What's new
By:

WWF Wrestlefest that does nothing.

Yes, I can do that and will tell you.
(The board does not reset with original roms)
I remember having done it already and the romident found a valid romset.

What do you think about the bg ram test ? Is it valid ?
 
Interesting, I wonder if my board has a busted watchdog. I had to do a couple repairs to the board to get it working originally.

The brief bg ram error says its unable to write to the bg ram. Which would imply some issue with the R/W signal from the CPU getting to the bg ram WE pin (via one of the TJ chips probably)

Can you give this a try

https://www.mvs-scans.com/misc/wwfwfest-mad-27c010-3.zip

It adds the watchdog ping.
 
Now it is working !
I occasionally had some resets but I could run all tests without issue. They all passed except bg. Well, it seems i have this bg issue to fix. I do see a pulse on the WE pin of the bg ram when the test is started but I get the error. Do you think I can use the ram model I mentioned as replacement in my previous post ?

One more thing : when the error occurs on the bg test, there is no button combination to exit the test. I needed to reset the board manually.

Let me know if you want me to test other versions of the diag rom. I would be happy to help.

Thanks a lot
 
ragefan said:
I occasionally had some resets but I could run all tests without issue. They all passed except bg. Well, it seems i have this bg issue to fix. I do see a pulse on the WE pin of the bg ram when the test is started but I get the error. Do you think I can use the ram model I mentioned as replacement in my previous post ?
Click to expand...
The ram model you listed before should be compatible.

The writable test works by doing
  • read word from ram
  • flip bits in read word
  • write flipped word to same ram location
  • re-read word from same ram location
  • compare originally read word with newly read word
  • if they match assume memory is unwritable
It looks like TJ-002 sits between the CPU and the BG ram. Perhaps there is something busted there causing the TJ-002 to always output the same data? If you select "BG1 RAM" under the memory viewer, what does it show? It should be all 0x00.

I assume the video dac test triggers the watchdog? any other locations that you can see that trigger it? I already had provisions in place for watchdog in the common code because of TMNT, but I might need to tweak them for this board. I will see if I can get my hands on another wrestlefest board or figure out whats up with mine.

ragefan said:
One more thing : when the error occurs on the bg test, there is no button combination to exit the test. I needed to reset the board manually.
Click to expand...

This is expected for now.
 
Indeed only video dac test actually triggers watchdog. BG1 Ram in memory viewer is total garbage. (Not 0 at all !). Same for BG2.
What to conclude ?
 

Attachments

  • PXL_20240819_184206406.MP.jpg
    PXL_20240819_184206406.MP.jpg
    231.3 KB · Views: 72
That kinda implies writes aren't happening and ram is just filled with gibberish since its never been initialized. Give this a try

https://www.mvs-scans.com/misc/wwfwfest-mad-27c010-4.zip

I hacked up the memory viewer so when you press button 1 it will write 0x01010101 to the first address line on the screen. The next time you press button 1 it will write 0x02020202, etc as you press is again and again. Maybe this can help you track down whats going on with writes?
 
Tried it :)

3 buttons on memory viewer :

- 1 does nothing
- 1 switches between word/byte. I see WE pulse when pressing with my probe.
-1 is back to previous screen. I see WE pulse when pressing with my probe. Scope shows a signal.

Is that expected ?

I cannot read anything new in the ram itself. Nothing is written.

Both rams in error seems strange (BG2 is having the same garbage). The signal is coming from TJ-002 which seems to react correctly. Now either I change 1 ram, either I reflow TJ-002 ? What would you do ?

By the way, I see another case where the boards resets : launch sprite ram tests, pause it for few seconds by pressing the button, it resets after a while.
 
ragefan said:
3 buttons on memory viewer :

- 1 does nothing
- 1 switches between word/byte. I see WE pulse when pressing with my probe.
-1 is back to previous screen. I see WE pulse when pressing with my probe. Scope shows a signal.
Click to expand...
I'm unclear what you are saying here

button 1 = switch between word/byte read + hack I mentioned above to write to ram
button 2 = exit memory viewer
up/down/left/right = scroll memory

ragefan said:
I cannot read anything new in the ram itself. Nothing is written.

