Problem with capacitance multiplier power supply

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I've built a capacitance multiplier power supply for testing/learning/measuring and Its not quite working right.

The circuit is the from here and looks like this:

An externally hosted image should be here but it was not working when we last tested it.


I'm getting 58.3 volts across the rectifier, 29.1V across the filter caps. From Gnd to -Ve I get
-26.8V, but from gnd to +Ve I only get 8.1v. I have tried the circuit with 2N3055's at Q3 & Q4 and with MJ15003's but the voltages are similar

The pages says I should adjust VR for 3v across the pass transistors - I get 21V across Q3 and 2.2V across Q4

I suspect this circuit was designed with a lower voltage secondary and that the resistor values will need to be adjusted to reflect this. I'm hoping for some guidance before I dive in. I have been looking at this page as a reference as well.

Cheers
 
Hi

I have built capacitance multipliers that are similar to those in the first circuit that Rod Elliot shows and they work.

I have not tried the layout shown in figure 2. However in figure 2 the relationship between Q2/Q4 the pass transistor does not seem right.

Don
 
Samuel: I'll measure the voltage at the base of Q1 and its emitter tonight,

h_a: ahh, ya I guess I was testing without a load. I do have some 5, 10, and 20 ohm 50W resistors. When you say it's not going to work without a load, can you be more specific? I do measure 29.1 volts at the -Ve side which is about what I would expect. It's just the +Ve side that I'm having a problem with.

mjf: I have checked the wiring about 6 different times, on 6 different days - I'm reasonably certain that the wiring is correct. On both sides of the circuit I have the 6.8k resistor going to the 5k pot. I also verified that the wiper of the pot is connected correctly on both sides.

on the outside chance that I did something to the BD139 I replace it as well but the voltages do not change. I also swapped the pass transistors to see if the problem followed them but it did not.
 
Ok, I have attached the schematic with some voltages on it (without load for now until I find my big resistors). I think I may try build the first circuit on that page just for comparison to see if the results are different.
 

Attachments

  • jlhcapmultfig2-with-voltages.jpg
    jlhcapmultfig2-with-voltages.jpg
    56 KB · Views: 365
The 20.8 V across Q4 doesn’t make sense relative to your other measurements, I’m guessing a typo? 27.3 V on the emitter of Q1 and 8 V on the emitter of Q3 sort of narrows it down to a faulty connection or trace at Q3, or a bad Q3. Any chance the base and emitter are interchanged on Q3 (easy to do on a TO-3 package that’s been hand wired)? I'm assuming nothing's getting hot.
 
kaos said:
The 20.8 V across Q4 doesn’t make sense relative to your other measurements, I’m guessing a typo? 27.3 V on the emitter of Q1 and 8 V on the emitter of Q3 sort of narrows it down to a faulty connection or trace at Q3, or a bad Q3. Any chance the base and emitter are interchanged on Q3 (easy to do on a TO-3 package that’s been hand wired)? I'm assuming nothing's getting hot.


You were absolutly right kaos, on both accounts!

1st, The voltage across Q4 was a typo. It should have read 1.8V.

2nd, I also thought that all roads pointed towards a bad Q4. I had swapped Q4's between the +Ve and -Ve several times and the problem never followed transistor. Also, the voltage output from the -Ve side also seemed about right. So I ASSUMED that the pass transistors were ok. The transistors were in fact ok, but I had the emitter and base swapped in my circuit. Wow. So now I feel a little foolish that I didn't catch this myself sooner. Well it was my 47th birthday yesterday so I will blame everything on that!

My goal now that I have this running is to clean up everything, put it in a chassis shorten the wires up and do some measurements. Are there any good websites that would get me going on doing the power supply measurements?

If I do some very basic measurements using the formula:

Vac/Vdc * 100 =% ripple

I get the following

Vac/Vdc * 100 =% ripple
%ripple=.002/26.9 * 100
%ripple = .007%

Thats .002 VAC measured from +Ve to ground and 26.9V DC from the same point. This seems very low indeed! Below is a picture of the PSU Board

Thanks to everyone who helped me on this. I really love this form everyone is a big help. Its just great to have a place to turn to for questions on this stuff and they actually make you think!
 

Attachments

  • _dsc9569.jpg
    _dsc9569.jpg
    92.4 KB · Views: 268
2mVac of output ripple is low and that would be consistent with a low output current. What was the test current during those measurements?

The nice thing about a capacitance multiplier is that as the output current rises the output ripple increases much more slowly than on a conventional PSU. This suits amplifiers that draw a high quiescent current and have a low PSRR. low ripple = low hum.

The normal way to quote ripple is Vpp. The DMM and any analogue voltmeter when set to AC tries to measure the average voltage and then scale it to read an effective RMS voltage.
Vpp~ 3* Vac.
Your ripple if measured on a scope would be ~ 6mVpp.

BTW,
some DMM do not work as your's did measuring AC superimposed on a DC voltage. They try to add some portion (or all of) the average DC voltage to the average AC voltage and give a very high total voltage reading that can overload or even damage an unsuitabel meter.
Check any other meters you have starting with the highest Vac scale to ensure you don't overload them.
 
Hi Andrew,

I didn't measure the actual current at that time however I did have a 50W 10 ohm resistor at the +Ve output and gnd. If my calculations are correct that should be about 2.7 amps. I'm planning on doing more testing later today. The board is setup to have two power supplies on the same board and I still have the second one to finish wiring up so I will do that too.

The meter I was using was a Radio Shack 22-168. I also have an old Simpson model 467 "True RMS" meter and a 200 MHz Tektronics Oscilloscope that I will try as well.

Cheers,
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.