Hi.
I need help to repair the power supply of a oscilloscope PM3207.
The oscilloscope came to me with a lot of problems with the power supply (the old owner tried to repair it without success).

I now have all the correct voltages except HT. In the HT circuit I found the diode V800 and transistor V801 burned. I Replaced the CI D801 and diodes V802 and V803 with new ones and checked all the other components (one by one) and they are all OK.
Now whenever I turn on the power, the V801 transistor burns...

What is wrong, for that to happen?
Is the CRT broken?

Many thanks.
Regards.
Carlos

--
From PORTUGAL, with regards
Carlos

No, it's probably not the CRT.

T801 and V801 form a simple oscillator, the output of which is used to generate a negative high voltage.
It is a little bit confusing that the third winding of T801 is drawn separated from the other two.

How fast does your transistor burn?
Also check the rectifier circuit (C804 806 807 and V802 803). A short here will overload the oscillator/generator circuit. V801 might go but not in a second.

T801 should be a small transformer with 6 terminals.
It might have a short circuit in one of the windings (hard to check, could be done with a square wave generator and a scope, but your scope is broken...). shorted windings are probable with HV transformers, but it depends on how well built it is and how much power it must deliver. I don't have experience with this one. Does it look ok?

An interruption or short in C805 might cause trouble. You might replace it but you might burn another transistor if it's not the cause.

You say V800 was also burnt. This is odd because I think it would never get too much power on this position in the circuit. Because of V800, also check R804.

Maybe C802 is interrupted. Check its capacitance. It is needed to make V801 oscillate, without it, V801 would probably stay in an on-state and maybe burn.

You would also like to check the regulator circuit around D801. With V801 removed, solder a 82k5 (like R802) (82k is also ok, or even 47k) resistor temporarily between -12V and pin 3 of D801. The output on pin 5 should be the same voltage as pin 3 now when you switch the scope on. (with V801 removed, there is no HV, no smoke and no stress on T801). If it is not, D801 is defective or one of its resistors.. But I don't see if it could really fry your V801. It could give +10 or12V output if defective and make V801 work too hard, causing a too high HV (which might then destroy T801).
By the way, if C802 is interrupted, D801 would give a high output too, because there is no HV.
Do not forget to remove the temporary resistor after checking D801.

So the first two things are to check the two capacitors and the rectifier circuit.
Then the most likely cause is the transformer, which is a bit hard to verify, and probably hard to find a replacement for.
The regulation circuit might have a problem, check that too, just to be sure.

Onno

I also been looking at it.

Put a 24V lichtbulb between Bc337 and T801, so the current is limited at 0,8A
the max current BC337.

Take R811 out at 1 side.

Than LM741 flip arround V1 to drive BC337.
Think cut that 10M resistor and put an external oscilator at V2
See what hapens

be careful: the -1500V across C806 will not leak away, it may become higher because there is no load.

Many thanks Onno and Maurice.

Maurice, i don't have a wave generator
At this moment a don't have more BC337 to check. Meanwhile I tested the ua741 and it is OK, and all the resistors and condensers (again). The tensions are all correct.
I removed the transformer, measured it and it looks ok. There are no signs of being shorted.

Regards
Carlos

--
From PORTUGAL, with regards
Carlos

It's verry hard to tell if there are short windings primair.

As I said check with a lichtbulb is series.
Also dubble check that feed back resistor 801 10 MHm and Focus network.
If that fails the opamp wil not flip and Transistor will be conducting.

Hi.
More news.
I found a trace in the printed circuit, open, at point A.
I put a 24 V lamp on the transistor collector and got the following measurements, and now the transistor doesn’t burn, but I don’t have any trace, or point on the CRT.
Are the tensions correct?
Do I have to turn off the points X and Y to see the point?
How can I measure the HT?

Many thanks.
Regards

--
From PORTUGAL, with regards
Carlos

Hi Carlos
The HV on the cathode is -1500V according to the manual. Do not try to measure this with a DVM, most are built for 500 or 1000V maximum. You may get an internal flash-over which will badly damage your DVM.
You need to use a HV probe, to divide the high voltage.

But your scope already has a voltage divider for the HV voltage: R801/R802, it provides the input for the control circuit around D801.
It has a 82,5/10082,5=0,0082 ratio, so with 1500V input you will get -0,37V on the +input of D801 (pin3).
Resistors R806, R807, R803 define the amplification of D801 to about 181 times.
You can calculate the input and output voltage of D801 (pin 5) for different values of the HV.
I put a simple calculation in a spreadsheet to produce a table:
HV +in D801 Output D801
-1600 -1,19 -12
-1500 -0,37 -12
-1400 0,45 +2,5
-1300 1,26 +12
-1200 2,08 +12
-1100 2,90 +12
-1000 3,72 +12
-900 4,54 +12
-800 5,36 +12
-700 6,17 +12
...
-200 10,27 +12
-100 11,08 +12
0 11,90 +12

Of course, this is theoretical, the output will be less than +12V and higher than -12V.
From this simple model of the amplifier, it will stabilize the HV to about -1405V (where the output voltage is -10...-5V)
But there are small errors in the model, probably it will be -1500V as Philips says.

Anyway, your results are 11,4 V on pin 3, meaning HV must now be 0..-100V, so if it is so low for now you _can_ measure the HV with your DVM. Don't do that if you measure lower than 5 V on pin 3

The low value may have 1 of 2 causes:
- the 24V lamp is limiting the power to the transformer T801
- T801 has a short in one of the windings (or the rectifier circuit C804-806 V802-803 has a bad component)

(by the way: when we are talking about a shorted winding in a transformer or coil, this means two adjacent loops of the wire are shorted, not the whole coil. This will absorb a lot of energy from the transformer)

You may replace the lamp by a 12V 5W car lamp or 10 ohm resistor, then 4,7 ohm, then 2 ohm etc to see if you get HV .... if you don't, probably the transistor will get hot again, because there is still a fault in the circuit.

Question.

Is the 24V light glowing? If so thats wrong.(a defect circuit that is)
It must be off or at least dimm.

V800 must be 2,8V across.

Than D801,
As seen in table by Onno, around the 0,4 there is a flip of the output.
That is the oscillation point were BC337 has to go oscillating.

That is only accomplished to subtract from that +12V power source.
That subtraction is done through R801 voltage.
+12V +(-R801 voltage) = voltage pin 3 D801, where R801 is a negative voltage divider from that -1500Vdc

As you measure +11,4V there is no negative voltage from resistor network
R801 etc.

The difficulty is , it is not a static power supply.
In that way you have to see if at start up for a short moment there'is a negative
peak or not.

The only way to see that is to manually start that BC337 and watch with a scope
that point 3 D801 if it briefly change, and so the output also.

To start the oscillation, at start BC337 conduct, a power peak provide charge the transformer, due to timing with capacitors the BC337 will close and the transformer will be discharge and That will make the first negative voltage pulse.
That must in fact be enough to flip D801 again.
After milli seconds it will stabilize in oscillation and minor changes in voltage will provide a longer or shorter pulse-lenght of conduction BC337.

DC measurements can't be done in running mode on that pulse network.
Only at the DC HV end of the circuit.

The problem., as you already checked the components there must be a problem
why at start the negative pulse is not ending up at the D801.

I think Pull that CRT plug first, and see what happens.
Than also check seperately those zeners D800 (2,4V) and D804 (39V)
On the internet is to find how to make e very simple test circuit.

Then also make with a potmeter to vary voltage at pin 3 D801.
Simply take a 50K or so potmeter, attache to +12V and -12V.
The middle should now be 0V, and by turning around that 0,4V it must flip the output.
Than see if Bc337 will close.

Pulse networks power supply's are not very easy to repair if there are big problems.
Its all about timing through coils and capacitors.

Thank you both for your patience.
Maurice the lamp turns on normal.
I'm going to do these tests and then I'll let you know.

Regards

--
From PORTUGAL, with regards
Carlos

Hi Maurice,
I'd like to disagree with you on the voltage over V800. It is not reverse-biased constantly.
Its anode is connected through R804 to the (relatively) positive output of D801, conducting to feed the bias current to V801, controlled by D801.
When V801 is oscillating, pulses from the feedback winding of T801 will make the connection point V800 C802 T801 go positive until V800 starts to conduct. I believe a diode is needed here to prevent T801 from being dampened by R804, but it must be a zener to prevent the voltage on this point to become too high.
So there will be a signal across V800 between peak values -0,7 and +2,8V, or maybe normally not even reaching +2,8. You cannot measure this with a DVM.
Carlos has measured 0,7 V across V800, which means it is conducting (or most of the time, there might be a train of narrow pulses) and V801 is not oscillating.

Onno,
Indeed so.
There is lot going on in those pwm's.
Don't see why that zener instead of a diode.
Maybe to level the C802 voltage with some timing issue that is.

As I stated think at the moment there is nothing going on and just a steady
dc mode. The reason BC337 get defect.

There must be a problem with that feedback not reaching D801 pin 3.
Or though those secondary capacitors, some defect resistors.
It is essential that secondary circuit to be in good condition.

May be at start up there is a little oscillation but due to f.i. bad capacitors
it is not reaching the voltage and so raising the voltage to point of pure dc
and oscillation shuts down.

There for TS needs a scope, but hé that one broke down.
Also a lack of puls generator to manually oscillate that BC337 to see if somehow
you get a little voltage at C807.

Hi.
Now i have an oscillator an a oscilloscope.
I pull up the resistor 10M and put the oscillator in pin 3 and put a 24V light bulb between Bc337 and T801.
I got these photos.
I wanted your opinion.

Regards

--
From PORTUGAL, with regards
Carlos

Interesting. The transistor and transformer do oscillate when the base is not too positive.
What is the frequency you used? The oscillator should operate at 30 kHz

The signal you use seems to switch the oscillator on and off. But it is oscillating when it should be off, and is drawing lots of current when it should oscillate.

Please disconnect T801 from R808. feed a 30kHz sine of 1Veff into R808 and look at the signals on T801 (not the HV winding).
Touch the connection of T801 to C804 with a small, well-insulated screwdriver, 750V AC should be here, you should see a small spark if you have the screwdriver 1 mm from the connection.

Interesting. The transistor and transformer do oscillate when the base is not too positive.
What is the frequency you used? The oscillator should operate at 30 kHz

The signal you use seems to switch the oscillator on and off. But it is oscillating when it should be off, and is drawing lots of current when it should oscillate.

Please disconnect T801 from R808. feed a 30kHz sine of 1Veff into R808 and look at the signals on T801 (not the HV winding).
Touch the connection of T801 to C804 with a small, well-insulated screwdriver, 750V AC should be here, you should see a small spark if you have the screwdriver 1 mm from the connection.

Hi.
I put new the V802, V803, C804, C806 and C807 and I did it (30 kHz, 1Veff). I don't have the 750V AC.
Then I removed the light bulb between Bc337 and T801, and the and the transistor burns.
For all this, I think the defect is in the T801 transformer.

--
From PORTUGAL, with regards
Carlos

I am also worry about that oscillating BC337.

Could indeed something with that transformer.
Pull secondary R811 and series with a 10M Ohm resistor to ground.

Now you have only made a rough power supply.
Again at R808 that 30kHz as Onno mentioned.

See what happens.
Remember that with BC337 you short out -12V to +12V
Only to the fact of delay of current in the coil T801 it will act as a resistor.
But if that delay is a bit too long it will act as a DC and there BC337 will blow.
As if any other big transistor at this point.

So don't lower the frequency, keep light bulb in series too.

That is the main problem with PWM supply's.
If there is even a minor problem like chain reaction all components give up.

Than what you can do is measure that transformer with frequency.
Think just solder out.
Put primair winding frequency and parallel your scope.
Find a voltage peak at 1 high frequency.

Then look up internet how to measure that coil impedance through scope and generator.

Two questions:
1- what were the waveforms on the base, collector of V801 and on the primary side of the transformer (where it was connected to the bulb)

2- did you have spare BY409's to replace V802, 803? (I wouldn't have the originals, but use BY176 from an old TV)
they are quite special rectifiers, rated at 11,5kV. In 1980-1990 color TV's there would have been one (or an equivalent) for the focusing voltage to the CRT.
They are quite unlikely to fail in this place, where they get a peak inverse voltage of 1500V (3000 if you take a wide margin)

I wouldn't replace them by an ordinary 1N4007, maybe a chain of four each if I was out of options.

By the way, it you test a BY409, they have a forward voltage of 36V so you might think they are interrupted if you test them using the diode test function of your DVM.

Hi Onno,
i know i can't test these diodes with the DVM, because that i put news. Not the BY409A but the equivalent BY713.
I did the test that Maurice said but I didn't get any HT signal.
these are the photos of the waveforms.

After that, i removed the transformer and made the circuit on a test board and
did not get any HT signal.

When I put the 30kHz signal the light bulb turns on.

--
From PORTUGAL, with regards
Carlos

After all this, the transformer is the remaining suspect.

But the oscillograms are a bit confusing: I'd expect the collector current to clip. But you have a nice sine.
What is the scale (V/div?)

Hopefully someone can help you to a new transformer.

Many thanks Onno.
I think that is the transformer

In the first picture i have 1 Vpp, in the second i have 10 mvpp, on the third i have i little less.

Best regards,
Carlos

--
From PORTUGAL, with regards
Carlos

Good job so far in debugging the problem.

Rewinding the transformer is duable.
Bit of a job but can be done.

If ferrit is glued heat it first up than hopefully the glue weakens.
You can buy those cores (link) and with some re engineering, counting it can be done.

See it as a challenge, i think.
They are not that expensive to buy on the internet.
But should be fun you get this one going again.

Or maybe there are other way's to get that 1500V.