Author Topic: Help repairing power supply  (Read 3735 times)

0 Members and 1 Guest are viewing this topic.

Offline crt_cowboyTopic starter

  • Regular Contributor
  • *
  • Posts: 71
  • Country: be
Help repairing power supply
« on: January 17, 2025, 08:02:31 am »
Hi all,

I am currently trying to repair a power supply.

From the supply voltage (6V), it generates +9V, -9V and +300V

It uses an oscillator consisting of transistors Q3 and Q4 and a transformer T1. This transformer has 3 secondary windings.

The voltages are regulated by a stabilization circuit consisting of a Zener network on the HV of an integrated circuit LM10 Z3 and transistor Q5. The base of Q5 is made more or less conductive by LM 10 depending on the variations in the HV.

Attached the diagram.

Adjusting R34 should adjust the output voltage. In my case changing it value doesn't change I can barely get +-8.4v and 270v by lowering it's value.

Anyone has suggestions on what to look at ?

I checked all resistors and they seems within acceptable range.

Many thanks in advance
 

Online Zero999

  • Super Contributor
  • ***
  • Posts: 20533
  • Country: gb
  • 0999
Re: Help repairing power supply
« Reply #1 on: January 17, 2025, 09:28:53 am »
What's the voltage on pin 8 of Z4?
 

Offline crt_cowboyTopic starter

  • Regular Contributor
  • *
  • Posts: 71
  • Country: be
Re: Help repairing power supply
« Reply #2 on: January 17, 2025, 09:59:16 am »
8.4v and I cannot get it higher adjusting R34
 

Online Zero999

  • Super Contributor
  • ***
  • Posts: 20533
  • Country: gb
  • 0999
Re: Help repairing power supply
« Reply #3 on: January 17, 2025, 10:04:42 am »
8.4v and I cannot get it higher adjusting R34
That can't be right. The voltage on pin 8 of Z4 (the LM10) can't exceed the power supply voltage to it, which is 6V. The voltage on pin 8 should be 200mV.

Are you sure you've not put one of the meter probes in the wrong place?

EDIT:
Refer to page 17 of the LM10 data sheet.
Pin 8 connects to the inverting input of an op-amp, with a 200mV reference connected to the non-inverting input. If the loop is closed, the input voltages to an op-amp should be equal, so the votlage on pin 8 should be the same as the 200mV reference.
https://www.ti.com/lit/ds/symlink/lm10.pdf
« Last Edit: January 17, 2025, 10:40:06 am by Zero999 »
 

Offline crt_cowboyTopic starter

  • Regular Contributor
  • *
  • Posts: 71
  • Country: be
Re: Help repairing power supply
« Reply #4 on: January 17, 2025, 10:41:45 am »
 Oh yes I went too quickly indeed 200mv on pin 8 (my ground probe was not connected to the right place)
« Last Edit: January 17, 2025, 11:58:25 am by crt_cowboy »
 

Offline crt_cowboyTopic starter

  • Regular Contributor
  • *
  • Posts: 71
  • Country: be
Re: Help repairing power supply
« Reply #5 on: January 17, 2025, 12:38:59 pm »
Here are some additional measurements:
Pin 1: 380mV
Pin 2: 4.8V
Pin 3: -430mV
Pin 4: 0V
Pin 5: 33mV
Pin 6: 1.7V
Pin 7: 6V
Pin 8: 200mV

Any suggestions based on this ?

Many thanks

EDIT:
I have redone the measures and for pin 1, 2 and 3 I now get 4.8v. I am not sure what is going on. Same voltmeter and same measurement procedure...
« Last Edit: January 17, 2025, 04:16:57 pm by crt_cowboy »
 

Online Zero999

  • Super Contributor
  • ***
  • Posts: 20533
  • Country: gb
  • 0999
Re: Help repairing power supply
« Reply #6 on: January 17, 2025, 11:21:28 pm »
So pins 1, 2, & 3 are now 4.8V? That sounds more reasonable, but what about pin 6? 1.7V doesn't sound right.

The op-amp in the LM10, takes the output voltage of the power supply, compares it to the reference and adjusts the voltage to Q5's bas which in tern alters the supply voltage to the oscillator driving the step-up transformer. It's an error amplifier.

Measurement errors aside, I suspect one of the transistors is faulty. The gain is probably lower than it should be, or it's leaking, so it still oscillates but doesn't reach the required output voltage. Have you tested any of them?
 

Offline crt_cowboyTopic starter

  • Regular Contributor
  • *
  • Posts: 71
  • Country: be
Re: Help repairing power supply
« Reply #7 on: January 18, 2025, 06:41:48 am »
I have tested them in circuit and did not find any short.
is there a way to check further Q3 and Q4 ? Oscillation frequency should be around 5KHz. I don't know if I can see more with an oscilloscope?
Q5 seems also fine (no short at least). On the emitter I find 2.6v which sounds a bit low.
 

Online Zero999

  • Super Contributor
  • ***
  • Posts: 20533
  • Country: gb
  • 0999
Re: Help repairing power supply
« Reply #8 on: January 18, 2025, 06:19:59 pm »
I have tested them in circuit and did not find any short.
is there a way to check further Q3 and Q4 ? Oscillation frequency should be around 5KHz. I don't know if I can see more with an oscilloscope?
Q5 seems also fine (no short at least). On the emitter I find 2.6v which sounds a bit low.
Unfortunately that won't tell you whether the transistors are working properly. One of them might be leaky or low gain and you wouldn't know.

