Glitched sprites using Game Doctor 6+ with Playchoice PPU

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

Moderator: Moderators

Post Reply
famiac
Posts: 19
Joined: Mon Dec 26, 2011 5:22 pm

Glitched sprites using Game Doctor 6+ with Playchoice PPU

Post by famiac »

Hi all,

occurs on my AV famicom, which was fitted with an RP2C03B.

After loading a game through the FDS to the Turbo Game Doctor 6+, some horizontal lines appear to scroll at high speeds over the sprites. Issue occurs on the three games that i have tried with it. I have also tried with an RC2C05 installed, and the issue persisted. I noticed that adjusting the trimmer cap changes the rate at which these lines are scrolling.

Does not occur on consoles that use NESRGB. all of my famicoms are modded, unfortunately.

PCB photos for the TGD6+ can be found here:
https://www.famicomdisksystem.com/game-doctor-copiers/
I can upload higher resolution photos if they would prove useful.

any ideas about this issue?
Last edited by famiac on Fri Jun 04, 2021 6:22 pm, edited 1 time in total.
pakosup
Posts: 58
Joined: Mon Apr 24, 2017 10:23 pm

Re: Glitched sprites using Game Doctor 6+ with Playchoice PPU

Post by pakosup »

For me it looks like low voltage on address bus, so rising front delay happen. Maybe I'm wrong.
famiac
Posts: 19
Joined: Mon Dec 26, 2011 5:22 pm

Re: Glitched sprites using Game Doctor 6+ with Playchoice PPU

Post by famiac »

which address bus? do you think there is a loading issue?
i can take a look with an oscilloscope if you can specify
pakosup
Posts: 58
Joined: Mon Apr 24, 2017 10:23 pm

Re: Glitched sprites using Game Doctor 6+ with Playchoice PPU

Post by pakosup »

CHR memory address bus. The noise is appearing due to not correct address set for chr memory, but after some time address becomes valid, and we see sprites more clear then the noise. So the question is why this delay happen. It can be result of slow memory, but device works fine with another console. And I think that logic level can be the cause. You should check contacts and clean them if necessary and check the voltage on CHR. Also it would be nice to check /oe and /ce signals on chr memory.
User avatar
aquasnake
Posts: 515
Joined: Fri Sep 13, 2019 11:22 pm

Re: Glitched sprites using Game Doctor 6+ with Playchoice PPU

Post by aquasnake »

Most of the time, the glitch of graphics is related to the RAM adapter/ROM emulator/romter used in your system.

This mechanism of using RAM to emulate ROM requires that the speed should not be too fast or too slow, the timing is very critical
famiac
Posts: 19
Joined: Mon Dec 26, 2011 5:22 pm

Re: Glitched sprites using Game Doctor 6+ with Playchoice PPU

Post by famiac »

OK. My AV famicom has a CD74HCT373E on the AD0-7 address lines, which has a 35ns propagation delay, and the on-board RAM chips are two BR6216C-10LL.

CHR RAM on the TGD6+ is stored on two TC514256AP-10 which are CMOS/TTL compatible. They are MUX-ed by two CD74HC157E chips. A GAL chip is involved in the process somehow. It's being driven by a 24MHz crystal.

it may be worth noting that on the playchoice pcb, the PPU ALE pin output goes through two buffers, which adds a 30ns delay
And, on the vs unisystem pcb, the output to the /CS pin passes through a 1.2kOhm resistor in parallel with a switching diode, which are shunted to ground with a 33pf cap before arriving to the PPU.

i'm not sure if these minor differences would have a significant enough effect on the timing of these signals, or if these specific pins would be relevant to the issue

EDIT: Added a scope shot of the PPU ALE pin vs AD7. There's quite a bit of noise, and the voltage does like to stay below 4V sometimes, but the chips should be tolerant 'til about 3V. I also checked PPU /CS but it looks clean despite some ringing.
I'm not sure if you want me to check CHR RAM of TGD6+ or the OE and CS of the on-board RAM.

EDIT 2: I resolved the issue by changing the clock source for the GAL16V8 chip from a 24MHz crystal to a 32MHz oscillator (i had it lying around). Either the clock signal was not clean enough for the chip or the speed was too slow for the multiplexing of the RAM.
I first tried changing to a 21MHz crystal but that worsened the graphical glitches. The fact that a 1.3x change in clock speed did not break anything makes me think that the GAL is doing sequential logic and 24MHz is some kind of minimum speed.
Attachments
PPU CS.png
PPU ALE vs AD7.png
Tomy
Posts: 56
Joined: Sat May 06, 2006 9:19 am
Contact:

Re: Glitched sprites using Game Doctor 6+ with Playchoice PPU

Post by Tomy »

Good job !!

As I know, the GAL16V8 chip is a DRAM Controller. So, increase the speed of clock source will result of read/write and refresh speed faster.
Post Reply