Need some help with Philips Scope

[david77]

Hi guys.

I've got an old analog Philips PM3212 scope that was sitting unused under the bench for years, it has some problems which I now decided to fix. The biggest problem were dirty switch contacts. I've cleaned the switches and they now seem to work fine.

There's still another problem and I'm not sure where to start looking.
The scope starts clipping waveforms if the input amplitude gets too high.
Here I'm measuring a 10kHz sinewave of 0,5Vpp vertical ampl. is set to 50mV/div and everything
is still fine.



Now if I switch to 20mV/div the clipping starts:



The funny thing is this is happening on both channels. So I somehow don't think the Y amplifier
is to blame.
I do have the servie manual for the scope but don't really know where to start looking.
Any suggestions would be much appreciated.

[amspire]

It is odd behavior.

If you put a 0.5 Vp/p into a 20mv/dev input, you expect it to clip, but in both photos, the waveforms are too small in amplitude for the channel sensitivity settings..

The input channels will have a vernier control that should be turned to the CAL position. Is this correct?

Perhaps you have them turned fully in the opposite direction.

Richard.

[Psi]

Yeah, first pic is showing a problem also, a 0.5V p2p waveform should take up 10 vertical divisions at 50mV/div.
First thing I'd do is check all the power rails are correct, especially for the analog front ends.

Does moving the entire waveform up/down effect the clipping at all?

Can you get a waveform to take up the entire screen on any setting?

[alm]

My guess is that the 1 div amplitude at 50mV/div is because David uses a 10x probe. Does the vertical position control have any effect? Does the clipping point remain on the same place or move with the signal? This may tell you if it's in the front-end or in the amplifier driving the CRT.

I'm not familiar with the service manual of Philips scopes, does it have detailed troubleshooting instructions like for Tek scopes, or just schematics and parts lists? Tek has a document about troubleshooting your oscilloscope, this may be contain some helpful hints.

[david77]

Yes, the first photo is also not correct. I'm not using a 10x probe, the generator is connected directly to the Y input. I missed that  .

The vernier controls are in the CAL position.

And no, I can not get a full screen display. Also if I change the Y position the waveform clips on the
positive or negative side if I move it up or down.
The first pic shows Y position turned fully up, the second fully down.
If I put the input to zero I also can't move the trace over the whole screen. It only moves about
1 div up and 1 div down.

That would point to a faulty power rail I'd guess. Probably shorted caps or something.

[tekfan]

My guess is the Y output amplifier going to the deflection plates. Check the voltages around that section. Check for overheated resistors which may have a higher resistance than normal thus reducing the deflection range.

[amspire]

Tekfan,

I think you are probably right. The waveforms look good, so I suspect the channel circuits are probably OK.

Something is reducing the Y amplifier gain to about 30% - like a failing resistor or a dry joint.

It looks like DC is affected too, so it is not a capacitor problem, and my suspicion is that the power supply rails will be OK.

So somewhere in the Y amplifier circuit where the gain can be affected by a single component.

Richard.

[david77]

You two are right. I have traced the error to the Y amp  but as yet have not been able
to identify the problem.
I have been measuring for the last 3 hours and my prime suspect at the moment is a transistor
array CA3086.
The voltages to the Y plates vary between 20,2 and 25,3V depending on the Y position pot's
setting.

I have to admit I have not quite understood how that circuit works, maybe someone would be
so good as to illuminate me a bit?

Attached is part of the Y amp where I think the error lies. I put my measurments in there in red.
The resistors seem to be fine, I've measured every single one.

[amspire]

Pin 3 of the CA3086 is open circuit somehow - probably a dry joint. Otherwise it has gone open circuit in the chip itself.

But if you resolder pin 3, everything will probably work. If you can measure a diode-type drop using the diode test on your multimeter between pins 2 and 3 on the CA3086 itself, then definitely a dry joint or broken track to pin 3.

If there is no base-emitter diode drop, then your CA3086 has an open circuit internally. It was a widely used IC, so it should be readily available.

Really great work testing out the circuit.  Makes it very easy to diagnose.

Richard

[amspire]

It just struck me that I assumed you had measured -12v on R843, but maybe you didn't.  You would get the same symptoms if R843 is open circuit, or if there is an open circuit to either the -12V, or pin 3 of the CA3086.

The last two transistors of CA3086 should have been providing a gain of about -10, but instead they are doing nothing, and so effectively the output transistors are being driven directly by the first two CA3086 transistors.  One of the consequences is that the signal you are seeing on the CRO is not only low in amplitude, but it is inverted.

When you fix it, you will have to get over the fact that things will work in reverse.

Richard

[david77]

I've already resoldered this whole section of the PCB - doesn't do the trick I'm afraid.
So I'll have to wait till Tuesday, I hope we still have some CA3086's at work.

Thanks for the help so far  .

[amspire]

Before you replace the CA3086, simple diode and continuity checks can find exactly where the problem is. Somewhere between 012v and the transistors connected to pin 3, there is an open circuit.

If you can detect the base-emitter junction between pins 2 and 3, and between pins 6 and 3 on the IC package, then the CA3086 is fine.

In that case it comes down to either one resistor, dry joints, or a broken track.

If a dry joint has corroded in time, resoldering is sometimes not enough - you have to remove the part and scrape the oxide off before resoldering.

Richard

[david77]

Just tested the CA3086, at least two transistors in there are dead.
Have now removed it and hope to get replacement soon.

[david77]

So, after replacing the CA3086 with a new one (Harris, Datecode Week 22/1984!) and cleaning up some very messy soldering left in there by a previous owner the scope is back in service!

Thank you everyone for your help.

[Psi]

NICE!

It's always a great feeling when a fix works and brings something back to life

[amspire]

Waveforms look great now.

Fantastic!

Richard

[tekfan]

So, after replacing the CA3086 with a new one (Harris, Datecode Week 22/1984!)

Nice one! Sharp traces and the lovely graticule illumination.

What's up with the resistor next to the blue electrolytic cap?

[david77]

I don't know. That part of the circuit is the triggering and somebody has run riot through there with what seems to have been a 100W plumbers iron. It was really messy in places.
I cleaned up as good as possible and replaced that resistor. It wasn't actually broken just lost some of its enamel coating.
I got that scope out of the estate of a deceased EE with some other stuff and he was apparently a very messy solderer.