This circuit can be used to test the hFE of a transistor. It will be equal to the ratio of the readings A2 & A2.

hFE = A2/A1
 
The following users thanked this post: watchmaker

Offline crt_cowboyTopic starter

  • Regular Contributor
  • *
  • Posts: 71
  • Country: be
Re: Help repairing power supply
« Reply #9 on: January 19, 2025, 06:52:21 am »
Thanks for the reply!
Would a failure of one of those transistors explain the low voltage on pin 6 of the LM10 ? I double checked and it varies between 1.6V and 1.7V.
For Q5, the 2.6V on the emitter seems low no ?


 

Online Zero999

  • Super Contributor
  • ***
  • Posts: 20533
  • Country: gb
  • 0999
Re: Help repairing power supply
« Reply #10 on: January 19, 2025, 11:50:46 am »
Thanks for the reply!
Would a failure of one of those transistors explain the low voltage on pin 6 of the LM10 ? I double checked and it varies between 1.6V and 1.7V.
For Q5, the 2.6V on the emitter seems low no ?
It depends.

What's the value of the base resistor?

If Q5's gain is low the excessive base current could pull the LM10's output low, assuming the base resistor's value is low enough to allow this.

Failing that, perhaps the LM10 is faulty.
 

Offline crt_cowboyTopic starter

  • Regular Contributor
  • *
  • Posts: 71
  • Country: be
Re: Help repairing power supply
« Reply #11 on: January 19, 2025, 12:38:08 pm »
Q5 is a 2N2222 and R38 is 220ohm.

On the base of Q5 I get 3.3 V.

I cleaned a bit more the leg of the LM10 and now I have 3.4V.

Should I assume the LM10 is working fine ? I am puzzled the emitter of Q5 only reads 2.6V
 

Offline watchmaker

  • Frequent Contributor
  • **
  • Posts: 675
  • Country: us
  • Self Study in EE
    • Precision Timepiece Restoration and Service
Re: Help repairing power supply
« Reply #12 on: January 19, 2025, 03:14:11 pm »
I have tested them in circuit and did not find any short.
is there a way to check further Q3 and Q4 ? Oscillation frequency should be around 5KHz. I don't know if I can see more with an oscilloscope?
Q5 seems also fine (no short at least). On the emitter I find 2.6v which sounds a bit low.
Unfortunately that won't tell you whether the transistors are working properly. One of them might be leaky or low gain and you wouldn't know.

This circuit can be used to test the hFE of a transistor. It will be equal to the ratio of the readings A2 & A2.

Thank you for that setup.  I just used it to select a transistor to match the gain of a transistor I pulled from an old Japanese radio that I cannot reinstall (it has those very early ribbon type leads).

Indeed, there is a great deal of variation within a transistor number.  Also, for the sake of others, it provides a different value than the Fnirsi.  I trust this method because it is so transparent.
Regards,

Dewey
 

Online Zero999

  • Super Contributor
  • ***
  • Posts: 20533
  • Country: gb
  • 0999
Re: Help repairing power supply
« Reply #13 on: January 19, 2025, 10:15:52 pm »
Q5 is a 2N2222 and R38 is 220ohm.

On the base of Q5 I get 3.3 V.

I cleaned a bit more the leg of the LM10 and now I have 3.4V.

Should I assume the LM10 is working fine ? I am puzzled the emitter of Q5 only reads 2.6V
The emitter voltage is 0.7V below the base voltage, which is to be expected.

Have you tried just powering the oscillator from a variable, current limited bench supply?
 

Offline crt_cowboyTopic starter

  • Regular Contributor
  • *
  • Posts: 71
  • Country: be
Re: Help repairing power supply
« Reply #14 on: January 20, 2025, 05:14:52 am »
Thanks for the suggestion. I will try that.
In practice, you mean injecting with the bench power supply somewhere after Q5 emitter to feed the oscillator? Do I need to disconnect anything from the regulation circuit?
 

Offline BILLPOD

  • Frequent Contributor
  • **
  • Posts: 400
  • Country: us
Re: Help repairing power supply
« Reply #15 on: January 20, 2025, 06:08:35 pm »
My eyes are old and don't always work very well, so would someone explain to me where Z4 is :-//
 

Online Zero999

  • Super Contributor
  • ***
  • Posts: 20533
  • Country: gb
  • 0999
Re: Help repairing power supply
« Reply #16 on: January 20, 2025, 06:21:49 pm »
My eyes are old and don't always work very well, so would someone explain to me where Z4 is :-//
The LM10. It was mentioned in the original post.

Thanks for the suggestion. I will try that.
In practice, you mean injecting with the bench power supply somewhere after Q5 emitter to feed the oscillator? Do I need to disconnect anything from the regulation circuit?
Yes, that was the idea. I was thinking that Q5's base-emitter diode will block current flow to the LM10, but I didn't consider that it might be faulty. I would recommend temporarily removing Q5 or at least disconnecting its emitter, before conducting this test.

The output must be unloaded and you should measure both the HV output and the input current. I would advise starting with a much lower voltage and slowly increasing it to 6V. You might have to suddenly apply the power to start the oscillator. It might not work, if you slowly increase the voltage from 0V. Set the voltage to something like 2V, then turn the power on, then increase by 0.5V if it doesn't oscillate. Repeat until the either the output voltage reaches 300V or you get to an input voltage of 6V.

If the oscillator works and all the transistors are good, then it could be one of the diodes on the output. Another possibility is a shorted turn in the transformer, which would result in the supply current being higher than it should be and it will heat up.
 

Offline crt_cowboyTopic starter

  • Regular Contributor
  • *
  • Posts: 71
  • Country: be
Re: Help repairing power supply
« Reply #17 on: January 21, 2025, 05:39:10 am »
 Thanks so I did as suggested.
When reaching 3.4 V on the bench power supply, I got the output voltage at the right value (9V and 300V).
Consumption was a little higher than it should 500mA while the expected should be 400mA. No heat that I could detect on the transformer (I don't have a thermal camera)
« Last Edit: January 21, 2025, 05:53:45 am by crt_cowboy »
 

Offline crt_cowboyTopic starter

  • Regular Contributor
  • *
  • Posts: 71
  • Country: be
Re: Help repairing power supply
« Reply #18 on: January 22, 2025, 09:44:40 am »
I am not sure what is the next step from here ? Is there something else I could do to further troubleshoot?
 

Offline Pfriemler

  • Regular Contributor
  • *
  • Posts: 159
  • Country: de
Re: Help repairing power supply
« Reply #19 on: January 22, 2025, 11:14:09 am »
3.x Volts sounds ok. Q5 did not deliver it by now.
R34 may (and should) not change the voltage on pin 8, but it should change pin 1 voltage in a wide range. Otherwise R34 or its internal connection is bad, so the resistor value does not change.
And pin 1 (reference output) and pin 3 (OPV + input) - or is it 5? - or in other words: both sides of R36 - should have (nearly) same voltage. Pin 3 cannot be negative.
Better you check all voltages on LM10 again.
« Last Edit: January 22, 2025, 05:53:50 pm by Pfriemler »
once you do it right, it works :-)
 

Online Zero999

  • Super Contributor
  • ***
  • Posts: 20533
  • Country: gb
  • 0999
Re: Help repairing power supply
« Reply #20 on: January 22, 2025, 07:13:25 pm »
Have you tested Q5? Disconnect pin 6 from the LM10, power the circuit from a 6V supply and vary the voltage on the base resistor. If the emitter sits around 0.8V to 1V below and you can get the oscillator working like this, then it means the  LM10 is faulty.
 

Offline Pfriemler

  • Regular Contributor
  • *
  • Posts: 159
  • Country: de
Re: Help repairing power supply
« Reply #21 on: January 22, 2025, 10:31:12 pm »
EDIT: ...I have redone the measures and for pin 1, 2 and 3 I now get 4.8v. I am not sure what is going on. Same voltmeter and same measurement procedure...
Ah .. I missed that. That answers one of my questions. These voltages should be the same if the circuit works in regulated condition. Nevertheless: If this voltage varies when R34 is changed, then the output voltages should vary, too.

once you do it right, it works :-)
 

Offline crt_cowboyTopic starter

  • Regular Contributor
  • *
  • Posts: 71
  • Country: be
Re: Help repairing power supply
« Reply #22 on: January 23, 2025, 12:20:47 pm »
I checked as suggested I injected voltage from the bench power supply to the base of Q5. Voltage drop between base and emitter was as expected.
I managed to get the oscillator working as well.
To get to the expected voltage, I had to push the voltage at the base of Q5 to 4.2 V.
Does it mean the LM10 is bad ?

« Last Edit: January 23, 2025, 12:55:49 pm by crt_cowboy »
 

Offline Pfriemler

  • Regular Contributor
  • *
  • Posts: 159
  • Country: de
Re: Help repairing power supply
« Reply #23 on: January 23, 2025, 12:30:06 pm »
LM10 regulates Q5 depending on the setpoint via R34 and the feedback from the 300 V rail (voltage from C17 via R29 to LM10 pin3). It's just guesswork if we don't know whether you can really vary the voltage at pin2 with R34 and how high the feedback voltage currently is if you got the right output voltages. If the feedback voltage is too high, the LM10 may regulate as it should, but with the wrong setting, and the setting range with R34 may not be sufficient. So for the third time: In what range does the voltage at pin2 change when you change R34 from one end to the other? Theoretically, it would probably only need a few tenths of a volt more at this point, but the question is whether this isn't masking another problem. (thx to DeepL :_))

edit: if the LM10 is bad, you should measure different pin2 and pin3 voltages. LM10 drives Q5 up until the two voltages are the same. If the output of LM10 is dead, there may be another path, so that the feedback voltage may feed the Q5 bypassing the LM10 via R37, but then I would expect a pin2/pin3 difference. And in this case, you may vary pin3's voltage by R34 without any change of the output.
« Last Edit: January 23, 2025, 12:41:35 pm by Pfriemler »
once you do it right, it works :-)
 

Offline crt_cowboyTopic starter

  • Regular Contributor
  • *
  • Posts: 71
  • Country: be
Re: Help repairing power supply
« Reply #24 on: January 23, 2025, 01:18:13 pm »
Here is what happens when I adjust R34:
- Pin 8 varies from 100mV to 200mV
- Pin 6 varies from 2.6V to 3.4V

Edit: confirming that pin 2 and 3 have exactly the same voltage.
« Last Edit: January 23, 2025, 01:34:58 pm by crt_cowboy »
 

Offline Pfriemler

  • Regular Contributor
  • *
  • Posts: 159
  • Country: de
Re: Help repairing power supply
« Reply #25 on: January 23, 2025, 01:42:24 pm »
OK. pin8 voltage should not vary in a wide range of R34, as @Zero999 mentioned earlier, as the reference OPV in LM10 should push it via R35. Only if pin1 is saturated at +6V, you may pull it further downwards.
But I asked for the pin1 variety. If it varies up to 6V (lower end is defined by the ratio of R34 and R35), then this part is ok.
Now: As long as pin3 is only a bit higher than pin2, LM10s ouput 6 should drive up to 6V. But it does not. You mentioned a 0.1v difference on R38, that's 0.5 mA base current to Q5. LM10 should deliver it.
Please report how pin1, pin2 and pin3 changes if you turn R34 into both directions.
once you do it right, it works :-)
 

Offline crt_cowboyTopic starter

  • Regular Contributor
  • *
  • Posts: 71
  • Country: be
Re: Help repairing power supply
« Reply #26 on: January 24, 2025, 08:11:36 am »
Thanks for the reply. Here is what I observe:

R34 low (0 ohm):
- pin 1: 3.9V
- pin 2: 3.9V
- pin 3: 3.9V

R34 High (1kohm):
- pin 1: 568mV
- pin 2: 568mV
- pin 3: 568mV
 

Offline Pfriemler

  • Regular Contributor
  • *
  • Posts: 159
  • Country: de
Re: Help repairing power supply
« Reply #27 on: January 24, 2025, 01:09:40 pm »
No difference between 1,2 and 3 - as it should if the LM10 works fine.

But I wonder why only 3.9V is reached now:

I have redone the measures and for pin 1, 2 and 3 I now get 4.8v. I am not sure what is going on. Same voltmeter and same measurement procedure...
There's no reason why 4.8v and more are not possible, except of 0 ohm isn't 0 Ohm. For R34=1k and 200mV at this point, R35 calculates to ~1,9k. Given this, the low end of R34 rather comes to ~105 Ohms. If you can read them, pls give us a few more values of R28,R29, R35..R37

Ah, I forgot: You put Q5 back in its place? You definitely should see an output voltage variation on the 300v. If you can reach, pickup the span on C17, too. If you got the 3.9 / 0.568mV without the inverter working, repeat measuring them, pls.
What are thes curious CR11-CR13?. Are C14-C16 ok (no leakage currents)? It's difficult to measure, I know, but if you can, check the voltages in this divider, too, and report. Still my idea is that the feedback voltage to the LM10 is too high.
And/or if the sliding contact inside R34 is poor.

Weird.


« Last Edit: January 24, 2025, 01:14:39 pm by Pfriemler »
once you do it right, it works :-)
 

Offline crt_cowboyTopic starter

  • Regular Contributor
  • *
  • Posts: 71
  • Country: be
Re: Help repairing power supply
« Reply #28 on: January 24, 2025, 02:17:19 pm »
Ok so yes Q5 was in place for the measures I provided.

Here are the resistors values:
R28: 100K
R29: 47K
R35: 2K (your guess was correct)
R36: 47
R37: 3.9M

I checked and R34 varies between 0 and 1kohm.

C17 when R34 is 0 measures 3.9V and when R34 is 1K measures 0.5V

CR11-CR13 are Zener diodes.

CR11 : 275V
CR12 : 184V
CR13: 92V
(Measurements made on cathode)

Sliding of R34 seems ok. I checked at different values between the min and max.
« Last Edit: January 24, 2025, 02:48:56 pm by crt_cowboy »
 

Offline Pfriemler

  • Regular Contributor
  • *
  • Posts: 159
  • Country: de
Re: Help repairing power supply
« Reply #29 on: January 24, 2025, 08:26:26 pm »
Good and normal values. We can forget the theory that Q5 is feeded bypassing the LM10 through R37 - not with 3.9M.
No other suspects.

So, let's go to the basics.
First: just for sure: The 300V output is bufferd by C7. Am I right that you do not have any oscilloscope equipment to check the 300V for any ripple? If not, try to measure the 300V ouput with a DMM in pure AC mode. You should not get a value of more than a few volts. If you have a ripple of 20-50v, there's maybe an issue with C7, C17 should buffer the (by Zeners) downscaled 300v output peaks, so the oscillator is regulated not to exceed 3xx volts. But with a significant ripple you will meausre a lower DC average on your DMM, and the output lowers more with any load than usual.
Note: When measuring in AC mode ... If you have an idea which frequency the oscillator works ... this could lead to false interpretations of AC readings at this point. I just checked my DMMs for that and got weired results with older (non rms) DMM, whereas two new OWON HDS120 and HDS242s are blind with frequencies >10 kHz. If you have any sine oscillator (AWG etc), check your meter for accuracy with various frequencys (if not done already, it's always good to know).

