Tron RAM wierdness

mcandrewsoun

Well-known member
Joined
Feb 26, 2010
Messages
1,602
Reaction score
63
Location
Fincastle, Virginia
On my Tron boardset, the video has a bunch of horizontal lines in it..kinda like a noise..I can make them go away by selectively yanking buffer rams. It seemed to get worse with brand new rams (AM93L422).. Are there different speeds of these that dont work in some games? I am starting to think there is more at play, because it is doing this on 6 of the 8 original ram chips and all of the new ones.. Anyone seen this?

Andrew
 
On my Tron boardset, the video has a bunch of horizontal lines in it..kinda like a noise..I can make them go away by selectively yanking buffer rams. It seemed to get worse with brand new rams (AM93L422).. Are there different speeds of these that dont work in some games? I am starting to think there is more at play, because it is doing this on 6 of the 8 original ram chips and all of the new ones.. Anyone seen this?

Andrew

Just to let you know Andrew all the RAM and EPROMS came off a working Video Gen board I had. The chips I took off that board worked too so I never swapped them back.

I also found out that if you lay the boards on carpet when powered up it can cause video distortion also, prob static electricity.
 
Last edited:
I think it might have something to do with that..I think theres an issue in the data loading to the buffer ram, or somewhere up the chain.. If I ground out all the bit inputs the static goes away, and its clean..
 
Fixed!

Fixed..Found a 74LS32 that was dead, popped in a new one, and its alive! Now..my spinner turns objects counter-clockwise irregardless of how I spin it.. Im pretty sure its on the cp (as it was a documented issue when I bought it)..

Andrew
 
Fixed..Found a 74LS32 that was dead, popped in a new one, and its alive! Now..my spinner turns objects counter-clockwise irregardless of how I spin it.. Im pretty sure its on the cp (as it was a documented issue when I bought it)..

Andrew



Yep it's got to be the CP, the controls worked fine when I tested your board. Hey good work man! Glad you got it working :)

Maybe post a little repair log on here to give some helpful info on troubleshootin MCR stuff.
 
Mini repair log...

Heres how I found it..
2 tools needed. Oscilloscope & Schematic/Flowchart. Basically, i was getting no foreground objects and bunches of lines on the screen. I tried ruling out the rams, by disabling them by grounding out the enable pins.. I could make the lines go away, but it wasnt consistant.. I tried grounding out the individual ins and outs on the buffer ram, and I found that I could clear the lines bu grounding out the ram inputs. So that told me that the problem was upstream of the rams, and that nothing was getting written to them. The flowchart showed that the foreground objects basically get loaded from the roms through a shift register, a latch, and then into the buffer ram. I figured either the clock was missing or one of the latch chips was dead, because it was a wholistic problem, not a corruption. Using the schematic, I take the scope and start looking at the clock and enable pins, and its flatlined... So I followed the logic, and found an OR gate that had a live signal on the input pins and 0v on the output pin. I replaced that chip and it came up.

Kinda the same thing for the SSIO board.. I think the watchdogging I saw might have been something I had hooked up wrong.. but I actually started checking clock signals first..Usually when the whole board is dead I check the processor and clock first. Sure enough the clock pin on the processor was dead.. traced the circuit and found a logic gate dead..

BTW the spinner issue was I think just a dirty optical sensor. I replaced the 339 because I thought it was dead, but that didnt fix it.. I was getting 2.8vs on the out of one of the sensors and 6v on the other. After messing with it, I started gettign 6 on the first one, and 2.8 on the second..hmm... kept cleaning, and got 6 on both... I did resolder all the pins too, so could have been that too..
 
Heres how I found it..
2 tools needed. Oscilloscope & Schematic/Flowchart. Basically, i was getting no foreground objects and bunches of lines on the screen. I tried ruling out the rams, by disabling them by grounding out the enable pins.. I could make the lines go away, but it wasnt consistant.. I tried grounding out the individual ins and outs on the buffer ram, and I found that I could clear the lines bu grounding out the ram inputs. So that told me that the problem was upstream of the rams, and that nothing was getting written to them. The flowchart showed that the foreground objects basically get loaded from the roms through a shift register, a latch, and then into the buffer ram. I figured either the clock was missing or one of the latch chips was dead, because it was a wholistic problem, not a corruption. Using the schematic, I take the scope and start looking at the clock and enable pins, and its flatlined... So I followed the logic, and found an OR gate that had a live signal on the input pins and 0v on the output pin. I replaced that chip and it came up.

Kinda the same thing for the SSIO board.. I think the watchdogging I saw might have been something I had hooked up wrong.. but I actually started checking clock signals first..Usually when the whole board is dead I check the processor and clock first. Sure enough the clock pin on the processor was dead.. traced the circuit and found a logic gate dead..

BTW the spinner issue was I think just a dirty optical sensor. I replaced the 339 because I thought it was dead, but that didnt fix it.. I was getting 2.8vs on the out of one of the sensors and 6v on the other. After messing with it, I started gettign 6 on the first one, and 2.8 on the second..hmm... kept cleaning, and got 6 on both... I did resolder all the pins too, so could have been that too..

Thanx Andrew, now instead of an Oscilloscope can one of those logic probes be used to troubleshoot these boards?

Like this one..
http://cgi.ebay.com/50-Mhz-HIGH-SPE...tem&pt=LH_DefaultDomain_0&hash=item53e10deaa0
 
Rob, Im sure that you can. I dont have a logic probe, but I think thats actually what that tool is for.

That's what I need! A tutorial on using a logic probe to debug board problems... An oscilloscope is too expensive!!
 
Is it good? I've thought about getting one before when working on Williams Pinball stuff...Allot of the resources speak to using them, though I've never seen a walkthrough on using it..I'm sure its pretty simply, I've just never used it before!!



Don't know, just ordered it. 50% of pinball issues are resolved with a multi-meter being alot of the problems are playfield issues and wire connectors.
 
a logic probe works just as well for most diagnostics. It tells you whether the logic level is high or low, or pulsing, or floating (no connection, or in between high and low).
The trick is figuring out what that specific pin is supposed to be doing. To determine that, you need to look at the datasheet for the chip (or something similar, like this - http://www.ionpool.net/arcade/giicm/GIICM.html), and the schematics.
In some cases, if your not sure if an ic is responding properly, you can look at the datasheet, and bend the outputs of the same chip out, then double the chip up on top of the chip on the board, and use your logic probe to compare the outputs. If they're different, then chances are, that chip is bad (or the one down the line could be holding it low, or high....).
There's not a simple answer to how to diagnose problems at the board level. There's reasons people like to work on certain boards. The more of them you do, the easier it gets. I will say this, though. There is nothing more rewarding than your first successful, "on your own", "i figured it out myself" board repair.
 
Back
Top Bottom