Help diagnosing power supply circuit failure

[jimkarl]

Having a problem with this power supply circuit (pdf attached). My main starting point issue is it's supposed to be putting out -15v and +15v to pins 1 & 4 respectively on PN201 connector. Pin 1 is showing -33v and pin 4 is showing 19v. I've noted on the schematic some various voltage points. These were measured with powersupply board isolated - no other boards connected. This is what I've checked so far:
1) ESR metered all caps 10uf and larger. They all register good. Triple checked those around the voltage regulator as well (C905/C907)
2) Pulled the Bridge rectifier out and diode checked between all 4 pins. Shows .556v in one direction, infinity in the other and from AC ins (2 to 4) .966v. Seems good, right?
3) Diode checked various other diodes and they seem to check ok as well but to be honest those were all done in circuit so I realize that's not 100% guarantee on the readings.
Hopefully someone can see something in the circuit that would be causing such a higher than expected rectifier output, especially the lopsided negative side vs the positive side.

 

[alec_t]

from AC ins (2 to 4) .966v. Seems good, right?

Wrong. With the bridge out of circuit you should see infinity between pins 2 and 4. If the associated reservoir caps have been getting the wrong polarity voltage as a result of bridge failure you'd probably need to replace them.

 

[jimkarl]

Wrong. With the bridge out of circuit you should see infinity between pins 2 and 4. If the associated reservoir caps have been getting the wrong polarity voltage as a result of bridge failure you'd probably need to replace them.

So it should read infinity in both directions (2 to 4 and 4 to 2), right?

 

[alec_t]

Right.

 

[jimkarl]

Thanks! Part on order! (plus a spare...)

 

[Tony Stewart]

Wrong. With the bridge out of circuit you should see infinity between pins 2 and 4. If the associated reservoir caps have been getting the wrong polarity voltage as a result of bridge failure you'd probably need to replace them.

For resistance open circuit is ok but in Diode test mode the results are good. THis mode applies a CC of 1mA and reads the diode voltage which is correct for power diodes at low current.

I suggest remove fuses and measure single phase AC w.r.t. gnd.. The transformer secondary centre tap may be imbalanced. No load DC will be +- 1.4* Vac plus drop expected due to winding loss at rated Vout.

 

[jimkarl]

Here are the voltages with fuses out:
Pin 3 -2.5vdc/-2.1vac
Pin 4 -2.5vdc/0vac
Pin 5 15.1vac
Pin 6 0
Pin 7 15.1vac
Pin 8 8.4vac
Pin 9 0

 

[Tony Stewart]

Thanks for checking that Jim. the only other thing that can cause it is a ground fault.

So I checked your schematic again and noticed the Supply cap were much different 3.3mF on V- and 6,8mF on V+, presumably due to more current, on the + side but pulling up the negative side means the negative load current is much more than the positive at the time of test. (Edited)

The imbalance is due to a lack of connection between secondary Centre tap and your regulator Gnds.

The centre tap is earthed but DC neutral is not. Fix that and you are good to go.

If you want an isolated supply remove strap to Chassis Gnd and connect to Regulator grounds. and use a small RF cap to Chassis from DC neutral. But keep in mind if AC entry is chassis grounded and Line filter with caps to earth will be coupled to your DC out and thus imposes a safety leakage restriction of 0.25mA and thus value of cap at 50/60 Hz even though your transformer is isolated. It also means lightning ground surges can still pass thru to the secondary although much attenuated.

 

[Tony Stewart]

I now see this is not your design, but a Harmon Kardon Amp
Note in the schematic corner Gnd to Chassis labelled AU.. This is either gold colour insulation of gold plated earth connection or just Another Uhoh. (AU) that is missing....

 

[Tony Stewart]

BTW good ESR in this case for a dampening factor of 100 on 8 Ohms is means ESR<< 1% of 8R or <<80 mOhm

Looks like some hot regulators there under load with excess voltage drop. Could be improved nowadays.

 

[jimkarl]

Thanks for checking that Jim. the only other thing that can cause it is a ground fault.

So I checked your schematic again and noticed the Supply cap were much different 3.3mF on V- and 6,8mF on V+, presumably due to more current, on the + side but pulling up the negative side means the negative load current is much more than the positive at the time of test. (Edited)

The imbalance is due to a lack of connection between secondary Centre tap and your regulator Gnds.

The centre tap is earthed but DC neutral is not. Fix that and you are good to go.

If you want an isolated supply remove strap to Chassis Gnd and connect to Regulator grounds. and use a small RF cap to Chassis from DC neutral. But keep in mind if AC entry is chassis grounded and Line filter with caps to earth will be coupled to your DC out and thus imposes a safety leakage restriction of 0.25mA and thus value of cap at 50/60 Hz even though your transformer is isolated. It also means lightning ground surges can still pass thru to the secondary although much attenuated.