Second: The Zener voltages are 91v, 92v and something between 88 and 91.5v, depending on the R34 position. Given you measured while R34 was at the low end (and C17 voltage 3.9v) all three Zeners are equal, otherwise there may be a problem with CR13. But I do not think so.

Third: Did you find any output voltage variation when turning R34 from 0 to 1k, and if, what variations of Q5 base or emitter voltage you see? If you find the output variation of let's say 20v and 0.X v at Q5, this may be normal.

Fourth: If you can vary the output with R34 and your are shure that C7 is ok and there's no significant ripple on the 300v output (a few volts are ok), and only then!, try to lower R28 a bit with a bypass resistor (start with ~ 5MegOhm in parallel). 3.9 volts on 100k result in a 0.039 mA current through them, resulting in ~12 mW total power dissipating, if I do not fuck the maths. With a 1MOhm in parallel, current and power dissipating rises by 10%, should be safe. 0.034 mA (or 4 mW) are not much for a Zener (rather 100mW are), and in this area they have a rather soft U/I characteristics, with only a bit more current you should get up to 100 volts per Zener and thus more output voltage. Do we have a spec for the Zeners? I do not think they are 91volts, rather 100 oder 110 volts.
Start the circuit with R34 on the 1k side and try to rise the output voltage.
Good luck!
once you do it right, it works :-)
 

Offline crt_cowboyTopic starter

  • Regular Contributor
  • *
  • Posts: 71
  • Country: be
Re: Help repairing power supply
« Reply #30 on: January 25, 2025, 01:24:33 pm »
Thanks for the very detailed reply.

1 - Yes I do have an oscilloscope. On the 300 V line I see very little ripple. Around 1.2 VPP.

2 - Zener diodes are BZX46C 100V

3 - R34 = 1K then HV measures 281V, Q5 base 3.1V, emitter 2.4V.
    R34 = 0 ohm the HV measures 285V, Q5 base 3.4V, emitter 2.7V

4 - as there is not so much change on the 300V line when changing R34 I am not sure I should proceed.
   
 

Offline Pfriemler

  • Regular Contributor
  • *
  • Posts: 159
  • Country: de
Re: Help repairing power supply
« Reply #31 on: January 25, 2025, 01:54:39 pm »
You’re right, the voltage variation of 4 Volts is a bit narrow. But: It still looks to me that the Zeners are not loaded enough. I would have a closer look to R28 again. Is it 100k, really?
Try to vary it, as I mentioned in the ost before.
once you do it right, it works :-)
 

Offline crt_cowboyTopic starter

  • Regular Contributor
  • *
  • Posts: 71
  • Country: be
Re: Help repairing power supply
« Reply #32 on: January 25, 2025, 02:11:58 pm »
Yes R28 measures 100k.
I will try as suggested and report the results
 

Offline crt_cowboyTopic starter

  • Regular Contributor
  • *
  • Posts: 71
  • Country: be
Re: Help repairing power supply
« Reply #33 on: January 25, 2025, 03:51:14 pm »
Just did a quick test of the R28 bypass with resistors from 5M to 1M. Strangely it made no difference in the output voltage.
 

Offline Pfriemler

  • Regular Contributor
  • *
  • Posts: 159
  • Country: de
Re: Help repairing power supply
« Reply #34 on: January 25, 2025, 10:08:10 pm »
A fault in C17 should rather lead to higher than lower voltages as intended.
Damned. I hate to say it, but I have no more ideas. Sorry!
once you do it right, it works :-)
 

Offline crt_cowboyTopic starter

  • Regular Contributor
  • *
  • Posts: 71
  • Country: be
Re: Help repairing power supply
« Reply #35 on: January 26, 2025, 07:28:54 am »
No problem thanks for all the advices.
There are not so many components involved but as we are facing a partial failure it is hard to find unfortunately.
 

Offline crt_cowboyTopic starter

  • Regular Contributor
  • *
  • Posts: 71
  • Country: be
Re: Help repairing power supply
« Reply #36 on: January 26, 2025, 08:38:41 am »
Here is what I observe on the base of Q3. Shouldn't it be 5KHz as this is the frequency it is supposed to run on ?
« Last Edit: January 26, 2025, 08:40:23 am by crt_cowboy »
 

Offline crt_cowboyTopic starter

  • Regular Contributor
  • *
  • Posts: 71
  • Country: be
Re: Help repairing power supply
« Reply #37 on: January 26, 2025, 04:09:30 pm »
Can my problem be related to the fact it doesn't oscillate at the right frequency? It is supposed to have 5KHz and I observe 4KHz. Thanks
 

Online Zero999

  • Super Contributor
  • ***
  • Posts: 20533
  • Country: gb
  • 0999
Re: Help repairing power supply
« Reply #38 on: January 26, 2025, 05:51:25 pm »
How do you know what frequency it's supposed to work at? 4kHz is a little low, but it wouldn't surprise me. My experience of this type of circuit is the frequency varies with the load.
 

