AES Randomly Soft Reboots

[Xian Xi]

Different from my other resetting AES thread.

Installed a Unibios and everything checks out, installed the SMK Diagnostics and everything checks out as well.

I did notice that the Z80 and YM2610 get pretty hot to the touch and if I put the AES in MVS mode there is no sound in the attract mode, is that normal?

Tested it with FF3 and was playing and about 3 matches in it just rebooted. Customer said he was playing Samurai Shodown and the same thing happened to him.

When it reboots, the Unibios doesn't display any exceptions at all, almost as if I hit the reset button. CRC comes out fine as well.

Could this be a problem from the Z80 and YM2610 comm to the 68K? Like sporadic noise or something that triggers a reset?


EDIT: Finally after many tries I got a result from the SMK-D bios,

Z80 Dead / Errored / Comm Issue

Expected: 00
Actual: FF

So is this Z80 going bad?

EDIT2: Z80 is not dead, swapped in a new one and got the same error.

 

[Neo Alec]

Finally AES behavior that resembles my first two used systems. One of the reasons I switched to MVS was reliability. Sorry I have no idea about it.

 

[NEO-GEO man]

Yes its normal for no sound in attract mode on an AES set to MVS mode.

 

[HeavyMachineGoob]

Concerning the random resetting, could it be a dirty/bad reset switch? I've also had dirty carts reset too.

 

[Tacitus]

Stupid question:

Are the cart slots clean?

 

[Xian Xi]

Concerning the random resetting, could it be a dirty/bad reset switch? I've also had dirty carts reset too.

Right now it's been narrowed down to either a faulty 68k or faulty Neo-B1.

Stupid question:

Are the cart slots clean?

Yup, very clean.

 

[HeavyMachineGoob]

The 68k has a reset pin doesn't it? There's no cold bridging or bits of metal on that line?

If the 68k was faulty, wouldn't the system be resetting at specific places?

 

[Xian Xi]

The 68k has a reset pin doesn't it? There's no cold bridging or bits of metal on that line?

If the 68k was faulty, wouldn't the system be resetting at specific places?

No bridges. If the 68k is faulty anything can happen. I just hope it's not the Neo-B1 as that would be a more expensive fix for the customer and I don't want them to say to just forget it because it's too high of a cost.

 

[smkdan]

Anything particular about the B1 that is making you suspect it? If it's a bad component causing it then the B1 could be it but so could most of the other customs. Since random resets are annoying to fix by just swapping components, have you ruled out stuff like a loose barrel jack for power? Nothing wrong with the power circuit? I'm guessing the constant WRAM test came out fine too?

 

[Xian Xi]

Anything particular about the B1 that is making you suspect it? If it's a bad component causing it then the B1 could be it but so could most of the other customs. Since random resets are annoying to fix by just swapping components, have you ruled out stuff like a loose barrel jack for power? Nothing wrong with the power circuit? I'm guessing the constant WRAM test came out fine too?

When the reset happens the power doesn't dip, it's generated like a normal soft reboot via the reset pin on the 68k going low then high again. I don't know if the 68k is failing to reset the counter or not and the B1 is kicking it.

 

[channelmaniac]

First thing I'd check if it were on my bench would be to see if there's any gunk around the socket that's eating through traces... and then I'd tap on various surface mount chips to see if you have a bad solder joint somewhere.

Too bad there isn't a mode where you can turn all DIP switches on and leave it running RAM tests. That would be another thing to try.

 

[Xian Xi]

Everything is clean on this board. Only corrosion on the board was around the LO rom but nothing breaking connection so I cleaned it up and covered it with polish. The QFPs are all solid, no loose pins. I need to burn version 16 of smkdan's diagnostics as it has a work ram loop test. I'll report back after that.

The 68k has 100% connection to the cart slot BTW as far as data and address lines.

 

[northlab]

Sorry for necroposting this old thread, but I can't post a new thread and this is one of the only ones on random resets that I can even find.

My random resets doesn’t follow any pattern, sometimes I can play for as long as I want, other times it resets after a few minutes in game and other times right at the boot logo.

The board checks out fine visually and when booting it up the only thing I can see that could
indicate that something is wrong is some light interference (over composite video) for a couple of seconds, which then calms down.

The test mode gives a blue screen.

I tried resoldering the CXA1145 legs, swapping the main 2200uf cap and tapping around on the different ic’s, but there is no difference. My Z80 also gets quite warm as Xian Xi stated in the first post.

Is there any knowledge about these random resets at all? Should I look into new ram modules? Whatever clues you guys can offer would be highly appreciated!

 

[Xian Xi]

Sorry for necroposting this old thread, but I can't post a new thread and this is one of the only ones on random resets that I can even find.

My random resets doesn’t follow any pattern, sometimes I can play for as long as I want, other times it resets after a few minutes in game and other times right at the boot logo.

The board checks out fine visually and when booting it up the only thing I can see that could
indicate that something is wrong is some light interference (over composite video) for a couple of seconds, which then calms down.

The test mode gives a blue screen.

I tried resoldering the CXA1145 legs, swapping the main 2200uf cap and tapping around on the different ic’s, but there is no difference. My Z80 also gets quite warm as Xian Xi stated in the first post.

Is there any knowledge about these random resets at all? Should I look into new ram modules? Whatever clues you guys can offer would be highly appreciated!

Check the power socket if it's corroded.

 

[wyndcrosser]

I had the same issue, but the rebooting only seemed to happen with larger MEG count games. I had to replace the power adapter. What adapter are you using Northlab?

XX said check out the power socket as well.

 

[kaicer]

I bought an AES that was having random reboots for some strange reason was the video din connector when I move the cable it reboots. At the moment the AES is working fine.

 

[StevenK]

I find it unnerving to think of Xian asking other people for neo answers.

 

[wyndcrosser]

I find it unnerving to think of Xian asking other people for neo answers.

lol, it's also from 2012. Look at posts from ANY of the "big-thymers" from 2006 on back.

 

[northlab]

Thanks for the input guys!

I checked the power socket and it sure was dirty. I used some contact cleaner and a flat screwdriver to fix it up somewhat but there's no difference. Also I can't force the reboot by wiggling the adapter so the only thing left there is to try if soldering the psu cable directly to the board makes a difference...

I use a Genesis 1 PSU which should be sufficient for my board revision.

One clue I've discovered is that it seems to correspond with joystick input. If I leave it in attract or in menu there's no reboot (30+ min).

 

[Xian Xi]

Don't solder it to the board. If you wiggle the power cord and it doesn't reboot then it isn't the problem.

 

[Atro]

One clue I've discovered is that it seems to correspond with joystick input. If I leave it in attract or in menu there's no reboot (30+ min).

If it reboots with joystick activity, check if you have the +5v wire loose.
It may be touching ground point and that leads directly to a reset by shorting.

Use another controller to be sure. If it's the case, it's an easy joystick fix.