Common moding technique on a Hameg HM203-7

[cosenmarco]

Hi everybody,
I'm troubleshooting a Hameg HM203-7 oscilloscope (see https://www.eevblog.com/forum/testgear/hm203-7-service-manual/ for the manual) with a problem in the vertical stage.
The problem shows up in the following way: the trace is either
a) heavily blurred especially going far away from the center and presents a "pillow" distorsion inwards
b) or jumps upwards randomly at higher intensity.
Timebase seems to work ok and also it triggers correctly.
In X-Y mode you can see the the dot jumping upwards similarly to the trace.

I checked power rails and they seem fine. For High Voltage I checked the potential across R3015 and reads ca. 480V and the trace is actually showing up but distorted, so I'm ruling out (for now, there might be a bad cap somewhere though) the HV stuff.

I checked also the voltages in the final Y stage and they seem quite close to what reported in the manual. BTW I think there is an error on the voltage reported at the node connecting R2038 and R2039. I am pretty sure it should be +5.88V instead of -5.88V.

So I read in the TEK troubleshooting PDF (posted in https://www.eevblog.com/forum/repair/repair-documents-and-links-sticky-me-please-mods/) about "common moding" technique and I'd like to use it to isolate the bad stage in the vertical final stage. So this should be similar to shorting the signal path to ground and the input to isolate the noisy stage in non-differential amplifiers. There is not much though in the TEK PDF about this technique and I'm wondering how should this be execute properly (at least in my circuit) and what can go wrong.

Let's assume I short the collectors of T206 and T207. So the way I understand it is that the differential part of the signal should not pose a problem when the output of a differential stage gets shorted because the current produced by one of the devices should be equal to the current absorbed by the other device.

Though if there is a heavy common mode signal could it happen that this procedure toasts the output devices which I'm shorting? In other words what's a safe way of doing this? Should I rather short the bases of the transistors in the next stage (in my case the bases of T208 and T209) instead of the collectors of the previous stage?

I also was thinking to put the multimeter in mA mode instead of abruptly shorting. At least it's fused and I can see if there are concerning currents going through the short. Does it make sense?

Thanks in advance for any reply

[oldway]

Looks like a defective CRT.... (broken internal connection) 

[cosenmarco]

Hi oldway,
how would you test if the CRT is defective?

The scope used to work the day I bought it (although with some distorsion of the trace) and the day after started behaving funky.
I can get a sharp trace with very low intensity or a very sharp dot in X-Y mode before the trace goes nuts, so I tend to exclude the CRT.

Thanks anyway 

[cosenmarco]

I suppose, as a simple way of testing, since the issue is "directional" (the trace jumps upwards) I suppose I could exclude that the CRT is defective by swapping the vertical connections to the CRT and observe that the trace jumps then downwards... Any other ideas?
I have no fancy equipment. As a matter of fact I cannot for example feed +80V  in the Y deflection plates to test disconnecting the CRT from the Vertical because I have no bench power supply.
Is it possible (desirable?) to test the CRT with the Vertical plates completely disconnected from the final stage? Would that result in damaging the CRT ? Should then the plates be grounded ? Can they be maybe connected in parallel to the X plates and observe a diagonal line forming on the screen?
In short how do you test a CRT with almost no equipment ?

Also I spent a couple of thoughts on common moding (which BTW is what this thread is also about) and I think it should be fine because the differential component of the current would flow between the two transistors of the same stage and the common mode component of the current would flow into the next stage. Anybody can confirm the correctness of my thinking?

[oldway]

In short how do you test a CRT with almost no equipment ?

You're trying to square the circle.
You need another oscilloscope to repair a faulty one. But sadly, focus changing with the position of the beam, that seems a bad CRT....A broken internal weld gives the kind of problem you described.

Best way to test the CRT without any instrument is to replace it by a known well working one.

These oscilloscopes have always problems with cold solders of X/Y end amplifiers....you must resolder all the connections of BF458, BF472 and of the power resistors.

Look also for bad solders of ground connections.

[cosenmarco]

Best way to test the CRT without any instrument is to replace it by a known well working one.

That's also not an option, as you might have guessed.

These oscilloscopes have always problems with cold solders of X/Y end amplifiers....you must resolder all the connections of BF458, BF472 and of the power resistors.

I see actually some solder joints in the final Y stage which seem not factory original (quite grossly made) so I assume somebody re-worked it or replaced some parts in that area already.

I did the test about swapping the Y plates and with my surprise the trace was still sparking upwards (it actually seems it does small little sparks upwards and then comes back) and it does more frequently when I increase intensity as usual.
I may actually post a short video of the issue if somebody feels it might help.
Now I'm open for everything but I still hope for a usable CRT as basically that makes the scope and I would have to re-sell it for scrap if the CRT is bad.
After all there's a bunch of grids in the CRT so I'll investigate in the area of focus and astigmatism. Maybe I get lucky and find some bad solder joint or some broken capacitor...

[cosenmarco]

Update: I was able to score in eBay a (presumably) compatible CRT, so I will post about replacing the CRT when I come back from vacation.

Another question: I was observing better the HV area and the blanking / un-blanking circuit might be also a good candidate for injecting noise / distorsion / insufficient current in the cathode. Could it be the cause of the problem? That's a bit harder to troubleshoot due to HV involved but the blanking signal is at low voltage and so also the circuit part before the optoisolator.
Any suggestion about troubleshooting that area? Eg. what about forcing the blanking signal always ON so that I can see if the problem is coming from the blanking circuit or the part after the optoisolator?

The optoisolator itself is mounted on socket and I checked it with the DMM ad seems good.

[cosenmarco]

Ok, I came back from vacation and I found the new CRT in my office, I replaced with the old one and the scope is now working OK. All functions seem to work and the traces are clear and with good intensity and focus.
Many thanks oldway for suggesting to replace the CRT.   

Now I'm wondering if there's a way to fix the old CRT...
nah: I'll get rid of it.

[oldway]

Well done 
Enjoy your oscilloscpe.
My HM605 is my dayly workhorse for repairing vintage amplifiers and receivers. It is a very good analog oscilloscope.