PM3394 glitches with less than 2us per division

[BurnedResistor]

see the attached video!

It seems that as soon as the horizontal time base is faster than 2us per division, the signal jumps back and forth between the actual value and ground.

I am gonna hazard a guess that getting this old model repaired is going to be hard? Could this be something I can do myself?

If you need any other info please let me know!

https://youtu.be/Y-saiWZ_-xU

[David Hess]

See how it behaves when using more than one channel, the other channels, and when using the external trigger input.

How does it behave in analog mode?

Check the levels and noise on the low voltage power supply outputs.

[BurnedResistor]

See how it behaves when using more than one channel, the other channels, and when using the external trigger input.

How does it behave in analog mode?

Check the levels and noise on the low voltage power supply outputs.

1) happens on all channels, even if used alone / together.

2) does not happen in analog mode

3) passes selftest

[BurnedResistor]

I won't have time to Crack it open for a while

[David Hess]

I think you are going to need the service manual to track this down; luckily a good scan is available.

I thought maybe the problem was being caused by equivalent mode sampling however that is used at 200ns/div and faster and not 2us/div.

Why is the display showing 1.6us/div sometimes in your video?

It would be a good idea to check the low voltage power supply outputs for level and noise while the problem is happening.

[oldway]

Unfortunately, the design of this oscilloscope does not make measurements easy.
These are electronic plug-in boards and components and test points are inaccessible.
It would require extension cards to troubleshoot it.

But to my knowledge, these extension cards are not available on the market.

The only solutions are:
- Static verification of components, including electrolytic capacitors, with the cards out of the rack.
- Replacement one by one of card by card known as good working, at least to locate the faulty card.

[David Hess]

Unfortunately, the design of this oscilloscope does not make measurements easy.
These are electronic plug-in boards and components and test points are inaccessible.
It would require extension cards to troubleshoot it.

For something as simple as measuring the low voltage power supply outputs, thin and short coaxial extensions can be soldered in to act as test points.  Just a wire could be used but a coaxial connection to a low impedance power supply output will yield much lower noise.

I connect a standard x1 or x10 probe to the coaxial extension using a probe tip to BNC adapter.

[oldway]

No need to do this, only check (value and ESR)  the electrolytics capacitors of the power supply.

Service Manual :

http://elektrotanya.com/philips_pm3382_3384_3392_3394_oscilloscope_sm.pdf/download.html

[HighVoltage]

But to my knowledge, these extension cards are not available on the market.

That is true.
I have been trying to get one for years and I know the guy who has one. But he is not even showing the adapter to me.


 

[grumpydoc]

I managed to construct an adapter out of several stacked connectors.

There is a point where the digitiser uses both DACs interleaved in parallel to get 200Ms/s rather than 100Ms/s which is what the individual DACs are capable of  - it sounds like that might be the problem although, from memory, I thought that this did not kick in until a faster  timebase than 2us/div

[BurnedResistor]

Thanks you all for the great help.

A few notes:

1) This is my only oscilloscope, so I am not able to perform complex measurments on it.

2) I have never taken something with a crt apart and am honestly a little nervous about doing it. OFC I wont touch it while it is on, but still, any safety tips?

[David Hess]

1) This is my only oscilloscope, so I am not able to perform complex measurements on it.

The oscilloscope can measure the ripple and noise on its own low voltage power supply outputs.  Be careful about where the probe's ground lead connects.  Do not allow it to wave around; remove it, tie it down, or solidly connect it to something.

2) I have never taken something with a crt apart and am honestly a little nervous about doing it. OFC I wont touch it while it is on, but still, any safety tips?

The CRT high voltages are not that dangerous.  The distributed PDA (post deflection accelleration) resistance will discharge the high positive PDA voltage within minutes and the high negative cathode voltage will also discharge within minutes.  You should not need to access the CRT sections while it is on.

[oldway]

I will be sincere with you, I know well this series of oscilloscope and considering
- the complexity of this oscilloscope, and even more of the digital circuits
- your knowledge and experience
- the technical means at your disposal,

The chances that you have to repair this fault are practically nil.
And worse, you can cause even more serious failures.

So, I advise you to use this oscilloscope only in analog mode.

[David Hess]

I agree with oldway which is why I keep suggesting simple tests which might reveal the problem.  I am pretty sure I could diagnose and repair the problem after extensive study of the service manual but it would be a lot of work and even more work to lead someone through the process.  The more likely result would be further damage.