Help with non-charging capacitor off of a bridge rectifier

[mapleLC]

Help with non-charging capacitor off of a bridge rectifier fed by a torrodial transformer

I have 2 brand new caps on a powered subwoofer circuit.  I am getting good voltages across the rectifiers, then the circuit charges 2 caps.  One caps charges to the 40v+ desired, but the other cap never charges past -1.5v or at least shows this voltage on the meter.

Can anyone think of a reason the other cap wont charge?  The cap is not giving me a bad reading on the surface that I can see so far, and its new.

What culprit might explain this set of behaviors?

[inse]

Photos from both sides, please.

[Ian.M]

I'd bet that if it doesn't have an inrush surge prevention circuit, something (or a connection) has gone high resistance 'upstream' of the capacitor, as anything that could drag down the output voltage by over 96%, without an upstream fault, would be drawing so much current it would be blowing fuses or even tripping the breaker on the mains circuit feeding the amp.

In addition to inse's request, provide the exact make & model number, or if its a home-built project, provide (a link to) the schematic.

[mapleLC]

Here is a photo of the board. 

One of the photos shows the star ground and the location of the cap that is not charging.

[mapleLC]

I'd bet that if it doesn't have an inrush surge prevention circuit, something (or a connection) has gone high resistance 'upstream' of the capacitor, as anything that could drag down the output voltage by over 96%, without an upstream fault, would be drawing so much current it would be blowing fuses or even tripping the breaker on the mains circuit feeding the amp.

In addition to inse's request, provide the exact make & model number, or if its a home-built project, provide (a link to) the schematic.

It has 2 fuses which were blown when I received the sub.

Unfortunately, no schematic.

There doesn't appear to be anything at between the rectifier and the cap.

I am going to go back to the variac and measure voltages again so I can share what values are around the rectifier. 

[mapleLC]

Please see the attached.  These were taken with the variac on at 80v.

I also wanted to note the resistors under the rectifier.

With the 2nd handwritten note, I am trying to point out that I am seeing 40v at stripe side of D8 and D9, and -40v if I reverse the leads.

[Ian.M]

Pull the fuses to isolate the input side of the bridge rectifier and use the diode check range on your DMM (or a low ohms range on an analog meter) to check each diode in the bridge, in circuit.  If all are good, their Vf should all be very similar.  You wont be able to get a reverse reading as the other diode in each pair will conduct, passing current through the load circuit. 

If you get significantly differing Vf readings, desolder the cathode ends of D8 and D9 to disconnect the positive side of the bridge and repeat your diode tests.  Now the load circuit is disconnected, you can test for reverse leakage of individual diodes and be certain that the Vf and reverse leakage you are seeing applies to the diode the test leads are on.

While D8 and D9 cathodes are disconnected, take the opportunity to check continuity from the bridge + and - pads to their reservoir capacitor terminals.

I suspect its either got a cracked track or the diodes D10 and D11 feeding the negative reservoir capacitor have gone high resistance.

[inse]

Where are the fuses FS1 and FS2?
Those are ceramic capacitors under the diodes.

[mapleLC]

Pull the fuses to isolate the input side of the bridge rectifier and use the diode check range on your DMM (or a low ohms range on an analog meter) to check each diode in the bridge, in circuit.  If all are good, their Vf should all be very similar.  You wont be able to get a reverse reading as the other diode in each pair will conduct, passing current through the load circuit. 

I have done this previous, they all check out.  I was intending to share the voltages I am seeing in circuit.

If you get significantly differing Vf readings, desolder the cathode ends of D8 and D9 to disconnect the positive side of the bridge and repeat your diode tests.  Now the load circuit is disconnected, you can test for reverse leakage of individual diodes and be certain that the Vf and reverse leakage you are seeing applies to the diode the test leads are on.

While D8 and D9 cathodes are disconnected, take the opportunity to check continuity from the bridge + and - pads to their reservoir capacitor terminals.

I suspect its either got a cracked track or the diodes D10 and D11 feeding the negative reservoir capacitor have gone high resistance.

I have performed most of these checks but not in the exact manner you prescribe.  Let me give that a try and see what I read.

[mapleLC]

Where are the fuses FS1 and FS2?
Those are ceramic capacitors under the diodes.

I removed them and wired in glass fuses so I could see if something blew.  I have replacements ready once the unit is sorted again.

Should those caps below the diodes be a worry?  I didn't think they should be.

[StillTrying]

Have you measured the AC voltages fron the star point to FS1, and from the star point to FS2, they should be about the same.

And checked that's not a short.

[mapleLC]

Have you measured the AC voltages fron the star point to FS1, and from the star point to FS2, they should be about the same.

And checked that's not a short.

I cleaned that up to no effect.

The AC voltages are just about identical at both points.

[mapleLC]

Pull the fuses to isolate the input side of the bridge rectifier and use the diode check range on your DMM (or a low ohms range on an analog meter) to check each diode in the bridge, in circuit.  If all are good, their Vf should all be very similar.  You wont be able to get a reverse reading as the other diode in each pair will conduct, passing current through the load circuit. 

If you get significantly differing Vf readings, desolder the cathode ends of D8 and D9 to disconnect the positive side of the bridge and repeat your diode tests.  Now the load circuit is disconnected, you can test for reverse leakage of individual diodes and be certain that the Vf and reverse leakage you are seeing applies to the diode the test leads are on.

While D8 and D9 cathodes are disconnected, take the opportunity to check continuity from the bridge + and - pads to their reservoir capacitor terminals.

I suspect its either got a cracked track or the diodes D10 and D11 feeding the negative reservoir capacitor have gone high resistance.

I disconnected one lead of each of the fuses so no power is in circuit and I am getting similar readings in the low mv range for each diode of about 200 to 300 mv

The continuity check for the diodes shows resistance in the Mohm range testing all 4 cap pads against both sides of the diode.

[StillTrying]

"The AC voltages are just about identical at both points."

That's a good start.

Measured from the startpoint, if the -Ve DC voltage was only around 0.7 of the +Ve DC voltage it would mean there's a no connection between the -Ve diodes and the -Ve DC smoothing cap., I think.

[Ian.M]

That's far too low a Vf for those diodes to be good.   I can make out enough of the part numbers in your photo to determine they must be 1N5402, which is an ordinary 200V 3A silicon rectifier diode. [Vishay Datasheet]

Vf should be in the 600mV to 700mV region and even at a very low test currents shouldn't drop below 500mV.

Either your DMM is faulty, or something is shunting the diodes, or at least two of them are bad and they are shunting each other.  Confirm it isn't your DMM by testing an ordinary silicon diode (*NOT* a schottky one) or a transistor junction out of circuit, then desolder D10 and D11 (my suspects for the bad ones) and check them out of circuit.  You can also then check D8 and D9 in circuit as the negative side of the bridge will be disconnected.  All should have Vf in the range above and no measurable reverse leakage.

[mapleLC]

That's far too low a Vf for those diodes to be good.   I can make out enough of the part numbers in your photo to determine they must be 1N5402, which is an ordinary 200V 3A silicon rectifier diode. [Vishay Datasheet]

Vf should be in the 600mV to 700mV region and even at a very low test currents shouldn't drop below 500mV.

Either your DMM is faulty, or something is shunting the diodes, or at least two of them are bad and they are shunting each other.  Confirm it isn't your DMM by testing an ordinary silicon diode (*NOT* a schottky one) or a transistor junction out of circuit, then desolder D10 and D11 (my suspects for the bad ones) and check them out of circuit.  You can also then check D8 and D9 in circuit as the negative side of the bridge will be disconnected.  All should have Vf in the range above and no measurable reverse leakage.

I didn't think it mattered to mention, but the transistors have been pulled from circuit to eliminate a root cause.

They are 1N5402 diodes, I have some 5404s here I can swap them for and retest the circuit.

[mapleLC]

"The AC voltages are just about identical at both points."

That's a good start.

Measured from the startpoint, if the -Ve DC voltage was only around 0.7 of the +Ve DC voltage it would mean there's a no connection between the -Ve diodes and the -Ve DC smoothing cap., I think.

I reconnected the fuses and turned the voltage to 120v.

Here are my readings, I have not changed any diodes or put the transistors back in.

I left the black lead on star ground and measured AC and DC across both sides of each diode

At D8 and D9 striped side, the meter cycles thru looking for a reading and then just beeps, it never actually reads anything for AC.

[StillTrying]

At D8 and D9 striped side, the meter cycles thru looking for a reading and then just beeps, it never actually reads anything for AC.

That's probably correct, with no load on the +40V there'll be almost 0 AC ripple.

