• Please review our updated Terms and Rules here

Circuit Board Troubleshooting

Chuckster_in_Jax

Veteran Member
Joined
Feb 28, 2006
Messages
1,362
Location
Jacksonville, Florida
I am trying to revive a MAI S-10 Basic Four computer. The unit appeared to be dead. There was no floppy access on boot-up and no video. The neck of the CRT does not glow at all either. I pulled the power supply board, replaced all the electrolytic capacitors and it is putting out the right voltages. However the unit is still inoperative.
This unit has two main boards: a terminal board and a CPU/Floppy board. I pulled both boards and tested for shorts with a DVM and found the following:

Terminal board
+5V to ground has 285 ohms resistance
+12V to ground has >40 megohms resistance (should be OK).

CPU/Floppy board
+5V to ground has 45 ohms resistance

My first instinct is that the problem is being caused by a shorted tantalum. It is next to impossible to pinpoint the short so I believe I am going to have to start replacing all of the tantalums until I get a good resistance reading on my meter.
Any other advice on finding shorts on circuit boards would be appreciated.
 
Tantalums fail by short circuit, so you should be able to tell if they have failed (and the voltage was no enough to pop them) by checking for a direct short. 285 Ohms in not a direct short.

Have you looked for fuses (pico fuses for example) to see if something had popped?
 
Run it till sum'n goes *POP*, and observe where the smoke came from?

--T

When they pop they really pop (and have potential to take an eye out if you are bending over the board). They don't always pop though. On a drive I recently fixed, it was a faulty tantalum which looked normal. A test revealed a short circuit though. No resistence at all.

Tez
 
and it is putting out the right voltages.
If the PSU is outputting the correct voltages with the terminal boards and CPU/floppy boards connected, then there can't be a short (or other cause of PSU overload), so please confirm that your statement applies only when the boards are disconnected from the PSU.

If so (i.e. one of the boards is overloading the PSU), the first step would be to remove each board in turn to identify which board is the cause of the overload.
 
When they pop they really pop (and have potential to take an eye out if you are bending over the board). They don't always pop though. On a drive I recently fixed, it was a faulty tantalum which looked normal. A test revealed a short circuit though. No resistence at all.

Tez

Yeah, I realize tants can go shorted sometimes, but there's a pretty good chance of detecting that with an ohmmeter. I just like the pyro display.

--T
 
I am trying to revive a MAI S-10 Basic Four computer. The unit appeared to be dead. There was no floppy access on boot-up and no video. The neck of the CRT does not glow at all either. I pulled the power supply board, replaced all the electrolytic capacitors and it is putting out the right voltages. However the unit is still inoperative.
This unit has two main boards: a terminal board and a CPU/Floppy board. I pulled both boards and tested for shorts with a DVM and found the following:

Terminal board
+5V to ground has 285 ohms resistance
+12V to ground has >40 megohms resistance (should be OK).

CPU/Floppy board
+5V to ground has 45 ohms resistance

My first instinct is that the problem is being caused by a shorted tantalum. It is next to impossible to pinpoint the short so I believe I am going to have to start replacing all of the tantalums until I get a good resistance reading on my meter.
Any other advice on finding shorts on circuit boards would be appreciated.


Hi
What makes you think these are bad reading? What made you replace
the power supply Caps?
Most CRTs use 12V for the filaments. The open there is a bad indication.
Have you measured the filament at the CRT pins?
Dwight
 
I guess I'd look to see what the PSU output readings are when connected to the CPU board. While it may be the case that you do have a bad capacitor somewhere, wouldn't it be better to take a more methodical approach?

Are those resistance readings the same in both directions? (i.e. swap the probes) -5 into 45 ohms is a twitch over 100 ma, which isn't unreasonable. I assume that you're using the low-voltage resistance range on your ohmmeter and not measuring the forward resistance of some diode.
 
I guess I'd look to see what the PSU output readings are when connected to the CPU board. While it may be the case that you do have a bad capacitor somewhere, wouldn't it be better to take a more methodical approach?.

OK, I checked the supply voltage at several TTL chips on both the terminal board and the CPU/Floppy board and got 5.1V. 12V supply is fine also. I guess my inexperience in troubleshooting on this level led me to believe there was a short somewhere. I did find a tantalum that was discolored and had a pinhole in the top of it. I removed it until I can replace it.

Are those resistance readings the same in both directions? (i.e. swap the probes) -5 into 45 ohms is a twitch over 100 ma, which isn't unreasonable. I assume that you're using the low-voltage resistance range on your ohmmeter and not measuring the forward resistance of some diode.

Yes, readings are for both directions and I am using the low range on my meter. Should have done some basic circuit analysis, I= E/R, and come to the conclusion that this wasn't a short.
 
Hi
What makes you think these are bad reading? What made you replace
the power supply Caps?
Most CRTs use 12V for the filaments. The open there is a bad indication.
Have you measured the filament at the CRT pins?
Dwight

I was thinking that a resistance that low would cause a large current draw on the power supply. Didn't realize that it would only draw 100ma. I did notice there is heat coming off of the terminal board. The chips are getting warm after a short period of time. I don't know what is considered to be normal operating temp and what is considered too hot. I know that TTL chips draw more current than CMOS chips. Recently I was working with a Tandy computer with an Arcnet board in it. The Arcnet board got so hot I thought it was burning up. The heat given off was equivalent to that given off by a halogen lamp. Then I noticed it was densely populated with TTL chips.

The reason I replaced the power supply electrolytics was I felt it would be prudent since they tend to fail with age.

I haven't taken any readings at the CRT yet. I will try that next. The high voltage section appears to be down too. No static charge on the front of the tube.
 
The next thing on my list would be to see what's wiggling.

Got a logic probe or an oscilloscope? Check for the presence of clock signals and see if the address and data lines on the Z80 are twitching. While a scope will give you a good picture of the quality, a logic probe with a pulse detector will tell you if there's any activity.

If you can find a video signal going to the CRT, is that wiggling also? How about horizontal and vertical sync?
 
I'm going to have to put this on hold until Tuesday. Got to work 12 hrs. tomorrow but I'll pick back up with your suggestions and get readings on the CRT pins and check for activity on the boards with a logic probe. I do have an oscilloscope.
 
Hi
What makes you think these are bad reading? What made you replace
the power supply Caps?
Most CRTs use 12V for the filaments. The open there is a bad indication.
Have you measured the filament at the CRT pins?
Dwight

Problem with this is that the 12v power supply does not run the heater filament on the CRT. It's used for RS-232 communications and other small things.

Unless you are dealing with ancient transformer based things the TVs and monitors made in the past 30+ years derived their filament voltages from a low voltage tap off of the flyback.

RJ
 
Test results

Test results

OK, I've had a chance to run some checks and am posting the results. I wanted this to be a learning experience for me and possibly give other members an idea of what to look for when troubleshooting these systems.
This is a MAI Basic Four S/10 computer with the following symptoms:
Appears to be dead. Disk drives aren't being accessed and the screen is blank.
The computer has 5 boards in it:
Power Supply
Terminal Board
CPU/Floppy Controller board
Video Board
Main board similar to a backplane.

The power supply should be one of the first modules to be checked. I replaced the electrolytic capacitors in the power supply only because of their age and potential to leak. Then checked the power supply voltages without it connected to the system and they are OK. This supply outputs +5V, +12V and -12V. I then connected it to the system and checked for voltages on the Terminal and CPU/Floppy boards. They are reading OK.

Next I checked for voltages and signals going to the monitor section. There are 5 lines going to it:
+12V supply
Ground
Video signal
Horizontal Sync
Vertical Sync
The +12v is reading fine and the horizontal and vertical sync signals look good coming from the Terminal board. The video signal has a problem. When the computer is powered up I get the video signal for about 3 seconds and then it stops. I'm not sure what is going on here. I would think that the video signal should be a steady stream.
I noticed there was no glow from the CRT neck and decided to investigate to see what was causing that. I have a manual for this computer and it has schematics for all of the boards except for the Video board. It does have a description of the circuits and references to some of the components that are labeled on the board. However without a schematic and not much experience with video circuits, I had to do some studying of the video circuits from other computers and trace through this board with a multimeter to see where the signals were going. I came up with the following:

There are 6 pins on the neck of the CRT. If you are holding the socket that mounts on the CRT neck with the pin connectors facing you, the pinouts are:

Counterclockwise from the "key"
1 - Video signal
2 - H+ Heater
3 - Ground
4 - G1 Control Grid
5 - G2 Screen grid
6 - G4 Focus Grid

The only pin getting voltage is the Heater. (Dwight Elvey was right. 12v from power supply). I measured the resistance between the heater and ground pins and got 30 ohms resistance. However when powered up the neck does not glow. Do the grid voltages also have to be present to get that glow?

Most CRTs use 12V for the filaments. The open there is a bad indication.
Have you measured the filament at the CRT pins?
Dwight

Found that all of the grid voltages are derived from the secondary of the flyback. Tested the Horizontal Output Transistor (HOT) with the diode setting on my multimeter and found the Base -> Emitter is shorted. That explains the high voltage section being down and no video. I would like to replace the flyback at he same time I replace the HOT just in case it is shorted but I can't seem to find a replacement on the web. The part # is KFS-20492. Any leads would be appreciated.
One last item. I tested the pins on the flyback to see if there was a short somewhere. There are four connections to the primary and there is continuity between all four pins (0.2 ohms). The secondary has 3 pins and a ground. Resistance is 1.7 ohms between it and the other two, the others were 0.2 ohms. There is no continuity between the primary and secondary windings.
 
I assume that the heater resistance measurement was made with the CRT socket pulled off--i.e., you're not measuring the resistance of the supply. Then 12/30 = 400 ma--not out of the question. Sometimes the glow is very dim.

Run the thing for about 10 minutes, remove power and feel the CRT neck near the base. Does it feel warm? Then the heater is probably fine.
 
Replace the HOT and test it again... You may have a very hard time finding a flyback for it and the flyback may be OK. The only way to really tell is to ring it and look at the signals to see if they are really nasty looking with a scope.

Check the caps around the HOT and flyback, the diodes off the secondaries of the flyback, and the solder joints on the flyback itself as those can cause the HOT to pop. If you see any burnt resistors then that's a clue that you may have a shorted or leaky diode. I've seen many old monitors with leaking or shorted fast recovery diodes blow things like the vertical output IC and take out the HOT at the same time.

RJ
 
I am trying to revive a MAI S-10 Basic Four computer. The unit appeared to be dead. There was no floppy access on boot-up and no video. The neck of the CRT does not glow at all either. I pulled the power supply board, replaced all the electrolytic capacitors and it is putting out the right voltages. However the unit is still inoperative.
This unit has two main boards: a terminal board and a CPU/Floppy board. I pulled both boards and tested for shorts with a DVM and found the following:

Terminal board
+5V to ground has 285 ohms resistance
+12V to ground has >40 megohms resistance (should be OK).

CPU/Floppy board
+5V to ground has 45 ohms resistance

My first instinct is that the problem is being caused by a shorted tantalum. It is next to impossible to pinpoint the short so I believe I am going to have to start replacing all of the tantalums until I get a good resistance reading on my meter.
Any other advice on finding shorts on circuit boards would be appreciated.
hello,

mayby this can help you, if you still have the S10 with problems, i have a original MAI course available how to troubleshoot /repair the S10
 
Back
Top