Final Fight minor sprite/object issue

[kernow]

http://forum.arcadeotaku.com/viewtopic.php?f=26&t=12284

I got this recently for free, when I plugged it in first it was completely dead. I took a closer look and noticed two of the ROM's had a leg poking outside of the socket, I pulled them, straightened the legs and re-inserted and woo! I have a working board.

However I am getting a bit of flashing lines on some of the objects, bg's are fine, Cody, Guy, stuff like phone boxes and barrels have a bit of white outline on them where I think its supposed to be black.

If I can just get this sorted the board will be 100%, however I am not sure what to attempt next. The JAMMA edge connector for some reason has solder on each pad, so I might remove this I suppose and clean it up a bit.

I've made sure all the ROM's are sitting tightly in the sockets.

Any ideas? Thanks

 

[MKL]

If you have another CPS board, swap the motherboards. It could be a problem with the OBJ ram on the A-board. Have you already done the RAM test?

 

[kernow]

It passes the short RAM test on boot, it flashes up but its took quick to see, I can see it passes though.

No, I don't have another A board sadly.

Another thing I've noticed on the motherboard is that it had a couple of resistors? joined to some of the chips, it looks factory but I'm not so sure.

This is the closest revision I can find on the internet, 4 instead of 5 like I have, but theres none of those components on the legs of the chips 6J/8J. .

 

[kernow]

http://arcade.ym2149.com/pcb/capcom/daimakaimura_cps1mobo_pcb_partside.jpg

Found the same revision as I have (-5) and it doesn't have those components on chips 6J/8J

I noticed when I put my hand around that area on the bottom it cleared up a bit too, is it stupid to try removing them?

 

[malignantpoodle]

Interesting that it has those caps on 6J/8J... I don't know what for. Me personally, I'd feel comfortable removing them and see what gives. At your own risk of course...

 

[kernow]

I'm tempted to, but not sure to be honest.

 

[malignantpoodle]

What about making a bridge where the caps used to be? Actually, that could be worse.

A bad or missing cap rarely does any permanent damage, especially when connected to an IC. At least, that's my experience with them, but I mostly build radio transmitters and active antennas.

 

[malignantpoodle]

Sorry for the double post, I can't edit my posts yet;

Are those disc caps wave soldered or done by hand?

 

[malignantpoodle]

Ok, went back and looked at the pics and did some thinking.

My GUESS is that the board was having some kind of issue, so solder was applied to the JAMMA connector in an attempt to establish better contact. Failing that, the disc caps were added to 6J and 8J if and when the problem was isolated.

Looking at the pic, definitely done by hand. So what we can surmise by that is that the board is by default not supposed to have those. Removing them at worst would probably render the board useless or with problems as it must have been powered up without the caps in the past, right?

If you remove the caps and the board doesn't work right, then I'd replace 6J and 8J chips entirely. That's just me, and my guess.

 

[kernow]

I'll have a think about it, they don't look 'standard' as such. I'll wait for channelmaniac to see this thread also

 

[metallizer]

just desolder one leg to break the connection and see what happens.

only do it if you´re capable of soldering back.

 

[DaytimeDreamer]

I dont know what revision my Final Fight PCB is but i will take a look at it when I get back from vacations. Will post some pics if its the same rev as yours just to compare boards

My board however has another issue. The sound suddenly died on me a couple of years ago. Maybe its a leaking cap who knows

 

[kernow]

Well, removing a leg from each of those caps did nothing pretty much, I'll just leave it now as its beyond my skills

 

[malignantpoodle]

Well, removing a leg from each of those caps did nothing pretty much, I'll just leave it now as its beyond my skills

Nothing eh? Put them back then.

Next step if you have a multimeter see if any of those pins on the JAMMA connector are bridged.

 

[MKL]

The caps are soldered across the output enable pin of those graphics buffers and ground, probably because there was instability issues with those two buffers in particular as all the other ones which have the OE in parallel with them are OK without caps.

 

[kernow]

I'll reattach the leg later on then Can't hurt.

edit: Done.

 

[channelmaniac]

Touch a small area of the board then move on to the next small area. Try to narrow down exactly where the issues are and you'll narrow down what chips to replace.

It's VERY possible for RAM to fail self tests and still glitch graphics. It's all in the timing.

RJ

 

[kernow]

The underside while running right? it was when I was touching it while on and feeling a slight tingle I noticed I could change the graphics a bit, and it was under where those caps were. Will try again later.

I might just end up giving this A board away to someone more knowledgeable and getting another one.

 

[Kyuusaku]

The caps are there to stretch the read pulse. Probably it's there from chip propogation speed issues, not a big deal. I think the problem is an address line to one of the VRAM (wherever attributes are stored) causing it to fetch the wrong palette.

 

[kernow]

The caps are there to stretch the read pulse. Probably it's there from chip propogation speed issues, not a big deal. I think the problem is an address line to one of the VRAM (wherever attributes are stored) causing it to fetch the wrong palette.

Would that be a visible gouge as such? or something as simple as reinserting the ROM's I straightened the legs on?

Sorry for such a dumb question, I guess we all have to start somewhere hehe

 

[smkdan]

Something like that will not be caused by the GFX ROMs. The tiles being fetched are correct and there's no artifacts visible with the actual tile data. Like kyuusakku said something on the main board could be causing incorrect palette reference or the RAM bits that contains the black palette reference is bad.

All I know is that the primary palette RAM is stored separately and can't be software tested. It passes through the OBJ RAM (that you say is good) and then gets stored somewhere else on the board. Wherever it is.

 

[kernow]

Touch a small area of the board then move on to the next small area. Try to narrow down exactly where the issues are and you'll narrow down what chips to replace.

It's VERY possible for RAM to fail self tests and still glitch graphics. It's all in the timing.

RJ

This works to some extent, if I touch my hand on a certain area at the back of the motherboard on the bottom, it clears up a little bit, but not 100% perfect.

Something like that will not be caused by the GFX ROMs. The tiles being fetched are correct and there's no artifacts visible with the actual tile data. Like kyuusakku said something on the main board could be causing incorrect palette reference or the RAM bits that contains the black palette reference is bad.

All I know is that the primary palette RAM is stored separately and can't be software tested. It passes through the OBJ RAM (that you say is good) and then gets stored somewhere else on the board. Wherever it is.

You're right, removing the ROM's made no difference so its got to be a RAM issue on the motherboard I believe.

It just surprises me it looks so clean and I can't see any bad solder joints, joined pins or broken legs etc, but I guess you're not going to visually see a faulty RAM chip, sigh.

 

[Xian Xi]

I'd still check the connection on the sockets.

 

[kernow]

So, I got sent a spare 'broken' final fight C board just to see what it'd do when I plugged it in, I get it today, plug it in, and BAM problem fixed. what the hell? It was causing errors on their board

Thanks for the help and tips anyway people. yay!

 

[DaytimeDreamer]

Enjoy this godly made game!