The 22 VAC and -17 DC on the -Ve of D10 and D11 are about the AC and DC the voltages you'd get for half wave unsmoothed. Which mught mean there's 2 faults, one of D10 or D11 is blown, plus there's no connection to the smoothing cap.

Which is similar to.
I suspect its either got a cracked track or the diodes D10 and D11 feeding the negative reservoir capacitor have gone high resistance.

So check the resistance from the -Ve diodes to the cap.

I can't see every connection in the photos, but can't see how it's not just this.

[mapleLC]

At D8 and D9 striped side, the meter cycles thru looking for a reading and then just beeps, it never actually reads anything for AC.

That's probably correct, with no load on the +40V there'll be almost 0 AC ripple.

The 22 VAC and -17 DC on the -Ve of D10 and D11 are about the AC and DC the voltages you'd get for half wave unsmoothed. Which mught mean there's 2 faults, one of D10 or D11 is blown, plus there's no connection to the smoothing cap.

Which is similar to.
I suspect its either got a cracked track or the diodes D10 and D11 feeding the negative reservoir capacitor have gone high resistance.

So check the resistance from the -Ve diodes to the cap.

I can't see every connection in the photos, but can't see how it's not just this.

Thanks for the response.

I changed D10 and D11 and this is the updated voltages.  It looks like the rectifier is working, but voltage is off on one side.

The cap still takes no voltage, but I wanted to address this first as it does seem like 2 problems

Any thoughts towards the imbalance?  Note I changed the diode to a 1n5404 as its all I had but it shouldnt matter for this.  And other than the imbalance, this looks fairly normal correct?

[StillTrying]

I would think the -35 VDC will go up to -40 VDC when/if you find out why the -Ve smoothing cap is not working.

[mapleLC]

I'll dive into that.  On the surface, it seems the bridge rectifier is now working unless someone still sees an issue.

[mapleLC]

I pulled the cap that was not charging from circuit and connected it directly via the leads as photographed here.

I went from the - voltage I read (about 25v) to the neg lead of the cap, from the + lead of the cap to the star ground.

No cap worked, I tried all of them including a brand new one out of the box.  Nothing will charge past about .5v.

Unless I am doing something wrong with the way I am leading it, I dont think so, since the behavior has not changed.  With the cap directly wired in, I still get the same charge.

After seeing the new voltage readings before, I thought perhaps this would expose a bad connection between the diode and the cap but it doesnt seem to.

[Ian.M]

Do you have a bench PSU with an adjustable output and current limiting?  If so, pull the suspect diodes and test them at at least 300mA.  If not, test them on whatever power supply you have with a low voltage filament bulb in series.  I suspect the lead to die joint has cracked and they are barely passing any current.

[StillTrying]

Have you checked all those crock leads, half the ones I've got have gone high resistance, intermittant or open circuit, especially when using them around a power supply that can supply pulses of many Amps.
When I've had to remake one I've always used thicker wire.

From some of you measurements I think it's unlikely there's a problem with the star point.

Have you checked again that the bridge is still working, measured from the star point, if the +V side is around 50V DC the -Ve side should be around -30 VDC because there's no smoothing on that side. If the -Ve side drops to around -17V DC again one of D10 or D11 would be blown, again.

I can't think of anything else ATM.

[mapleLC]

Have you checked all those crock leads, half the ones I've got have gone high resistance, intermittant or open circuit, especially when using them around a power supply that can supply pulses of many Amps.
When I've had to remake one I've always used thicker wire.

From some of you measurements I think it's unlikely there's a problem with the star point.

Have you checked again that the bridge is still working, measured from the star point, if the +V side is around 50V DC the -Ve side should be around -30 VDC because there's no smoothing on that side. If the -Ve side drops to around -17V DC again one of D10 or D11 would be blown, again.

I can't think of anything else ATM.

So the voltage has dropped again to -17. 

I went ahead and soldered in some leads and soldered the cap to the lead so it was well separated from the board, then checked those for the voltages again.  It seems like I am getting a reading across the capacitor now.

I want to get back to your request but I want to make sure I am getting a measurement correctly and the capacitor problem was perplexing.

I am getting 30v DC now at the points that were showing 54v before.   The cap working in circuit is the reason for this now perhaps?

Should I replace those 2 diodes again and try once more?  Or a different way?

I do want to take a moment and thank everyone for their help.  It has been YEARS since I touched this stuff and its a mountain to climb re-learning it.