Both rams in error seems strange (BG2 is having the same garbage). The signal is coming from TJ-002 which seems to react correctly. Now either I change 1 ram, either I reflow TJ-002 ? What would you do ?
Click to expand...
BG1/BG2 are the same ram chips (IC16/upper byte and IC17/lower byte). Its weird that both chips are having the issue. What kinda of activity do you see on the CE1 and CE2 pins of the ram chips?
 
Sorry if was unclear.
Button 1 changes from word to byte read. I see a pulse on WE everytime I push it but no change on screen.

CE1 is low on both chip.
CE2 is high on both chip.
 
How about the OE line when you do the write? Does it pulse high?
 
OE will pulse high for me when doing the BG write test in the memory viewer. My BG ram chips are UM6264AK-10L and the datasheet says

Screenshot 2024-08-19 at 3.24.03 PM.png


Its not clear to me if the high pulse is needed or not. Write Cycle 1 shows OE should be high, but write cycle 2 says low. The sony chips you listed seem to completely ignore OE when WE is low. So dunno.
 
ack said:
OE will pulse high for me when doing the BG write test in the memory viewer. My BG ram chips are UM6264AK-10L and the datasheet says

Screenshot 2024-08-19 at 3.24.03 PM.png


Its not clear to me if the high pulse is needed or not. Write Cycle 1 shows OE should be high, but write cycle 2 says low. The sony chips you listed seem to completely ignore OE when WE is low. So dunno.
Click to expand...
I will try to change one by this week end and see what happens, moving from UM62 to sony one.
 
I unsoldered one of the ram... And tested it off circuit. It is correct. So, the problem comes from the custom chip as the OE is directly connected to it. What is weird is that I do see a pulse when the test starts. Maybe the signal is not "clean" enough to enable the ram ? Or could it be only one chip failing (the other one) leading to both being reported in error by the test ?
 
do you get the pulses when you are in the bg ram memory viewer and pressing button 1?
 
Yes when it's changing from word to byte read at the screen.(Button 1). I see 1 pulse (only one) with my logic probe everytime I push the button.
 
@ack, it is kind of magic... I unsoldered 1 ram, put a socket , changed the ram with another model and ... now I can see your memory viewer working on BG1 (BG2 failing). Changed again to the original BG1 ram ... Still working... Did the same with the BG2, now the whole test is working ... with the original rams.🫣🫣🫣😂😂... Background is back ! I still have a sprite glitch remaining. Shall I conclude it is a mask rom ?
 
  • Like
Reactions: ack
@ack great job on the test roms, I tried the Wrestlefest one I did notice the screen stays black on sprite ram test though. It does display when pressing button one but blink and you miss it, not sure if this is a bug?

Would it be possible to add a rom test to test the graphics maskroms?
 
Ace` said:
@ack great job on the test roms, I tried the Wrestlefest one I did notice the screen stays black on sprite ram test though. It does display when pressing button one but blink and you miss it, not sure if this is a bug?
Click to expand...
This is expected. The foreground (text) and sprites share the same ram chips. The test is testing those ram chips which overwrites whats on screen. If you press and hold the first button it will pause the test and do full re-draw of the screen.

Ace` said:
Would it be possible to add a rom test to test the graphics maskroms?
Click to expand...

The cpu doesn't have direct access the graphics mask roms, so there isn't much that can be done. There is a tile viewer under the debug menu that will display foreground graphics, not sure how much that will help. At somepoint I will get around to adding a sprite veiwer, but not looking forward to it since each board does its own thing for sprites.

Are you using the rom from this thread or building it from source? The one in this thread might pre-date me adding the tile viewer (and other stuff).
 
You must log in or register to reply here.
Back
Top