I have to admit I'm completely lost on your answer so you'll have to take it a little slower for me to catch up. Old guy here learning new tricks LOL. Or simply just learning. First, are you saying there is nothing wrong with the voltage regulator that I just ordered and should be here in a couple of days? That something else is wrong instead? Is there another part failure that I need to be looking at? If it's the power supply (transformer) itself I think I'm SOL because there's no way to get one of those for this unit anymore. I'm also not clear on the whole "remove strap to chasis" purpose. I didn't want o change the way was designed to work. Really just identify the source of the problem and replace the part(s). Unless you are saying the problem is the power supply (transformer) itself is bad and this is a fix to that transformer?

 

[jimkarl]

I now see this is not your design, but a Harmon Kardon Amp
Note in the schematic corner Gnd to Chassis labelled AU.. This is either gold colour insulation of gold plated earth connection or just Another Uhoh. (AU) that is missing....

View attachment 92175

Correct this is a HK CD recorder. So if I read this correct you are saying to check the path for continuity on those lines?

 

[Tony Stewart]

yes
earth to reg gnd


"The imbalance is due to a lack of connection between secondary Centre tap and your regulator Gnds."

 

[jimkarl]

yes
earth to reg gnd


"The imbalance is due to a lack of connection between secondary Centre tap and your regulator Gnds."

So check every point on the diagram that shows a ground point and make sure it actually is grounded on the board to the same point on the left side of the transformer? Or am I checking every point on the left side to some point on the right side? I get confused when you say earth to reg gnd.

 

[Tony Stewart]

"The imbalance is due to a lack of connection between secondary Centre tap and your regulator Gnds." (1)
Your 125Vac test to earth ground, showed that it was connected.
Therefore the connection (1) is missing because the Regulators are not all grounded.

THe yellow square in the corner of the schematic shows it should be.
Look for it.
If you cant find it. make it connected to earth chassis, preferably close to where the Transformer centre tap is grounded to chassis.

 

[jimkarl]

Tony - First - Thanks for your patience with me. I realize it's baby steps. Second - please don't think that I didn't read the "the imbalance is due to" feedback all 3 times. It wasn't a question of missing that comment as a cause - it was a question of what exactly to do about.
Now you mention the yellow square - none on my diagram but looks like there is one on yours around the AU ground (two of them). So for clarity sake - I am supposed to check those yellow lines, make sure they all tie out to where ever they go (ICs or pin connectors etc) as well as to a point on the *chassis*? to which they should all be terminated to? Or the ground of the power plug line? I didn't and still don't, understand the difference between a AU ground and a regular ground. The regular ground is clearly where pin 6 of the transformer ties to. I've never seen a schematic with two types of grounds before. I'm sure very common, but just nothing that I've come across.

 

[Tony Stewart]

I can't be more explicit.

Plan B. You don't need to find the broken connection.

Just make one.

Can you make a connection from the secondary Tap to the +/- regulator pair "Common" or "neutral" connection ?

This will be a success. Using the same wire gauge or heavier, in a reasonable short path or.... if shorter... connect to same chassis ground.

I added the yellow lines for clarity, sorry to confuse you.

The reason for careful grounds or separate grounds is to avoid creating a few millivolts of ripple on this ground connection near an audio signal ground . Then you get a buzz on your audio.

This neutral ground wire will carry 10x the average current or more as current pulses the capacitors to keep them with l,ow ripple. Any ohmic loss or resistance in the wire or chassis or connection, will create an audible buzz. Basic Ohm's Law.. V=I*R

A thought for Engineers...
As usual, schematics are also called "logic diagrams" and ground connections are often the most "illogical" and go the furthest point on the drawing, or worse, aren't shown and assume you should know. It is even more critical to show proper connections when separating grounds for power, analog and digital to prevent cross-talk by conduction loss.

If you understand, no explanation necessary, if not, no explanation possible. No Worries.

 

[jimkarl]

Can you make a connection from the secondary Tap to the +/- regulator pair "Common" or "neutral" connection ?
.

Yes - if I'm clear on the from and to points to connect are! I'm left guessing where these are.
"the Secondary tap" - Exactly which pin on the secondary side of the transformer are you referring to? Pin 6?
"the +/- regulator pair common or neutral connection" - Please explain to what component (by part number) I'm connecting this to? If it is Pin 6 above, then I assume you are saying to C905 (ground side) but please confirm or correct part reference to connect to.

 

[Tony Stewart]

first explain what you used for ground in
on PN201 connector. Pin 1 is showing -33v and pin 4 is showing 19v.
If using the chassis gnd.

, then voltage on PN201 pins 2,6,8,10 should all be 0 V

If not they should have been grounded at one time.

The reason they are separate is called a Star Ground design where each circuit has its own dedicated supply return , so there is no crosstalk between circuits on the supply, which can get into pre-amps.

Then if there is an open ground, why and where should it be depends on construction details such as the earth ground symbols in yellow I made to

 

[jimkarl]

Today was a good day! Part came and I solved the ground problem I thought the board was grounded to the case from a wire but in reality it was direct contact to the case. Since I had taken it out of the case to debug the bad voltages I was reading, I didn't realize I then had lost my ground connection. Stupid! Anyway, with the new regulator part, power board back in the case, some touched up solder points and a resetting of all connectors - everything works fine now! I want to thank everyone for their help and patience through my (beginner) learning process. Thanks again!