MV-1 turns on, now has the click of death

[HeavyMachineGoob]

Got a broken MV-1. When I plug it into my Supergun, the LED on the board won't turn on. There's two LEDs for 5 and 12v on my Supergun as well. The 5v LED won't light up, but the 12v dims on and off (lit, dims to off. lights up instantly, cycle repeat).

I replaced the 7805 and there's no change. What's wrong with this board?

EDIT: Power up problem solved, I think a pin on one of the big black connectors was touching two others, causing 5v to be grounded. Scroll down to post #16 for the latest problem.

 

[Xian Xi]

What supergun are you using?

Sounds like short.

 

[HeavyMachineGoob]

Home made one. Theoretically JAMMA compliant. I have a second working MVS board that works fine with this SG. When plugged into this MV-1, it has that no power up problem.

Yeah, a short would be the only possible explanation as to why the Supergun doesn't power up properly either. But why would 12v turn on and dim off at a consistent tempo?

 

[Xian Xi]

Yeah, a short would be the only possible explanation as to why the Supergun doesn't power up properly either. But why would 12v turn on and dim off at a consistent tempo?

Hard to say, I don't know the specs of your SG. If you're using an ATX PSU it's possible to happen.

 

[HeavyMachineGoob]

Nope, I'm using a dual voltage PSU off JNX.

Actually that gives me an idea... While either voltage connected to ground wouldn't do anything, what would happen to a Neo Geo board if 5 and 12v were touched together?

 

[mjs256]

Yikes! Don't touch 12v to the 5v line! That would create a short and could fry your power supply, and half the components on your board! To investigate the problem I would suggest a multimeter that can measure resistance and voltage. Disconnect the power supply from everything and try measuring the voltage at the 5v line with nothing attached. If you don't get 5v between the 5v line and GND then you know you have a bad power supply. With everything still disconnected, try measing the resistance between 5v and GND at the JAMMA interface - you should get a resistance of the order of hundreds of ohms. If you get zero or near zero then you know you have a short. If possible, disconnect the supergun from the MVS board and similarly measure the resistance between the 5v line and GND.

 

[Xian Xi]

Ya when you solder to the 5 pin din you should use heat shrink so that they don't touch at all. Check the din connector and make sure they aren't.

 

[HeavyMachineGoob]

Err no I was just asking about 5 and 12v shorting in case that happened with this PCB. The PSU and Supergun work with a different MVS board (MV-1AX). The dead board is the MV-1 and it not only doesn't power on, but the PSU seems as if it shorts too.

Since the PSU does work, I'll try measuring voltage and resistance on the MV-1. Wouldn't touching 5v and gnd with my multimeter to check resistance short out the voltage though?

 

[shadows]

Wouldn't touching 5v and gnd with my multimeter to check resistance short out the voltage though?

You dont test that while its powered on.

 

[HeavyMachineGoob]

Dumb question, but if there's no current, how can the resistance be measured? Sounds impossible to me.

Well, with normal operation (MV-1AX), my PSU outputs 4.99v on the +5v line and 11.94 for +12v. When plugged into the MV-1, 5v is virtually nothing. 12v floats between 2 and 3 volts.

 

[mjs256]

It sounds like the MV-1 board probably has a short somewhere on the 5v line, but use a multimeter to verify. As mentioned, you should measure the resistance with the power turned off. You are correct that you cannot measure the resistance without any current flow, but that is the multimeter's job - my understanding is that the multimeter supplies the circuit you are testing with a tiny voltage which induces a tiny amount of current to flow, and then figures out the resistance based on the tiny voltage and the tiny current.

If you measure zero (or very near zero) resistance between 5v and GND on the MV-1 board (and I suspect you will) then you will need to track down the short. The first thing to do would be to look at the underside of the main board to make sure none of those little pins sticking through are touching each other. Also look for signs of physical damage to the board. If you do not find any issues there then I would remove all the capacitors and then check the resistance again between 5v and GND to see if the short is gone. Capacitors are quite easy to remove and they can cause shorts (especially if someone accidentally put one in backwards!). If you still have not found the short after that then things will get considerably harder. For now let's hope this will be enough to find and fix the short.

 

[bustedstr8]

I found that a common suspect in shorted 5V are QFPs. A small piece of metal shaving can get caught in between the legs or a small bump can knock two legs together. Grab a loupe and give them a quick look.

 

[Nightmare Tony]

the 7805 on the neo geo only provides 5 volts to teh sound circuits and wont cause a 5 volt shoret. that sounds like one somewhere.

A resistance seeintg on a metter run s a TINY bit if current to measure resistance by...

 

[HeavyMachineGoob]

Well, I didn't do much looking over the board, but I tried checking resistance (seemed like the same method I do continuity with) and yeah I got connection between 5v and ground... I'll work with this board more thoroughly another day (already quite late where I am). I'm not holding my breath over the PCB though. Thanks for all the help so far guys, very awesome of you all.

 

[Xian Xi]

On an MV1 take a look at the DIP ICs. Sometimes I see repairs that bridge pins on the DIP ICs. Also make sure there are no bent pins on the riser card.

 

[HeavyMachineGoob]

Well, I looked over the board a lot more thoroughly today. Checked the caps, QFPs, DIP ICs, all the legs underneath the board etc... I didn't find a whole lot wrong, cleaned the board a bit more, didn't find any bent pins in the riser card plugs. I did find one odd thing though, on one of the big black connectors on the left side of the board (relative to where the JAMMA edge is), I found one pin was bent far enough to touch two others, so I bent it slowly back into position. I powered the board up with no riser card, LED came on and my PSU wasn't grounding out. I then plugged the riser card in and re-attached the metal cartridge guide. Again, successfully turned on. Thanks for all the help so far guys!

Now onto level 2 I guess heh, it's stuck in POST. I don't know what exactly the board is doing, must be some corroded traces (or bad chips). The signal is probably unstable, since a CRT won't pick the video signal up. It works with a Plasma though, so here's some pictures.

This displays with no cartridge inserted:

With a game plugged in, this happens.

Please keep in mind that with both pictures, the video signal flashes different garbage partially. The images above aren't static like the pictures may suggest.

I notice the board sort of looks corroded in a couple places, like one side of the big ASIC in the middle, the traces on one side of a little surface mount DIP chip and a couple other parts. I get the impression maybe this board isn't worth fixing, but hey I'm not the doctor here, perhaps it's not as bad as I think. I can take pictures of the PCB if you guys want.

EDIT: Pics of suspicious parts of the board added (and one for just the general sound area...) I can provide larger versions of these images if need be.

On one other note, I put the cart I tested with the MV-1 into a working MV-1AX and it go to a black screen after post. I wiped the cart edges with a shirt and black dirt/dust came off, so me thinks the MV-1 is also suffering from a dirty cartridge slot as well.

 

[dragonpt]

Well, I looked over the board a lot more thoroughly today. Checked the caps, QFPs, DIP ICs, all the legs underneath the board etc... I didn't find a whole lot wrong, cleaned the board a bit more, didn't find any bent pins in the riser card plugs. I did find one odd thing though, on one of the big black connectors on the left side of the board (relative to where the JAMMA edge is), I found one pin was bent far enough to touch two others, so I bent it slowly back into position. I powered the board up with no riser card, LED came on and my PSU wasn't grounding out. I then plugged the riser card in and re-attached the metal cartridge guide. Again, successfully turned on. Thanks for all the help so far guys!

Now onto level 2 I guess heh, it's stuck in POST. I don't know what exactly the board is doing, must be some corroded traces (or bad chips). The signal is probably unstable, since a CRT won't pick the video signal up. It works with a Plasma though, so here's some pictures.

This displays with no cartridge inserted:

With a game plugged in, this happens.

Please keep in mind that with both pictures, the video signal flashes different garbage partially. The images above aren't static like the pictures may suggest.

I notice the board sort of looks corroded in a couple places, like one side of the big ASIC in the middle, the traces on one side of a little surface mount DIP chip and a couple other parts. I get the impression maybe this board isn't worth fixing, but hey I'm not the doctor here, perhaps it's not as bad as I think. I can take pictures of the PCB if you guys want.

EDIT: Pics of suspicious parts of the board added (and one for just the general sound area...) I can provide larger versions of these images if need be.

