Chinese Hiland 0-30V/Audioguru PSU build (down the rabbit hole)

[bushtech]

Will be ordering some new 2N3055's
But my old 3055's confuse me: Using an analogue MM black probe on common, red on base I get no connectivity. Swing the probes around and I get connectivity.
Using DMM I get OL on meter both ways
Is the high voltage of the analogue meter causing a breakthrough somehow?

 

[Audioguru]

A transistor does not have a "common", it has a base-collector diode and a base-emitter diode. It should not conduct collector to emitter when its base is not connected. When the base-emitter diode is reverse-biased with about 6V or more it conducts like a zener diode which damages the junction.
My multimeter uses only 0.2V to measure resistance so that a silicon transistor junction does not conduct. Then on "diode test" it uses 3V so that a silicon transistor junction and many LEDs will conduct when forward biased.

 

[bushtech]

Thanks Audioguru. So could I have damaged them with the analogue MM.? I think they push 9V.

When I switch on the power supply, the light in series goes bright and off which is good. But then it starts with a dim erratic flickering which I can also see faintly on the onboard led. Does this signify something?

 

[bushtech]

You wrote:
The power supply output was 41V because one or both output transistors are shorted from collector to emitter.

When I test them with DMM on continuity I don't get OL but 558 on both

 

[Audioguru]

If the light in series flickers then something is wrong. I do not know which of your many problems is causing it.

If a multimeter has more than about 5V in the continuity test and its red lead is on the emitter and its black lead is on the collector then the emitter-base junction is reverse biased through the series forward biased base-collector junction and it normally conducts like a zener diode. Its current might be low and for a short duration so it might do no harm. The red lead should be on the collector and the black lead should be on the emitter for an NPN transistor which is how it is in a circuit. With the base disconnected it should measure no current.

 

[bushtech]

Have replaced D5, D6 and 2 x D7. 2 Resistors check out ok and 2 caps on order
U2:
1. 0V
2. 14V
3. 0.1V
4. 0V
5. 0.1.V
6. 18V
7. 40V
8. 0V

U3(overheating)
1. 4.4V
2. 3.2V
3. 2.9V
4. 3.8V
5. 4.4V
6. 4.2V
7. 30.6V
8. 0V

If I remove U3 pin 4 of U3 goes to 0.01V. Cannot for the life of me get it to -1.3V

 

[Audioguru]

D7 is two forward biased diodes, not a reverse biased zener diode.
Here is how -1.3V is made:

 

[bushtech]

Thank you for your help Audioguru.

We know that problem areas are pin 4 of U3 which should be -1.3V. I have also replaced Q3, the BC557, so I am hoping the problem lies with C2 or C3 (waiting for replacements) and pin 7 of U3 which voltage is too high.

How are the other voltages on U2 and U3?. Do I have to concentrate on the 2 problems above?

Some observations:

Pin 3 of U2 goes from 0 to 4V when turning Pi (with or without U2 removed)
Pin 6 of U2 stays at 18V when P1 turned

The 1N4740 zener has 10V across it.

BD139:
Base: 40.8V
Emitter: 18V

 

[Audioguru]

pin 7 of U3 which voltage is too high.

no, its voltage is 10V less than the +40V positive so the 10V zener diode is doing its job.

How are the other voltages on U2 and U3?

I will look at them later.

Pin 3 of U2 goes from 0 to 4V when turning Pi (with or without U2 removed)

That is wrong. The output of U1 is +11/2V and feeds pin 3 of U2 though R8 and R9 which total only 29.2k so pin 3 of U2 should go from 0V to +11.2V. Maybe your voltmeter is an old low input resistance analog one instead of a modern digital one with an extremely high input resistance?

Pin 6 of U2 stays at 18V when P1 turned

During this test the BD139 is disconnected from pin 6 and R12 and C6 are disconnected from the output and connected to pin 6 instead then the opamp U2 should have a gain of 1+ [R12/(R11+RV2)]= about 30V/11.2V so that when the voltage pot is turned the output of U2 goes from 0V to 30V.

The 1N4740 zener has 10V across it.

That is good.

BD139:
Base: 40.8V
Emitter: 18V

Impossible. It is an NPN so its base to emitter voltage is 0.7V with a low current or 1V maximum with a high current, yours is 22.8V. The BD139 is destroyed or has its pins connected wrong (or both).

Oh, pin 6 of U2 is 18v and does not change with the voltage pot AND the emitter of the BD139 is also 18V. Then you wrongly connected U2 to the emitter instead of to the base?
Oh again, the base of the BD139 is 40.8V so maybe it is wrongly connected to +40.8V instead of to pin 6 of U2? Look at the datasheet of the BD139 to see which pin is which, the Chinese transistor (and the Chinese pcb) probably has a different pins layout.

