The 'scope of many faults

[throbscottle]

I thought I'd better get my oscilloscope fixed (Cossor CDU 150). Been rather like chasing a poltergeist.
So far I've fixed:
constant very bright display which was due a resistor gone high connecting the blanking/bright-up circuits to the tube,
duff triggering due a resistor gone high in the trigger circuit,
the alt sweep which had a duff transistor
the channel 2 vertical position pot which had a broken wiper (being a an obscure dual-gang one I went hardcore and soldered the wiper back together)
Aforementioned pot being very stiff, I cleaned and greased the shaft. Knob wasn't fitting properly because of spindle being stiff it had worn grooves in it, some thread-lock fixed that,
The most bizarre fault was regular-ish transients in the "A" timebase, causing the display to flick off and on every few seconds. Trying to track it down I killed both transistors in the multivibrator which produces the flyback, eventually found the j-fet which receives the signal from that multivibrator has a jittery gate. Lucky I had an old j-fet - wrong type but works ok
Another bizarre fault in the timebase is what I'm calling the cheese triangles. When the larger value caps are engaged in the timing circuit, the flyback becomes a slope. So I'm getting cheese triangles instead of sawteeth. Nothing wrong with any components in the t/b afaict and nothing connected which can affect it. Leaving that one for now (and it did it before I swapped the fet so it's not that).
Whilst I was messing about with the "A" timebase, the "B" timebase stopped working. At least that was easy, ramp transistor gone. I will have to replace it another day.
So I'm left with faults consisting of:
Bright spot at start of display - t/b delay not being blanked I guess
Z-modulation - I expect to find a duff cap in the hv supply. Also now gets gradually dimmer as it warms up.
Stiff channel 1 position pot, getting like the ch2 one was
Stuff I don't know about yet but will doubtless turn up.
Still looking out for a suitable switch to replace the one for ch2 "invert". The original fell apart when I first did fault fixing on this 'scope. It's a slide switch but because of the pin spacing I've put in one of the sort normally found in gangs on older stereo systems. Be nice to find a slide switch with that pin-spacing, to replace it properly. Or maybe I could go hardcore again and mould a new slider for the old one, out of something...

 

[throbscottle]

In case anyone is actually reading this...
The jfet wasn't the cause of jumpy signal, as it started doing it again. Funny how it had settled down though. So now I'm thinking the cause is in the output of the trigger circuit somewhere, which I know is faulty anyway (it gets less able to trigger as it warms up)
I realised the brightness control / blanking / bright-up is ac coupled to the CRT cathode. I had it in my head that maybe it works like a charge pump, but since the simulation oscillates I guess it's supposed to do that and gets turned into a DC level at the tube. So maybe the origin of the z modulation is there, not the HT supply.

 

[tomizett]

The jfet wasn't the cause of jumpy signal, as it started doing it again. Funny how it had settled down though.

I had a repair do that the other day - replaced a part, fixed. Put the case back on, not fixed again. Probably todo with jiggling the real culprit about or warming it with the iron...

Thanks for sharing your experiences - nice to see people sharing the love for the older kit. My father has a Cossor scope that I used to play with as a kid. That's going to be one hell of a project for somebody one day, as it's been sitting on a shelf in the shed for a good few years now. I suspect it will want a bit of attention by now.
I believe that one had Nuvistors in, which are probably quite rare these days. Does yours use them?

 

[throbscottle]

Nah, all solid state Though most of the transistors aren't made any more, it's hard to find equivalents in through hole packages because they are high speed switching ones and have been replaced by SMD's. I believe the design dates from the late '60's so it just missed out on the valve era. I think my one was probably made in the '70's but that's more of a guess, there's nothing to actually say when it was made unless someone comes up with a list of serial numbers to go from. Except it's s/n has been rubbed off. Ho-hum.
40MHz dual trace, dual timebase 'scope, still respectable even today. The old Motorola transistors have gold leads and "ebc" moulded on the casing. I don't know when they stopped doing that. I keep looking for NOS ones on fleabay. Expensive though. I only paid 99p for the 'scope!

 

[atferrari]

Z modulation deals with HV?

 

[throbscottle]

Think about it. The high voltage supply affects the brightness of the display, so any variation in it will create z-modulation, such as if a smoothing cap has gone. The alternative is that a bogus signal is getting in on the blanking/bright-up input (which still modulates the grid). Actually thinking about that may be more likely.
I really must invest in some freezer spray. I know some of these faults are thermal. It would make the task so much easier!

 

[atferrari]

Yes, you are right. My question was wrong.

For the design mentioned here I used a very low voltage to cut the beam off, after every dot was produced. Just by looking at the circuit it is evident that HV is involved.

My question should have read something like: have you considered that the control circuit could have failed? You answered my question in your last post. Sorry for disturbing your thread.

 

[throbscottle]

That's what threads are for, being disturbed!
Anyways, there is a 470pF 16KV doorknob capacitor in parallel with the final anode, it was crackling so I gave it and the voltage multiplier rectifiers a good clean. Now it makes a whirring sound instead! So I think it can't stand the voltage anymore. The anode loses its charge straight away too. I tried running the scope with this cap disconnected, display looks the same, noise stopped. I'm wondering if I can replace it with 2x 10KV 1000pF caps in series. But I'll try turning down the EHT first I think.

 

[throbscottle]

I made a high voltage probe to check that final anode voltage. Only about 9KV - it's supposed to be about 14! So I need to replace that cap before I adjust it.
I found the cause of the z-modulation. The 120V supply had a small transistor (acting as current source? Not sure what) which must have a crack in the silicon, responds to heat and mechanical pressure on the leads. I thought it was a dry joint at first, especially since it worked when I re-soldered it. Stopped again soon though. So I replaced that and it's fine. There was 3V of ripple on the supply, it was 14V down and the voltage was gradually changing.
So now the scope has the fault remaining which was the original problem when I got it, trigger circuit not working correctly. It did work after I'd cleaned all the switches, but now is intermittent and mostly not at all. It plays merry hell with the timebases.
Extra hard to diagnose because it's connected via rotary wafer switches to get all the different trigger and t/b modes, so I'm having to look at 3 different schematics and a drawing of how the switches are wried, all on different pages and in different orientations in my photo-copied manual. Also if I want to get to the underside of the board to unsolder/solder anything I have to un-solder 5 wires and 2 air-wired resistors to be able to move the board. Do that a few times and it starts looking unhappy

 

[throbscottle]

Oh ok I got the EHT voltage wrong - it's supposed to be between 8.8 and 11.2 KV. So I suppose it's correct then. Still need to replace that cap though, but at least I can get a less obscure (and therefore cheaper) one.