On one other note, I put the cart I tested with the MV-1 into a working MV-1AX and it go to a black screen after post. I wiped the cart edges with a shirt and black dirt/dust came off, so me thinks the MV-1 is also suffering from a dirty cartridge slot as well.

Check, and re-check your din/socket video conections, my problem was similiar to this (looking at the pictures), also check your PSU maybe its too weak to power the board
Also you can go clean that board, put it in the washer.
Here's how mine look like :





Uploaded with ImageShack.us

 

[HeavyMachineGoob]

I don't mean to be rude, but your advice is confusing me. Firstly my problem appears to be 100% digital (eg. error in the video processing), not in any way related to analog video (encoding RGB into composite and then to the TV). How could garbage on the screen equal a problem with the AV connections? It's impossible.

Also, my PSU is just fine, it's a dual voltage PSU I bought on Jamma-Nation-X and it can comfortably handle 25w (XX has said it can do higher but whatever). I've also tried an ATX PSU with like 200w capability and there's no improvement.

And about cleaning the board, I really don't see how dust could be causing this problem, seems more like bad traces are causing the system to hang in POST. And I don't feel safe putting this thing in the dish washer, it's already had enough corrosion as it is.

Most prominently though, you forgot to say what you did to fix your faulty Neo board... Either that or your post is just plain hard to read. Did cleaning it all over fix it? Repairing traces? Replacing RAM? What fixed it??

Lastly, why did you quote my big long post? It's completely unnecessary...

 

[Xian Xi]

The PSU is more than capable of powering that board. That's a lot of corrosion dude on the LSPC, MV1s aren't very expensive but if you want to do this as a challenge then good luck dude.

 

[HeavyMachineGoob]

Hmm, I was expecting that. So I should just try fixing as many traces on the big LSPC and the little ram chip as I can and see where that gets me? Should I try and scrape corrosion off the pins or will that not help?

Also I never see dead MV-1s for less than $60, working ones are way higher. Am I just looking in the wrong places?

 

[mjs256]

Nice job tracking down and fixing the short. I'm sorry your board still is not working, but you are for sure a step closer. If you want to keep trying to fix it then I agree you should look for bad traces. Use your multimeter and you should get zero (or almost zero) resistance between where a trace starts and where it ends. Check all traces that have any signs of corrosion or damage. Also, looks can deceive. I once repaired a board that was badly corroded with a lot of bad looking traces. Using my multimeter I only found two traces that were not working, but the strange thing was that the two bad ones did not look nearly as bad as some of the other traces that still worked.

It might also be a good idea to reflow the solder under the SMDs.

Another thing you can do is start up the system with no BIOS installed. The system of course will have no chance of working without the BIOS, but it should display a predictable pattern that you can compare with a known working MV-1 without a BIOS.

For now I would just remove the board with the cartridge connectors and try to get the crosshatch screen and do the self tests and worry about inserting a cartridge later.

 

[dragonpt]

I don't mean to be rude, but your advice is confusing me. Firstly my problem appears to be 100% digital (eg. error in the video processing), not in any way related to analog video (encoding RGB into composite and then to the TV). How could garbage on the screen equal a problem with the AV connections? It's impossible.

Also, my PSU is just fine, it's a dual voltage PSU I bought on Jamma-Nation-X and it can comfortably handle 25w (XX has said it can do higher but whatever). I've also tried an ATX PSU with like 200w capability and there's no improvement.

And about cleaning the board, I really don't see how dust could be causing this problem, seems more like bad traces are causing the system to hang in POST. And I don't feel safe putting this thing in the dish washer, it's already had enough corrosion as it is.

Most prominently though, you forgot to say what you did to fix your faulty Neo board... Either that or your post is just plain hard to read. Did cleaning it all over fix it? Repairing traces? Replacing RAM? What fixed it??

Lastly, why did you quote my big long post? It's completely unnecessary...

well in that case (pictures) the problem was on the psu, your psu is a good one and new, but still in can have some problems, did you measure the amps on the 5v, 12v lines ?
also even if its a new psu, doenst mean, its working properly.

By looking the pictures, my problem looks very similar to yours.

Grab another psu (loan from a friend) and test, i only managed to power my board with a 300 or 400 watts psu/with 7 or 8 amps on the 5v line

Also cleaning the board is always a good thing to do, very dirty boards, can give sometimes problems.

B.Regards



------------- edit:

On another board i had, the problem (bad video/sync on screen), was just a case of bad conections on the din/supergun, the sync line, was conected to a whrong pin on the scart socket.

 

[Xian Xi]

well in that case (pictures) the problem was on the psu, your psu is a good one and new, but still in can have some problems, did you measure the amps on the 5v, 12v lines ?
also even if its a new psu, doenst mean, its working properly.

By looking the pictures, my problem looks very similar to yours.

Grab another psu (loan from a friend) and test, i only managed to power my board with a 300 or 400 watts psu/with 7 or 8 amps on the 5v line

Also cleaning the board is always a good thing to do, very dirty boards, can give sometimes problems.

B.Regards



------------- edit:

On another board i had, the problem (bad video/sync on screen), was just a case of bad conections on the din/supergun, the sync line, was conected to a whrong pin on the scart socket.

All 1 slot MVS boards use between 1-2A on the +5v line and that is measured. 300 watts to power an MVS board? 7 or 8 amps better be a 6 slot, anything smaller and your MVS board is pulling way too many amps.

 

[HeavyMachineGoob]

Nice job tracking down and fixing the short. I'm sorry your board still is not working, but you are for sure a step closer. If you want to keep trying to fix it then I agree you should look for bad traces. Use your multimeter and you should get zero (or almost zero) resistance between where a trace starts and where it ends. Check all traces that have any signs of corrosion or damage. Also, looks can deceive. I once repaired a board that was badly corroded with a lot of bad looking traces. Using my multimeter I only found two traces that were not working, but the strange thing was that the two bad ones did not look nearly as bad as some of the other traces that still worked.

It might also be a good idea to reflow the solder under the SMDs.

Another thing you can do is start up the system with no BIOS installed. The system of course will have no chance of working without the BIOS, but it should display a predictable pattern that you can compare with a known working MV-1 without a BIOS.

For now I would just remove the board with the cartridge connectors and try to get the crosshatch screen and do the self tests and worry about inserting a cartridge later.

Thanks for all the help man, greatly appreciate it!

So far every corroded trace I've tested on the big LSPC, that one ram chip closest to the battery and the traces to small QFP located behind the battery, are all making proper connectivity... Either that or I'm doing these checks wrong. This is very puzzling...

Never thought of reflowing the solder, I'd only do that as a last resort.

I tried starting the system up without the BIOS, no change. I cleaned the socket with a tooth brush and 99% alcohol (all I have), doesn't seem to have helped. Though to be fair, that socket was rather dirty.

Yes you're right, removing the cart slot riser card is a good idea, no need to boot a game for now. Crosshatch is more important. Besides, without the cartridge guide and all that, it's easier to work with the board.

On another note, I tried running audio from the board to the TV and turned the speaker slider up. There's a clicking sound... Click click click click click... That can't be good.

 

[mjs256]

That clicking is just the watchdog circuit - there is an automatic counter which restarts the system if the counter reaches it's target. When the system is working properly it is constantly resetting the counter to prevent the system from restarting. If the system crashes (or is not working properly) then there is nothing to stop the counter so it forces the system to restart. This was really meant as a way for the arcade hardware to automatically recover from a system crash instead of just hanging indefinitely or showing a "Blue Screen of Death" like so many Windows users have come to know and hate. The clicking you are hearing is just the system restarting over and over again because of this "watchdog" circuit.

Since it made no difference when you removed the BIOS then you know the system is probably not even getting far enough through the startup sequence to get any data or instructions from the BIOS, so that helps to narrow it down some at least.

Unfortunately, your problem may be within one or more of the ICs. If that is the case then the IC would need to be replaced. In order to diagnose which IC is bad you would probably need a logic probe or an oscilliscope.

For now I would just keep checking for bad traces, but if you do not find anything then you may not want to keep on going because it will be very hard to find the problem especially without the right tools. Unless you are up for a challenge then you might consider selling it to someone who might want it for parts.