Advanced troubleshooting

[Ikaria]

Having trouble figuring out why this one isn't working. I'm getting power into the main board, but there's no power making it out to the features.

The only things I can see are what looks like leakage from the crystal onto the 2 resistors, and leakage from either the bottom of the diode, or the inductor.

I'm going out to buy a digital multimeter with diode test tomorrow, but wanted some thoughts from those in the know.

[Ikaria]

Was able to test the diode and it tests @ .152 forward, open reverse. Not sure if this is a high enough reading or not. But from this I'm going to say the diode is probably good.

Anybody out there have more experience with these things that could give me a better direction to possibly fix this, I'd appreciate it.

[daverob]

Really need a bit more information to go on before I would be comfortable giving advice on troubleshooting a power problem. I'm not going to start saying to randomly poke around with a multimeter, unless I know it's going to be reasonably safe to do so.

For all I know this might be a board from a machine with a CRT monitor in it, and it could have several thousand volts on it. If I just asked you to measure voltages on it with the power on, I might never hear back with the results from your testing.

Really need to be able to see pictures of the power board in question (both sides would help), and also to be able to identify the machine, as different era machines use different electronics. Then it might be possible to think of a suitable direction to point you in...


The good thing is that your pictures only show the glue that was often used to hold large components onto PCBs, over time it goes brown and brittle, and can often be mistaken for leakage, but it's nothing to worry about. It can sometimes cause problems in high frequency circuits as it ages, but it's rarely used in this area, so isn't usually a problem.

[Ikaria]

Thanks for posting Dave. This machine is Pachinko Big Awards 13 hanemono. Identical to the one seen in trailtraumas post. http://www.pachitalk.com/forums/mode...o-machine.html

I'm sure it's the board because the 'tattletale' led (d6) is not lit when the power is on.

Hopefully the pics are good.

[daverob]

OK, to me it looks like the red LED is driven by the transistor below it, so will probably require the CPU on the board to be working before it will light. So it's not absolutely certain that this is a power supply issue.

If it's possible to re-mount the board to the machine but leave the covers off, it would be a good idea to measure some voltages on the board, to see which voltages are there and which are not. Assuming that this machine is fed from a 24v transformer, there shouldn't be any dangerous voltages on the board, so as long as you're careful that your meter leads don't short anything out, this should be perfectly safe to do.

With the power on, if you measure the voltage across C14 (to the left and slightly above the LED in your photo). Meter set to 200v DC range with the red lead on the left and the black lead on the right, you should get a reading of between 28v and 36v (actual value doesn't really matter as long as it's somewhere near these values).

This will prove that there's power getting to the board, and that the power conversion circuitry is also getting power.

Secondly make the same measurement on C43 (just to the right of the orange inductor). This time the meter is set to 20v DC range and the black lead goes to the left hand side of the capacitor. You should read 5v on this capacitor (give or take half a volt).

If this voltage is missing, then the power circuit is not working. If it is correct, then the problem is with the CPU part of the circuit board.

If either voltage is missing, try powering up with all of the connectors on the right hand side of the board disconnected, to see if the voltages return.

If it's a power problem, then I'm going to need to know the markings on the 8 pin chip just below the inductor, and some clearer photos of the left hand side of the board (both sides) so I can trace out the connections before I can give any more advice on where to look next (might be an idea to email me the photos as well as posting them here, as the resolution of posted photos isn't really high enough to do this easily).

If both voltages are correct, then fault finding is going to be a lot more complicated, and is likely to need a lot more equipment than a basic multimeter.

[Ikaria]

I appreciate you taking time away from your own projects to help with this.

The readings on C14 = 34.4v, and C43 = 4.86v

Looks like it's going to be complicated...

[new in town]

Me THINKX now why you needed the Shunt pins sent over.

Did you get them O.K.?

[Ikaria]

Me THINKX now why you needed the Shunt pins sent over.

Did you get them O.K.?

You hit the nail on the head with that one! I did get them, Thanks a ton!!

[new in town]

Your Welcome Sir. . .

Good Luck with the Repairs.

But Remember. If you need H E L P

Spoiler

I know C P R

[daverob]

The readings on C14 = 34.4v, and C43 = 4.86v

Looks like it's going to be complicated...

This does have the potential of being a complicated one, as the voltages are exactly what I'd expect them to be, so the PSU part seems to be working fine.

The only simple thing that is left to try is reseating the EEPROM chip in it's socket (the M2764 chip with the label with the name of the game in Japanese written on it). It's a very long shot, but it's possible that it might just be a bad connection between the chip and it's socket.


If this has no effect the next step will require checking for signals on the circuit board and will need either a 'logic probe' (some multimeters have this function on them, so it's worth checking the manual for the one you've bought) or an oscilloscope. You might be able to get away with using the 'frequency' setting on a multimeter (if you have this setting) for some of the tests, but the results are harder to interpret.

If your multimeter doesn't have a 'logic' function on it, then it's probably not worth investing in buying more test equipment, as you'd probably be (financially) better off trying to source a replacement board or sending the board out to someone who has the necessary equipment (and experience!) to test and fault find on the board.

[Ikaria]

Did the reseat on the EEPROM. No change.

Manual that came with the meter is pretty bare. So I included a photo.

If there's anything I can test with what I have, let me know.

[daverob]

That meter is pretty basic, it does have a frequency range, but it doesn't measure high enough frequencies to be of any use for the sort of tests that are needed here.

The only other thing I can think of is to ask for other owners of similar (but working!) machines, if they can remove all of the connectors from the board (on the right hand side of your photo) and see if the LED lights with them disconnected. If theirs does (and yours does not) then we know that there is definitely a fault on your board.

If their LED also doesn't light, then there is a remote possibility that the CPU isn't starting due to a problem with something other than the circuit board.

[TailTrauma]

The only other thing I can think of is to ask for other owners of similar (but working!) machines, if they can remove all of the connectors from the board (on the right hand side of your photo) and see if the LED lights with them disconnected. If theirs does (and yours does not) then we know that there is definitely a fault on your board.

I'm willing to help. I own an identical working machine (although I can't be sure the board is 100% identical... the enclosure is a sealed metal box and I don't want to break the seals.) I can see the LED's light from outside the enclosure and it does light when the machine is operating normally.

I'm cautious, though. Can you assure me that I will not damage my board by performing the test you suggest?

[monkeyboy]

Speaking as an electronics engineer there is no need to worry, you won't damage the board by having the connectors disconnected.

[daverob]

The board will be fine if you run it with the connectors disconnected. But to be extra safe I'd recommend only powering it on for the few seconds needed to confirm whether the LED it lit or not.

Though you do need to make sure that all connectors are removed (apart from the cable that supplies power to the board!), as if you leave something like a motor connected, but remove the connector to a switch that monitors when the motor has reached it's limit, it is possible that the board may keep driving the motor until it burns out.

Also it's important to make sure you replace the connectors exactly how they were originally, as faults could be introduced if they are replaced incorrectly.

[TailTrauma]

Speaking as an electronics engineer there is no need to worry, you won't damage the board by having the connectors disconnected.

The board will be fine if you run it with the connectors disconnected. But to be extra safe I'd recommend only powering it on for the few seconds needed to confirm whether the LED it lit or not.

Though you do need to make sure that all connectors are removed (apart from the cable that supplies power to the board!), as if you leave something like a motor connected, but remove the connector to a switch that monitors when the motor has reached it's limit, it is possible that the board may keep driving the motor until it burns out.

Also it's important to make sure you replace the connectors exactly how they were originally, as faults could be introduced if they are replaced incorrectly.

Thank you both! I am reassured.

I will perform the test and post the results this afternoon as soon as possible, probably this afternoon.

[TailTrauma]

The LED on my board lights when power is applied, even when there are no connections other than power.

The test I performed:
-Disconnected all cables from the bottom of the board. The only connection was the power cable.
-Powered on and checked onboard LED. (It lit.)

[Ikaria]

TailTrauma, Thanks for checking your board out.

Looks like this machine is going to be retired for now.

Time to move on to the next one...