turns out my scrap famicoms all have bad cpus so that sucks.