Offline Pfriemler

  • Regular Contributor
  • *
  • Posts: 159
  • Country: de
Re: Help repairing power supply
« Reply #39 on: January 26, 2025, 06:32:07 pm »
Can my problem be related to the fact it doesn't oscillate at the right frequency? It is supposed to have 5KHz and I observe 4KHz. Thanks
Normally not. Lower frequency may result in a lower energy transport (simply spoken), but the feedback circuitry is DC and this should result in a higher supply voltage to the oscillator via Q5. Still I am wondering why changing the R34 has such a low effect to the output voltage and why it is not possible to get >5V in the feedback regulation. Any ideas to this, Zero999?
once you do it right, it works :-)
 

Offline crt_cowboyTopic starter

  • Regular Contributor
  • *
  • Posts: 71
  • Country: be
Re: Help repairing power supply
« Reply #40 on: January 27, 2025, 05:22:44 am »
How do you know what frequency it's supposed to work at?
It is written in the manufacturer specifications. So we can assume that 4KHz is indeed too low.
 

Online Zero999

  • Super Contributor
  • ***
  • Posts: 20533
  • Country: gb
  • 0999
Re: Help repairing power supply
« Reply #41 on: January 27, 2025, 11:48:09 am »
How do you know what frequency it's supposed to work at?
It is written in the manufacturer specifications. So we can assume that 4KHz is indeed too low.
Not really. The typical frequency is 5kHz which will vary depending on the load and the tolerance of some components, such as capacitors, can be as much as 20%, I wouldn't say that it being 4kHz means it's definitely faulty. It's on the low side, but not necessarily a cause for concern.

To make things easier, annotate the schematic with component values and the measured voltages. We should then be able to work out what the voltages should be.
 

Offline crt_cowboyTopic starter

  • Regular Contributor
  • *
  • Posts: 71
  • Country: be
Re: Help repairing power supply
« Reply #42 on: January 27, 2025, 12:52:14 pm »
Here is the annotated circuit diagram. Measures taken when R34 equals 0 ohm
 

Offline crt_cowboyTopic starter

  • Regular Contributor
  • *
  • Posts: 71
  • Country: be
Re: Help repairing power supply
« Reply #43 on: January 27, 2025, 01:11:04 pm »
And here is he version when R34 equals 1K.
« Last Edit: January 27, 2025, 01:15:24 pm by crt_cowboy »
 

Offline Pfriemler

  • Regular Contributor
  • *
  • Posts: 159
  • Country: de
Re: Help repairing power supply
« Reply #44 on: January 27, 2025, 04:03:55 pm »
That's interesting news and different values!

R29: 47K
Now 47 Ohm? what's right?

Quote
C17 when R34 is 0 measures 3.9V and when R34 is 1K measures 0.5V

That's completely different to your pics.
If in R34=0 C17 has 3.97v and pin2@LM10 3.9v, that may be ok with a R29=47k(!) but not with 47 ohms, and with R34=1k the voltage @CR13 is now only 88v instead of 92v.
Regarding the feedback voltage "divider", 3.97/279v and 3.46/275v sounds reasonable, but the voltages @CR11 and @CR12 should be the same as @CR13.

And neither with R29=47R nor R29=47k the right part of LM10 with its feedback of R37=3.9M should be able to lower the voltage on pin2 down to 590mV.

If you are sure you got good connections to the measuring points, check all connections around R29.



once you do it right, it works :-)
 

Offline crt_cowboyTopic starter

  • Regular Contributor
  • *
  • Posts: 71
  • Country: be
Re: Help repairing power supply
« Reply #45 on: January 27, 2025, 04:13:10 pm »
My bad for R29. Sorry. I confirm 47 ohm is correct.

EDIT: I quickly checked between C17 and the pin 2 of LM10 and I measure 47 ohm.

I will take check again this area and re-do measurements to confirm.
« Last Edit: January 27, 2025, 05:14:56 pm by crt_cowboy »
 

Offline Pfriemler

  • Regular Contributor
  • *
  • Posts: 159
  • Country: de
Re: Help repairing power supply
« Reply #46 on: January 27, 2025, 05:20:45 pm »
Rather good: For the 1k-R34: 3.46 -> R29 -> 0.59V would be a current of 62 mA, if the connections are good. No such current can come from the CR11..CR13 (they would be blown up from heat), nor there's a sink at the LM10 - only if the LM10 input pin2 is defective, rather unlikely if the other R34's end with 3.9v works.
Heat dissipation of R29 should then be 180 mW, that would make it considerably warm.
There must be a failure. Solder joint, hairline crack. Check for continuity.

EDIT: ah, you did. Makes no sense until now. We'll see..
« Last Edit: January 27, 2025, 05:25:23 pm by Pfriemler »
once you do it right, it works :-)
 

Offline crt_cowboyTopic starter

  • Regular Contributor
  • *
  • Posts: 71
  • Country: be
Re: Help repairing power supply
« Reply #47 on: January 27, 2025, 06:05:28 pm »
R28 measures 47K in circuit and not 100K. I will desolder it tomorrow to confirm. Probably a measurement issue with the in-circuit measure.
 

Offline Smile2016

  • Regular Contributor
  • *
  • Posts: 58
  • Country: us
Re: Help repairing power supply
« Reply #48 on: January 27, 2025, 06:34:25 pm »
Are CR11, CR12, and CR13 identical in value? If they are, and the measured voltages are accurate, CR13 seems a bit lower at around 88.5V. Can you check the Zener values or markings on them? Considering they are biased at only a few tenths of microamperes, the circuit is sensitive to small leakages, CR13?. I would replace CR13, [&the capacitor C17, and Q5]. Standard values are 91V...and next up is 100V. 91V won't get you to 300V output...Also check if battery is still 6V when Pot is at 0 Ohms...