 

[bushtech]

Thanks a mil again Audioguru.
Just 2 notes:
"Pin 6 of U2 stays at 18V when P1 turned
During this test the BD139 is disconnected from pin 6 and R12 and C6 are disconnected from the output and connected to pin 6 instead then the opamp U2 should have a gain of 1+ [R12/(R11+RV2)]= about 30V/11.2V so that when the voltage pot is turned the output of U2 goes from 0V to 30V."

Currently the power supply is fully connected

"Maybe your voltmeter is an old low input resistance analog one"

I have both, for these readings I used my DMM

Please note that all readings taken are relative to the neg pole of C1, the big cap.

When you say: It is an NPN so its base to emitter voltage is 0.7V. Does that mean I have to take a reading with one probe on base and other probe on emitter?

Now to trace the BD139 connections

 

[bushtech]

Just to confirm:

My BD139:


With the legs facing down the connections from left to right: E C B

 

[bushtech]

Just seen I've made an ooopsie reading BD139

The corrected readings are:
E 28V
C 40V
B 18V

 

[Audioguru]

Something wrong is causing the emitter voltage of the BD139 to be too high. First I wanted to see if U2 works properly by disconnecting the BD139 and connecting the negative feedback from R12 and C6 to U2 pin 6 without the output transistors. You measured not enough voltage swing at the input of U2 when U2 IC is removed which is another problem, but only the voltage pot P1, R8, C4, D9 and R9 can cause this problem since the output of U1 is correctly at +11.2V that feeds the voltage pot.

Your pins are correct for the BD139 but the emitter voltage is 10V higher than its base when it should be a little less than its base. Your BD139 has a reverse emitter-base voltage of 10V when the datasheet says its maximum allowed reverse voltage is 5V so it might be destroyed. Try another BD139 but without the output transistors which might be causing the BD139 emitter voltage to be too high.

I do not see a manufacturer's logo on your BD139 (I expect to see Philips or NXP) so it might be a cheap Chinese fake.

 

[bushtech]

This is the BD139's I ordered: http://www.mantech.co.za/Stock.aspx?Query=35M6218+and

Checked BD139 again and base to emitter voltage 9.6V. Is the 5V you are talking about Vebo on the datasheet. If so, then yeah, probably history. I have 1 more BD139 but I will now wait for the replacement caps and 2N3055s

 

[bushtech]

Some good news. I replaced the Chinese P1 pot with a decent pot and Pin 3 of U2 now goes from 0 to 11V when P1 adjusted

 

[bushtech]

BD139 readings with 2 output transistors disconnected:
E 11.3V
C 41V
B 1.7V

B to E = 9.6V
E to C = 5.8V
B to C = 0.1V

 

[Black Smock]

I don't know whether I should start a new thread for this or not. I purchased two of these kits from Banggood recently, with a view to using them to replace the board in my Farnell L30 PSU which has suddenly failed. It's years since I did much fault-finding to component level and the Farnell circuit uses ancient components, such as OC44, OC24 & OC26. It has a mixture of pnp, npn, silicon and germanium transistors. It has a nice metering system and the pots have the same values as the Hiland circuit.
I propose to disconnect the +ve end of the bridge rectifier from the board and introduce an LM317 regulator circuit to drop the 45V or so from the Farnell transformer to around 30V dc for the remainder of the board. I would leave C1 & R1 at the output of the LM317 regulator. I only need 500mA maximum current to match the Farnell unit's current-handling capability and as I rarely need voltages above 20V, I could reduce the LM317s output voltage sufficiently to make replacement of the TL081s unnecessary.
What do you think?

 

[bushtech]

Black Smock: This build is about a highly modified version of the chinese product using the chinese pcb. We are pretty far down the rabbit hole, so I suggest it would be better to start your own project.

And welcome to EP

 

[bushtech]

Is BD139's E too high or is B too low? Just so I know where to start looking

 

[Audioguru]

BD139 readings with 2 output transistors disconnected:
E 11.3V
C 41V
B 1.7V

B to E = 9.6V
E to C = 5.8V
B to C = 0.1V

This is an NPN transistor and in this circuit the emitter must NEVER higher than the base voltage. Since your circuit does not have a load then maybe the BD139 leaks a little current from its 41V collector that is causing the emitter voltage to rise too high.
I do not understand how you measured the emitter at 11.3V and collector at 41V but then say E to C = 5.8V; and say the base at 1.7V and the collector at 41V then say B to C = 0.1V.

As a test, make sure that there is no short or anything from the emitter of the BD139 to the collector on the pcb and the 2N3055 transistors are removed. Install the spare BD139 that you have and add a 10k testing resistor to the output. Connect it and measure it like this: