You are here

CRT scope high voltage supply problem?

Customising, building or repairing your own gear? Need help with acoustic treatment or soundproofing? Ask away…

Re: CRT scope high voltage supply problem?

Postby Adam Inglis » Wed Feb 05, 2014 6:59 am

Just to be clear, I'm simply making a high-resistance probe here, and not adding a voltage divider, as suggested here?

The resistors I bought are high voltage 15W ones, and so too big to fit into a biro. Electrical tape and heat-shrink perhaps?
User avatar
Adam Inglis
Regular
Posts: 422
Joined: Wed Sep 01, 2004 12:00 am
Location: Gold Coast Queensland Australia

Re: CRT scope high voltage supply problem?

Postby Folderol » Wed Feb 05, 2014 7:57 pm

Adam Inglis wrote:Just to be clear, I'm simply making a high-resistance probe here, and not adding a voltage divider, as suggested here?
Their justification for a potential divider setup is safety, but at the voltages you are working and with the series resistance this really isn't a problem. Absolute worst case a 90M probe at 1kV will pass 11 micro amps, which if you are in contact with the bottom end would feel slightly unpleasant but would be no risk at all. In this situation we are specifically trying to minimise the loading of the high voltage supply, and a shunt resistor would tend to detract from that. When plugged in to the meter, the meter provides the lower part of a potential divider and if you use a nice insulated plug you won't be able to come into contact with it at all. You should, of course securely connect the 'cold' end of the meter to 0V before trying to use the probe.

The resistors I bought are high voltage 15W ones, and so too big to fit into a biro. Electrical tape and heat-shrink perhaps?
Those are rather big resistors! Electrical tape would be likely to produce a noticeable leakage path and I wouldn't be too trusting of heatshrink if you touched the probe part way up. The idea of the biro tube is to provide physical support as well as insulation. However, any rigid plastic tube would probably do. The joint with the lead to the bottom resistor should be well inside the tube and finger proof. The hot end of the top resistor should be turned into a probe point, and anchored to the tube in some way - a dollop of epoxy would do the trick.

Finally, you can calibrate the probe by using it to measure a known lower voltage. For example, the 50V supply (as measured with your meter directly) should read 5V via the probe. Up to 20% difference is acceptable. A dramatic difference needs investigation.
User avatar
Folderol
Jedi Poster
Posts: 11356
Joined: Sat Nov 15, 2008 1:00 am
Location: The Mudway Towns, UK
Yes. I am that Linux nut.
Onwards and... err... sideways!

Re: CRT scope high voltage supply problem?

Postby Adam Inglis » Wed Feb 12, 2014 5:35 am

OK thanks Will.
I've managed to make a safe probe, I estimate it to be about 91.5 Mohhm from segmental readings I took before I epoxied it up, so its pretty close to ideal. Checking it on the 150 and 50 volt supplies confirms it divides by 10.
So now checking the high voltage points shows I'm getting 1.8kvolts where there should be 1kvolt, and 4.8kvolts where there should be 3kvolts - probably why I cooked my VOM!!

Any ideas?

schematic link
User avatar
Adam Inglis
Regular
Posts: 422
Joined: Wed Sep 01, 2004 12:00 am
Location: Gold Coast Queensland Australia

Re: CRT scope high voltage supply problem?

Postby Folderol » Wed Feb 12, 2014 9:07 pm

That's very interesting. All the supplies except the 3kV one are about twice what they should be, and that one is far too high. That suggests to me that the voltage control circuit isn't working correctly. My first suspicion would be C72 going leaky.
User avatar
Folderol
Jedi Poster
Posts: 11356
Joined: Sat Nov 15, 2008 1:00 am
Location: The Mudway Towns, UK
Yes. I am that Linux nut.
Onwards and... err... sideways!

Re: CRT scope high voltage supply problem?

Postby Adam Inglis » Thu Feb 13, 2014 12:55 am

OK, that might explain why those three transistors have weird voltages on them.
User avatar
Adam Inglis
Regular
Posts: 422
Joined: Wed Sep 01, 2004 12:00 am
Location: Gold Coast Queensland Australia

Re: CRT scope high voltage supply problem?

Postby Folderol » Fri Feb 14, 2014 12:31 am

Actually, it's a weird circuit anyway!

Other likely candidates are R131, RV14B, RV20B and to a much lesser extent R129. If any of these have gone high resistance there will be a more positive voltage on the base of Q40, which will conduct more and thus cause Q41 to also conduct more, driving Q42, the oscillator, harder.

Doncha just love DC coupled circuits :)
User avatar
Folderol
Jedi Poster
Posts: 11356
Joined: Sat Nov 15, 2008 1:00 am
Location: The Mudway Towns, UK
Yes. I am that Linux nut.
Onwards and... err... sideways!

Re: CRT scope high voltage supply problem?

Postby ef37a » Fri Feb 14, 2014 8:03 am

Folderol wrote:Actually, it's a weird circuit anyway!

Other likely candidates are R131, RV14B, RV20B and to a much lesser extent R129. If any of these have gone high resistance there will be a more positive voltage on the base of Q40, which will conduct more and thus cause Q41 to also conduct more, driving Q42, the oscillator, harder.

Doncha just love DC coupled circuits :)

Weird indeed Will!
Generally such generator circuits have an obvious feedback path from the HT side. This is usually a high value resistor, often 1 to 3.3meg and they used to go high/O/C in TV PSUs with devastating results.
I can see no such feedback here and the whole thing seems to run open loop on a wing and a prayer!

Dave.
ef37a
Jedi Poster
Posts: 12024
Joined: Mon May 29, 2006 12:00 am
Location: northampton uk
#They did not listen, they are not listening still...Perhaps they never will?#

Re: CRT scope high voltage supply problem?

Postby Adam Inglis » Fri Feb 14, 2014 8:03 am

Many thanks for your help with this, Will.
I replaced C72 - no joy.
R131 did indeed measure high, over 5 meg, but replacing it caused no change to the base voltage on Q40 which remains at -13v. The other resistors check out OK.
Might be time to start replacing transistors?

(From my basic understanding of transistors, Q41s emitter should not be more negative than its base, it should be more positive.
And likewise, Q42s base should not be more negative than its emitter, it should be more positive, non?)
User avatar
Adam Inglis
Regular
Posts: 422
Joined: Wed Sep 01, 2004 12:00 am
Location: Gold Coast Queensland Australia

Re: CRT scope high voltage supply problem?

Postby Adam Inglis » Fri Feb 14, 2014 8:07 am

ef37a wrote:I can see no such feedback here and the whole thing seems to run open loop on a wing and a prayer!
Dave, would the feedback you're talking about be the connection via R131 (2.7Mohm) to the base of Q40?
User avatar
Adam Inglis
Regular
Posts: 422
Joined: Wed Sep 01, 2004 12:00 am
Location: Gold Coast Queensland Australia

Re: CRT scope high voltage supply problem?

Postby ef37a » Fri Feb 14, 2014 8:14 am

Adam Inglis wrote:
ef37a wrote:I can see no such feedback here and the whole thing seems to run open loop on a wing and a prayer!
Dave, would the feedback you're talking about be the connection via R131 (2.7Mohm) to the base of Q40?

It could indeed!!! DOH! HOW did I miss that? There is a V set pre-set as well! If it turns out to be that resistor and the pre-set R is cheap, open frame thing I would replace it with at least a new enclosed type or in fact go for a multi-turn Cermet device, only a pound or so.

Dave.
ef37a
Jedi Poster
Posts: 12024
Joined: Mon May 29, 2006 12:00 am
Location: northampton uk
#They did not listen, they are not listening still...Perhaps they never will?#

Re: CRT scope high voltage supply problem?

Postby Folderol » Fri Feb 14, 2014 6:15 pm

I must say it is a very badly drawn schematic. It took me quite some time to work out what was what.


From my basic understanding of transistors, Q41s emitter should not be more negative than its base, it should be more positive.

Now that's rather interesting.
Q41 emitter should be at 0V and it's base should be about -0.6V. If the base is positive it means that Q40 and the resistor chain feeding it is actually trying to regulate the drive. I wonder if Q41 has gone S/C collector - emitter.



And likewise, Q42s base should not be more negative than its emitter, it should be more positive, non?

Q42 is an inductively coupled oscillator circuit, and as such the average base voltage is very likely to be negative with respect to the emitter.
User avatar
Folderol
Jedi Poster
Posts: 11356
Joined: Sat Nov 15, 2008 1:00 am
Location: The Mudway Towns, UK
Yes. I am that Linux nut.
Onwards and... err... sideways!

Re: CRT scope high voltage supply problem?

Postby Adam Inglis » Mon Feb 17, 2014 9:17 am

Well, this is embarrassing..
In replacing those transistors, I realise I incorrectly identified their pinouts when I took those voltage measurements.
So now they are replaced, and after tracing and double checking the board and the schematic, the voltages are
Q40 collector 10v, base -24v, emitter -13v
Q41 collector -15v, base 10v, emitter 0 v

in other words, little change. High voltage side of things much the same.
The large negative voltage on the base of Q40 looks wrong to me - I wonder if C90 could be leaky pulling the base towards the -1000v. I'm still looking for high voltage caps in these values - they're getting thin on the ground.

Actually the base-emitter voltage difference on both these transistors looks wrong.
User avatar
Adam Inglis
Regular
Posts: 422
Joined: Wed Sep 01, 2004 12:00 am
Location: Gold Coast Queensland Australia

Re: CRT scope high voltage supply problem?

Postby Folderol » Mon Feb 17, 2014 6:34 pm

I'm relatively happy about the Q40 voltages. These seem 'correct' assuming the circuit is trying to pull back the voltage, but is prevented from doing so due to some issue with Q41. I'm puzzled by its collector voltage. If the transistor is in cutoff (as the base voltage suggests) the collector should be well below Q42 emitter voltage - R122 is trying to pull it down to -50V.

My next suspect is C71 (especially if it's an old wax paper type). If it is leaky it could be supplying uncontrolled current to Q42 base.

Another far more remote possibility is R122 itself being O/C. On very rare occasions, if there is no resistor to draw away current, there can be enough base-collector leakage to bias a power transistor like this into operation. I assume there actually is -50V on the 'bottom' end of this resistor.
User avatar
Folderol
Jedi Poster
Posts: 11356
Joined: Sat Nov 15, 2008 1:00 am
Location: The Mudway Towns, UK
Yes. I am that Linux nut.
Onwards and... err... sideways!

Re: CRT scope high voltage supply problem?

Postby Adam Inglis » Wed Mar 05, 2014 9:57 am

OK, so I'm in the process of ordering replacement caps for this thing, but I'm having a problem with caps C78 and 79, on the high voltage side of the coil.
Schematic here.
Image here.
What kind of caps are these?
The schematic says they are 0.033uF rated at 1 kV, but the image shows the numbers "0.033/10/1600"?
... and yet, looking at the circuit, surely they must have something close to 4000 volts across them in normal operation??
User avatar
Adam Inglis
Regular
Posts: 422
Joined: Wed Sep 01, 2004 12:00 am
Location: Gold Coast Queensland Australia

Re: CRT scope high voltage supply problem?

Postby ef37a » Wed Mar 05, 2014 10:58 am

Adam Inglis wrote:OK, so I'm in the process of ordering replacement caps for this thing, but I'm having a problem with caps C78 and 79, on the high voltage side of the coil.
Schematic here.
Image here.
What kind of caps are these?
The schematic says they are 0.033uF rated at 1 kV, but the image shows the numbers "0.033/10/1600"?
... and yet, looking at the circuit, surely they must have something close to 4000 volts across them in normal operation??


No because they are in series with C76 C77 and they are both much smaller in value and rated at 4kV so I would estimate that those caps have well under 1kV on them. But! Tis a bugger of a circuit to read!

Oh! ***! the bottom end of the .033s goes to deck!

Dave.
ef37a
Jedi Poster
Posts: 12024
Joined: Mon May 29, 2006 12:00 am
Location: northampton uk
#They did not listen, they are not listening still...Perhaps they never will?#

PreviousNext