Gould 4074 display problem

[Noopy2014]

Looks good!
I'm glad I could help.

There is a single jumper connecting the battery with the rest of the circuit. It's placed in the bottom left "corner" of the battery. Is this jumper missing?

[mole42uk]

Originally Posted by Noopy2014

Looks good!
I'm glad I could help.

There is a single jumper connecting the battery with the rest of the circuit. It's placed in the bottom left "corner" of the battery. Is this jumper missing?

The jumper is there, but there's also some corrosion from an old, leaked, battery. I will have to troubleshoot that sometime.

Right now I'm still chasing the main fault which is that I can't get any reliable traces from any of the channels. I suspect that a power fault that increased the +12V to near 17V has damaged something but it's being elusive. I need a checklist - I have started on the calibration list in the service manual and so far got all the supply voltages correct. The Master Clock pll isn't adjustable to quite the correct voltage, at best it's about 200mV too high. I suppose if I go through and get everything correct, I'll find the actual fault....

[mole42uk]

Originally Posted by dave cox

Does this scope use a single ADC, with multiplexing for the channels ?

Yes, it's a single ADC with multiplex.

Probing around there discovers that sensible signals are present on only one channel before the ADC. Looking at the inputs to the CCD cards reveals that all four signals from the pre-amp board are there - three are correct, one is a little awry but nonetheless it seems that I have problems with both CCD cards.

I must put the spare CCD into a slot and check that.

[mole42uk]

Updates:

I have tested all three DAC boards and the ADC board this morning and the results are promising. All ADC inputs are good, applying a known signal to each of the four inputs produces a good signal on the 'scope screen.

Applying a known sine wave to the front panel inputs and checking the outputs of the pre-amp board, channels 1 to 3 are good, channel four has a low signal output with a distorted waveform.

Applying a known good signal to the inputs of the CCD boards, one board has no outputs at all, another has one good output, the third has two good outputs. All the working outputs are noisy, each trace has hash even though the trace seems to be close to expectation in terms of wave shape and height.

There are several irritating but less severe problems, such as one channel doesn't respond to the up/down position control and there's some triggering issues to resolve.

[chriswood1900]

Richard
When you say 'All the working outputs are noisy, each trace has hash even though the trace seems to be close to expectation in terms of wave shape and height.'
Could it be the norm, digital scope traces often look noisy compared with analog although they are working correctly.

[mole42uk]

Originally Posted by chriswood1900

Richard
When you say 'All the working outputs are noisy, each trace has hash even though the trace seems to be close to expectation in terms of wave shape and height.'
Could it be the norm, digital scope traces often look noisy compared with analog although they are working correctly.

I rather hope not Chris, it's pretty awful noise and I'm sure I've seen photos of the 4074 screen with much cleaner traces. I suspect damage from the 12V excursion to un-regulated 17V.

[mole42uk]

Here are two photos, the better quality image comes from Noopy2014, the other is mine:

[Noopy2014]

Be aware that the CCD signals need a lot of correction due to the imperfections of the CCD. The correction data has to be generated while self calibration. Have you run a full calibration (menu, 999...) each time you changed the CCD boards?
It's also possible that one of the correction circuits is dead...

[mole42uk]

Okay, I did the full calibration twice. The good news is that if I apply a signal from the cal output and press the Auto Setup, during the time that the machine tries to resolve a displayable signal it briefly shows a very sharp, well-defined, noise-free square wave trace.
Downside is that after the Auto Setup routine has completed the display resolves to a random set of dots.
It's a fabulous project for the duration of the lockdown!

[Noopy2014]

You really have a very interesting project for the lockdown!

What do you mean with random set of dots? Is everything messy or only the traces? You are not in ETS-mode (<250ns/div)?
Strange...

I think you will have to go all the way trough the oscilloscope. Input stage is 3/4 good, I think that is good enough for the Moment. Now you have to dig through the big complex CCD-board…

[mole42uk]

Yes, it's only the traces that are messed up, the alpha is good, sharp and stable.

I have seen a couple of photos of the oscilloscope displaying four different wave patterns - square, triangle, sine, etc, but there is nothing connected to the input sockets. Is there a hidden menu to get a 'demo' mode active?

[Noopy2014]

That's no demo mode. Beside one-time-trigger you can save traces (up to 8 I think) and activate them whenever you want and in addition to the traces you measure at the moment. At the bottom of the screen you have the a save- and a recall-button.

[mole42uk]

Time to revisit this beast. I have to put it under the bench and ignore it every few days otherwise I might use it to fill in a hole in the ground.....

I can now observe a fairly faithful representation of the CAL output signal (1v p-p, 1kHz square wave) on channels 1,2 & 4 but the trace flicks up and down by about 1½ divisions, irrespective of the V/div setting at a rate of about half a second. It happens both normal and inverted, AC / GND / DC the same and all the working channels are affected. The alphanumerics remain stable and don't jump up and down.

I tried a replacement Display board but that does the same. I have been looking at the block diagram to try and establish a common path, it looks as though it must be the ADC board. I have a spare so I'll try that next.

[Noopy2014]

So you have a good trace to follow.

I additionally would try this: Save a curve and display it on the screen. If it looks stable you definitely have no problem with the display CRT amplifier and so on...

Next I would check the signal coming out of the CCD board to be sure that the CCDs work properly. But since most hannels have the problem the CCDs are probably ok.

Perhaps there is a half dead bit in the digital bus between ADC and display. Dead logic chip...

[mole42uk]

Okay, I didn't save the display. I did change to No.2 ADC board and the problem is not there any more. So I know there is a problem with the No.1 ADC board - I imagine it's a problem that started when the 12V line went to 17V....

I am now back to checking the CCD boards.