I just noticed you mentioned, "2 - Zener diodes are BZX46C 100V." Probably these diodes were pre-selected to have a Zener voltage (Vz) of around 100V at ~ 40µA, they have an initial spread of 94–106V according to the datasheet. Probably your PSU works just fine except other than Vz is not tailored to give you exactly 300V. again... make sure battery holds 6V
« Last Edit: January 28, 2025, 04:31:14 am by Smile2016 »
 

Offline crt_cowboyTopic starter

  • Regular Contributor
  • *
  • Posts: 71
  • Country: be
Re: Help repairing power supply
« Reply #49 on: January 28, 2025, 12:19:02 pm »
Yes but as the 300v line is low the two other 9v lines are also low as the regulation relies on the 300v line.

Removed R28 and it has the right value when measured alone.

 

Online Zero999

  • Super Contributor
  • ***
  • Posts: 20533
  • Country: gb
  • 0999
Re: Help repairing power supply
« Reply #50 on: January 28, 2025, 12:25:33 pm »
Then Q5 must be faulty. The transistor has partially failed. The hFE has gone down, so it draws too much base current for it to work with R28 in place. Even if you can get it to work by omitting the base resistor, I still recommend replacing Q5.
 

Offline crt_cowboyTopic starter

  • Regular Contributor
  • *
  • Posts: 71
  • Country: be
Re: Help repairing power supply
« Reply #51 on: January 28, 2025, 12:30:40 pm »
Thanks, makes sense. I will try to find a suitable replacement and post an update here after replacing it.
 

Offline Smile2016

  • Regular Contributor
  • *
  • Posts: 58
  • Country: us
Re: Help repairing power supply
« Reply #52 on: January 28, 2025, 01:30:20 pm »
can you add for a quick test a 27V zenner in series with one of CR11..CR13 to see if everything else works right and get to 300V?
 

Offline crt_cowboyTopic starter

  • Regular Contributor
  • *
  • Posts: 71
  • Country: be
Re: Help repairing power supply
« Reply #53 on: January 28, 2025, 02:20:21 pm »
I will check if I have a Zener in stock to do the test.
When I was driving base Q5 manually I could reach 300v
 

Offline Pfriemler

  • Regular Contributor
  • *
  • Posts: 159
  • Country: de
Re: Help repairing power supply
« Reply #54 on: January 28, 2025, 05:21:37 pm »
Then Q5 must be faulty. The transistor has partially failed. The hFE has gone down, so it draws too much base current for it to work with R28 in place. Even if you can get it to work by omitting the base resistor, I still recommend replacing Q5.
Looking at the 0k-pic, I read 3.4v from LM10 pin7 and 3.37 at base of Q5, that is 0,13mA through R38. Far away from "too much".

Can anyone else explain how the 3.46v @ C17 may drop to 0.59v at pin2@LM with a R29=47R between it?
once you do it right, it works :-)
 

Offline crt_cowboyTopic starter

  • Regular Contributor
  • *
  • Posts: 71
  • Country: be
Re: Help repairing power supply
« Reply #55 on: January 28, 2025, 05:35:52 pm »
Quote
Can anyone else explain how the 3.46v @ C17 may drop to 0.59v at pin2@LM with a R29=47R between it?
I will redo the measurement either I screwup or there is an intermittent bad contact somewhere.

EDIT: confirming that my measures were bad.
« Last Edit: January 29, 2025, 06:48:35 am by crt_cowboy »
 

Offline Smile2016

  • Regular Contributor
  • *
  • Posts: 58
  • Country: us
Re: Help repairing power supply
« Reply #56 on: January 28, 2025, 09:25:27 pm »
I will check if I have a Zener in stock to do the test.
When I was driving base Q5 manually I could reach 300v
When you manually drove Q5, did you happen to check the voltage across R28/C17?  I noticed that Z3 pin 3 never exceeded 4V in any of the scenarios shown here. Z3.3 fails to go higher only under one of the following conditions:
    It’s limited by the 6V power supply (e.g., the supply voltage drops when required to deliver more current).--> have no values to confirm this
    The servo/feedback system dictates it. --> I'm leaning towards this , specifically the zenners
    Z3 is partially defective.--> unlikely [Z3 should be capable driving +-20mA and +-0.4V away from rails]

You mentioned that a 4.2V input at Q5's base results in 300V being present.  This also implies that Q5's emitter voltage would be approximately 3.4–3.5V. At the measured 0.5A, Q5 would dissipate around 1.3W, with the remaining 1.7W dissipated across the rest of the circuit. While older TO-39 cases can handle this level of continuous dissipation, it’s still excessive for Q5. Q5 might have intermittent failures or h21 decay as others mentioned here, since dissipates a lot in a continuous mode common for older gen medium power transistors failure modes—another reason I suggested replacing it. the reason I suggested to check the zenners is because of this extra 100mA sharp current jump  adjusting 275V to 300V which translates into a power jump from 1.1W to 1.7W . If you don't have a zenner in house you can still repeat the experiment with 4.2V on the base and checking the feedback voltage on R28 while 300V output. that voltage has to fall in Z3's reference voltage range set on pin1 , let's say with R43 at the mid point [1V on Z3 pin1]. !caution here higher voltage might be applied to Z3 trough R29 and if divider string is malfunctioning. If your voltage on R28 goes higher than 5..5.8V while 300V out then you know where the issue is.
« Last Edit: January 28, 2025, 09:34:43 pm by Smile2016 »
 

Offline crt_cowboyTopic starter

  • Regular Contributor
  • *
  • Posts: 71
  • Country: be
Re: Help repairing power supply
« Reply #57 on: January 29, 2025, 06:53:04 am »
Just checked as suggested the voltage at R28/C17 when applying voltage directly to the base of Q5.
It goes up to 10v and I was not yet at 300v...
Does it mean the Zener diodes have to be replaced?
« Last Edit: January 29, 2025, 07:07:21 am by crt_cowboy »
 

Offline Pfriemler

  • Regular Contributor
  • *
  • Posts: 159
  • Country: de
Re: Help repairing power supply
« Reply #58 on: January 29, 2025, 09:35:53 am »
Wow. I imagine the 10v going through R29 into LM10's input, driven with 6V. That's unsane for it, normally it must not exceed UB +.7v. otherweise the LM10 may die or go defective.
But it proofes that's there an issue with the HV feedback circuit.
What the heck happened to that device?  ???
once you do it right, it works :-)
 

Offline crt_cowboyTopic starter

  • Regular Contributor
  • *
  • Posts: 71
  • Country: be
Re: Help repairing power supply
« Reply #59 on: January 29, 2025, 10:04:50 am »
Yes I will replace the Zener diodes. Not sure what is a suitable replacement for BZX46C 100V ? I suppose any Zener 0.5 Watt would work ?
« Last Edit: January 29, 2025, 11:04:46 am by crt_cowboy »
 

Offline Smile2016

  • Regular Contributor
  • *
  • Posts: 58
  • Country: us
Re: Help repairing power supply
« Reply #60 on: January 29, 2025, 03:03:41 pm »
can you add for a quick test a 27V zenner in series with one of CR11..CR13 to see if everything else works right and get to 300V?

 For precise accuracy, have to pre-select 3 of them to total 300V and at 10...40µA. Standard high-voltage zeners can vary widely in tolerance or become costly with tighter tolerances. pick lower wattage, those are more accurate at such low currents. you'll dissipate less than 50mW in normal operation. Any DO35 500mW zener diode should work. the zenners has not to be perfectle equal... a 100V will work with a 110V plus a 92V as long the total is around your targeted voltage [@ 10..40uA]
 

Offline Smile2016

  • Regular Contributor
  • *
  • Posts: 58
  • Country: us
Re: Help repairing power supply
« Reply #61 on: January 29, 2025, 03:11:13 pm »
My bad for R29. Sorry. I confirm 47 ohm is correct.

EDIT: I quickly checked between C17 and the pin 2 of LM10 and I measure 47 ohm.

I will take check again this area and re-do measurements to confirm.

R36=R29 as for older op-amps :)  identical resistors to maintain a balanced impedance,might help minimize input offset voltage and errors caused by different bias currents flowing into the non-inverting and inverting inputs , even original designer might have missed something or had just 47 OHms on hand.
 

Offline Smile2016

  • Regular Contributor
  • *
  • Posts: 58
  • Country: us
Re: Help repairing power supply
« Reply #62 on: January 29, 2025, 03:18:59 pm »
Wow. I imagine the 10v going through R29 into LM10's input, driven with 6V. That's unsane for it, normally it must not exceed UB +.7v. otherweise the LM10 may die or go defective.
But it proofes that's there an issue with the HV feedback circuit.
What the heck happened to that device?  ???

got lucky , beefy LM10 ! the input voltage can exceed the supply voltages provided that the voltage from the input to any other terminal does not exceed the
maximum differential input voltage ... that's +-40V
 

Offline Smile2016

  • Regular Contributor
  • *
  • Posts: 58
  • Country: us
Re: Help repairing power supply
« Reply #63 on: January 29, 2025, 03:49:10 pm »

What the heck happened to that device?  ???
[/quote]

a shift in the breakdown voltage, caused by long-term degradation of the semiconductor material or by overheating [unlikely]??  . zenner start conducting too early seems...
 

Offline crt_cowboyTopic starter

  • Regular Contributor
  • *
  • Posts: 71
  • Country: be
Re: Help repairing power supply
« Reply #64 on: January 29, 2025, 04:45:05 pm »
can you add for a quick test a 27V zenner in series with one of CR11..CR13 to see if everything else works right and get to 300V?

 For precise accuracy, have to pre-select 3 of them to total 300V and at 10...40µA. Standard high-voltage zeners can vary widely in tolerance or become costly with tighter tolerances. pick lower wattage, those are more accurate at such low currents. you'll dissipate less than 50mW in normal operation. Any DO35 500mW zener diode should work. the zenners has not to be perfectle equal... a 100V will work with a 110V plus a 92V as long the total is around your targeted voltage [@ 10..40uA]

Thank you very much. I have ordered more than needed to be able to select ones that are suitable. I will post progress once I have replaced them. I plan to replace all 3 Zener.
« Last Edit: January 30, 2025, 06:48:21 am by crt_cowboy »
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf