[SOLVED] Problem with an old 10KV/3mA DC high voltage module

[betwixt]

I suspect the -1.5V is the average measured by your meter, there is no negative supply to produce a steady -1.5V at that point.

Quick check - run it fo a few minutes, switch off and put your finger on R84. It should be very hot, if it isn't, change it for a new one, it has to be open circuit.

Brian.

 

[pplus]

Swapping non oscillating BDY56 with oscillating BDY56 from the working unit I saw it start oscillating on faulty unit as well. So it seems I need to find an oscillating BDY56 as a pair for the BA157 diod at its right.

Swap new non oscillating BDY56 with oscillating BDY56 to the working unit and check start oscillating. Try find a match pair.

 

[agoo]

Brian

Yes as you said I have a sin wave signal with about 3V peak to peak (below zero line) and about 800 kHz frequency at that point. After a about 20 sec R84 is warm and BDY56 transistors are a bit warmer but not hot. Checked R84 with ohmeter. It is OK about 10K.

pplus

I can't risk working unit because it is the only working pair of BDY56 I have. Prefer to find the problem in a more risk free ways.

- - - Updated - - -

When I try different BDY56s I see voltage at point "P" changes from about 8mV up to 70mV (at working unit 10V) depending on the BDY56. Also peak-to-peak of the signal at point 16.1 changes slightly. Peak to peak at the working unit is about 5V while at the faulty unit it is about 3V. Also in faulty unit signal is close to sinusoidal shape but in the working unit it does not looks like sinusoidal shape.

 

[agoo]

I found a document showing frequency of the oscillator is 100 KHZ. I measured about 800 KHZ at point 16.1 (my oscilloscope shows this). Is this the oscillator frequency? Could this give a clue about the problem? Which parts contribute to this frequency (I can check or replace them)?

 

[betwixt]

The frequency does indeed seem wrong. Given it's age, even 100KHz is quite high so when you measured 800KHz it would be far too high.

It's a self oscillating circuit, each transistor pushes the other so they 'see-saw' with one then the other conducting. The frequency is decided by a number of factors, the main ones would be the three capacitors across the transformer windings (left side on the schemaic). If one of those went open circuit it would increase the frequency and cause a corresponding drop in efficiency. Note that there is nothing to maintain frequency stability in the circuit and it could vary maybe +/- 10% from one unit to another and it is likely to drift as the temperature changes.

The other possibility is that you have a short circuit inside the transformer. It is unlikely the windings connected to the transistors would be damaged because I would guess they are only a few turns of fairly thick wire and hence almost indestructable. The high voltage side will have hundreds of turns and to fit it all in, would be much thinner wire and more likely, especially with high voltages between the turns, to suffer insulation breakdown. If that has happened you really need a new transformer. You might be able to rewind it yourself but it isn't an easy job. If the wires are deep inside resin or otherwise unreachable the job would be virtually impossible.

Brian.

 

[pplus]

The frequency does indeed seem wrong. Given it's age, even 100KHz is quite high so when you measured 800KHz it would be far too high.

I agree with Brian.
BDY55 – BDY56
Turn-on time --0.5μs
Turn-off time --2μs
Fmax ~ 1/(2*2.5μs)=200KHz

The other possibility is that you have a short circuit inside the transformer. It is unlikely the windings connected to the transistors would be damaged because I would guess they are only a few turns of fairly thick wire and hence almost indestructable. The high voltage side will have hundreds of turns and to fit it all in, would be much thinner wire and more likely, especially with high voltages between the turns, to suffer insulation breakdown. If that has happened you really need a new transformer. You might be able to rewind it yourself but it isn't an easy job. If the wires are deep inside resin or otherwise unreachable the job would be virtually impossible.

Brian.

The absence of oscillations is the result of a low Q oscillation circuit. Check the C26 C27 C28 and and measure inductance of the transformer windings. It seems that the transformer windings are shorted.

 

[agoo]

Monitoring signal at points "H" and "R" using oscilloscope, I have almost square signal with about 60V peak to peak at both "H" and "R" on working unit with 45 KHZ frequency but in faulty unit only DC voltage about 60V on these points with no oscillation. The only thing left unchecked is the transformer. I will try to change it and see how it will go.

 

[agoo]

Good news guys. :-D After changing transform primary the unit works like a charm. Thank you all for the help.

I knew I am a kind of a person who luck never come close to me. It seems this is right after all. Transformer primary with 1 mm thick copper wire is the last thing to go wrong and this should happen to me. What do you call it except s*** luck? :bang:

Thank you all again. :thumbsup: