Oscilloscope repair help

[stev.dk]

I have an old cheap oscilloscope that works perfectly in normal mode, but in digital storage mode, it adds noise to the signal - making it useless in storage mode 

My knowledge goes only as far, as not getting myself killed probing this thing I only have a multimeter to troubleshoot with. It is a Fluke 87 III, so i think the voltage readings should be pretty accurate.

Here im in normal mode, with no input. A nice straight line as expected.



Here im in digital storage mode, and as you can see there's some noise on the display. This might not look too bad actually, but send a sinewave through and its caos!



I read the service manual, and found something about y-amplification (page 27) - Probing the pins 7,8 on IC301 (LM733CN) reads 3.390V and 3.445V, which i guess is reasonable within spec. (3.5V). Pins 1 and 14 however is different. When i probe these in normal mode and storage mode, it adds extra noise. Is this expected? (check the attached ...mode-probe pictures) I have no idea if this is a clue to solve the problem, or if it's a no brainer, that probing these will add noise. However, the service manual says pins 1 and 14 should be +1V ±0.8V, and they both read just about -0.300V, only half a volt, but still 50% of its intended value of 1V.

Measuring the voltage on the test pins on the top board: +11V ± 0.5V (+10,65V), -11V ± 0.5V (-11,29V), +210V ±21V (+215,2V), +7V±0.4V - 0.6V (+6.84V), - 5V (-5.39V), so i guess the main power supply is working properly. And the Ramp test pin (not mentioned in the service manual?) reads +1.934V.

At this point im pretty much stuck. I have absolutely no clue where to go from here I really like this analog scope, and with the service manual and (as far as i can tell) only jelly bean components, i really have high hopes for a repair.

Please share your wisdom, i will be forever grateful

I uploaded the service manual here: http://stev.dk/gould_os1420_2x2mv,20mhz_oscilloscope_1982_sm.pdf

[84GKSIG]

i bought an old cheap oscilloscope that had been messed with before I had a chance and I found all the issues it was having were RF related and the cause was some one had removed various aluminum covers inside to have a play and never put them back, after making some new ones which was pain, pretty much perfect, but looks like possible rf interference from the storage circuit or possibly a coupling issue.

step one inspect for possible missing RF shielding
step two clean all contacts and switches
step three insure a good strong secure ground between each circuit ( assuming its not a single board construction )

 

[oldway]

Use it in analog mode only.

[nfmax]

Can you post pictures of the display with the exact same external sinewave input in both analogue and storage modes, please? It looks like the issue might be bad, stuck, or shorted bits in the digital path, hopefully between the memory & the display DAC.

[jancumps]

Ask a fellow EEVblogger from your region to help

That's what I did when I was fixing my oscilloscope. I asked andete from the forum here (he lives two villages away from me) if he could come over with his scope, so we could trace the signals.
We managed to trace the issue, and that scope is fixed.

[stev.dk]

Use it in analog mode only.

OMG, Einstein? Is that really you?

Can you post pictures of the display with the exact same external sinewave input in both analogue and storage modes, please? It looks like the issue might be bad, stuck, or shorted bits in the digital path, hopefully between the memory & the display DAC.

Sinewave is CH1, Sinewave-x10 is CH2 in x10 mag mode (dont know if this makes any help).

step one inspect for possible missing RF shielding
step two clean all contacts and switches
step three insure a good strong secure ground between each circuit ( assuming its not a single board construction )

 

I dont think any shields are missing, ill try the other steps later.

[robrenz]

None of your pictures show anything connected to the input channels bnc connectors. Whats up with that?

[stev.dk]

I just wanted to show that the oscilloscope have a fault on the y axis. I thought using no signal would be fine. Check my latest post though, im sending in a sinewave here.

[Anks]

Looks like some of the bits on the ADC are stuck or something on the bus is curupting the data

[oldway]

Use it in analog mode only.

OMG, Einstein? Is that really you?

You're wasting your time trying to repair an old digital oscilloscope that is worthless.
Digital technology has evolved considerably since.
Already an analog oscilloscope of just 20 Mhz, it's not interesting.
What use digital oscilloscope part of this?
it's a piece of shit
And again, these old digital technologies are very complicated to troubleshoot and it is difficult to find replacement components.
For shure, I am not EINSTEIN but I have some experience in repairing CRT scopes ! 

[84GKSIG]

i cant see harm in trying, worst case you may learn something from it

[oldway]

i cant see harm in trying, worst case you may learn something from it

But my recommendation "use it in analog mode only" makes sense !

[Shock]

But my recommendation "use it in analog mode only" makes sense !

We would have to rename the forum to "workaround".

[Cyberdragon]

Does the digital section use the same power section? It may just be some bad bypass or filter capacitors.

[ManateeMafia]

It seems to have the appearance of every other stored value is correct or it is just an illusion? What is the effect of using Dual mode?

[ornea]

It doesn't make sense if the OP is hoping to use the storage feature, which he does.

[voltz]

Hi,
It wont be easy to isolate the fault in a case like this. But here are some ideas if you want to keep going..
Check all low voltage power rails for noise. Stay away from anything over 50v.
You will most likely need another working scope to find the fault with this scope. Using a multimeter just isn't enough in a case like this. You will need to see noise and data activity, a meter cannot show that.
After checking the supplies (especially the digital +5v rail) and you are happy the low voltage rails are reasonably clean, expect a tiny bit of noise, below around 50mv +/-, then next:

Look at the digital section on page 63. Here there is a DAC, some RAM and another chip (on the far right) which i cannot read because of the very bad print. In my experience with old storage scopes like this, the most suspect component would be the RAM. In this case its marked as U817 which is a 6116 ram chip. A lot of these chips gave up working years ago. There is a good chance its that. But again, you would need to examine the address and data buss with a scope and look for 'stuck' pins, either high or low, while injecting your sinewave. This would indicate a fault on the buss somewhere.

Finding the input to the ADC (following the analog path to it) and checking for noise there too would also rule out anything bad feeding noise to the ADC. I suspects thats probably ok.
If the chips are on holders, then go pull out the RAM and see what effect that has. But if not, then its a tricky repair and narrowing the fault down will take time to locate. At that point, maybe ask whether its worth it or not.
Good luck.

[Stonent]

I would suspect bad ram chips. Look for any data lines always stuck high or low when data is going in. Same for address lines stuck. A stuck address line might indicate something else like a 3 to 8 chip may be faulty.  Look for any 74xx138, 137, 238, or 237 chips near the ram or main processor they might be difficult to test in circuit unless you can trigger on a latch and see if again there is one output consistently stuck high or low.

Edit: didn't catch that checking for high and low pins had already been suggested. Well hopefully my thoughts on 3 to 8 decoders might be useful though. If the memory is socketed you could swap them around to see if the bad capture changes.

[stev.dk]

Thank you guys! You have been very helpful.

Special thanks goes to oldway for calling my oscilloscope worthless and a piece of shit. That's really constructive.

The problem was on the bottom board, where one or more socketed chips had vibrated loose (at least i think so), causing one or more legs to be intermittent. After gently pressing these chips into their sockets, the scopes storage function works like a charm :-)

[tautech]

Thank you guys! You have been very helpful.

The problem was on the bottom board, where one or more socketed chips had vibrated loose (at least i think so), causing one or more legs to be intermittent. After gently pressing these chips into their sockets, the scopes storage function works like a charm :-)

Many old scopes with socketed IC's and transistors can give problems in this way.
The quality of sockets can be the cause but not always. In humid climates they can be problematic with corrosion at the points of contact.
Standard troubleshooting in these scopes is to re-seat components in the problem areas of the circuit.
PCB interconnects can also be troublesome.

You'll know where to look next time. 

[oldway]

Thank you guys! You have been very helpful.

Special thanks goes to oldway for calling my oscilloscope worthless and a piece of shit. That's really constructive.

The problem was on the bottom board, where one or more socketed chips had vibrated loose (at least i think so), causing one or more legs to be intermittent. After gently pressing these chips into their sockets, the scopes storage function works like a charm :-)

This is what I really think of these "combi-scopes" of older generation and there is no reason that prevents me from saying what i thing about it.

Why waste time on scopes like this?

Here you have been lucky, but you could spend days trying to fix it and do not get there.

Digital feature of these old scopes is unnecessary and totally outdated.
Nothing to do with the current digital oscilloscope.
Digital technology has improved considerably since then !

I worked in a big SMPS factory (about 400 workers) and we used combiscopes Philips / fluke PM3394A ... I did an investigation with other technicians: none was using the oscilloscope in digital mode!
And yet, there is no possible comparison between a 20 Mhz Gould and 200Mhz PM3394A !

[particleman]

Good job with the repair.  Don't worry about what some people say. You learned about socketed chips/components and now have a working scope. I just picked up a tek 485 with intermittent issues caused by a socketed transistor.  Have a good holiday season!

[voltz]

Good result, glad you found the problem. And simple.