Opinions on state of AU-20000 sought

[Hyperion]

There are quite a few 'key' resistors in that list which have the most difference in value to the expected, and thus would make the most difference to the relay delay.

Out of curiosity was the LED connected to X & Y when you were observing the relay time delay, as I think the current draw through R23 with the LED connected would slow down relay 1 activation... just a thought...

 

[KeithD]

There are quite a few 'key' resistors in that list which have the most difference in value to the expected, and thus would make the most difference to the relay delay.

Out of curiosity was the LED connected to X & Y when you were observing the relay time delay, as I think the current draw through R23 with the LED connected would slow down relay 1 activation... just a thought...

Only on my initial test, when I had the board back in the amp. All subsequent tests did not have the LED connected. I was testing it on the bench with an external 20 VDC supply. Should I connect an LED across X-Y and give it a test? The only things I've changed at this point are KSC1845 for TR03 and 04.

 

[KeithD]

OK, tried putting an LED lamp across X-Y, but no change in relay. Looking at the schematic @richardpp annotated in post #98 I realized he was right about looking at some of the diodes. I decided to replace D17 and D18 which are in the relay time delay circuit as indicated on his diagram. Also put a single 22 µF/50V cap in at C10 (closest I had on hand) and did a better repair on the + pad on the foil side (had been damaged and then ripped off). I'm now getting 4 seconds delay on the relay, so moving in the right direction. Will replace the rest of the IS2473 diodes and see where things are at.

 

[richardpp]

I agree with @Hyperion that something has fundamentally changed when you restored the protection components and some key resistors that affect the delay time are significantly out of tolerance, R15 for example.

@KeithD your resistance tolerance chart is very revealing! It helps explain the change in behaviour of the timing of both relays. And I agree with you that this, or at least something you changed, was masking another problem. It looks like changing your resistors corrected some timing issues with the relays and this has revealed other issues to do with charge bleeding off C10.

 

[KeithD]

I agree with @Hyperion that something has fundamentally changed when you restored the protection components and some key resistors that affect the delay time are significantly out of tolerance, R15 for example.

@KeithD your resistance tolerance chart is very revealing! It helps explain the change in behaviour of the timing of both relays. And I agree with you that this, or at least something you changed, was masking another problem. It looks like changing your resistors corrected some timing issues with the relays and this has revealed other issues to do with charge bleeding off C10.

Slightly off topic but the transformer will be less stressed now that RL01 is closing on time.

Just to be clear, I did not keep track of which 100 kΩ resistor was in R15 or R18 position. They could have been switched from the way I listed them.

I think we are getting a better picture of the state of this amp. Given that all output transistors were replaced at some point and the damage to the trace on F-2511 where the entire length melted off the mask and lifted from the board (+68V supply), I'd say this had a significant OT failure at some point. I'm thinking that failure caused damage to components on F-2511 including resistors and diodes. Somehow this unit kept working with the original complement of components, but yeah, brining the resistors into spec seems to reveal the damage to the diodes. Actually, given the damage to the trace in the +68V supply, I'd have thought D01 would have been replaced, but it seems original to me with markings that match the SM part. R03 and 04 did measure in spec.

I'll replace D19 and 20 next and do another test to see if the relay delay lengthens some more. Then continue on replacing diodes. I'm now thinking it would be a good idea to replace all of the 1S2226 diodes as well (seems 1N4007 is recommended, which I have). Thanks again @richardpp for your insights and pointing in the right direction!

 

[richardpp]

All, good Keith! I'm confident you'll resolve this. D09 through D14 are of particular interest.

Again slightly off topic, but I think you should consider changing R01 and R02 back to 3.9Ω.

 

[KeithD]

All, good Keith! I'm confident you'll resolve this. D09 through D14 are of particular interest.

Again slightly off topic, but I think you should consider changing R01 and R02 back to 3.9Ω.

OK, plan to work on the diodes tomorrow.

I did not replace R01 or 02, did you mean R03 and 04 (3.9 kΩ)? I do have the originals and they tested in spec.

 

[richardpp]

R01 R02 are the soft start resistors. The original resistors would have been 3.9Ω 7W (Noble) cement wire wound. Yours are 4.7Ω 10W .

The lower resistance value will allow more current to flow. The amount of current flowing and duration of time that the soft start relay is open for determines how much charge builds up in C10. Once the relay closes C10 stops charging. After the first relay closes, the second relay closes in as long as it takes for the charge build-up in C10 to drain to a level that turns off TR03.

If the relay was staying open longer than normal because of aging board components it could explain why the delay time was originally passable. Now that the timing has been fixed we're trying to find out if either C10 is not getting sufficiently charged or if it is, is that charge then dissipating too quickly by leaking to somewhere it shouldn't.

I'm trying to figure this out as I go so sorry for the twists and turns. Also, I deleted an earlier comment in post #144 about the transformer being stressed because it was simply wrong.

This post by Trnsfmr gave me some valuable insight : https://audiokarma.org/forums/threa...st-fix-for-safe-operation.733713/post-9929380

 

[KeithD]

R01 R02 are the soft start resistors. The original resistors would have been 3.9Ω 7W (Noble) cement wire wound. Yours are 4.7Ω 10W .

The lower resistance value will allow more current to flow. The amount of current flowing and duration of time that the soft start relay is open for determines how much charge builds up in C10. Once the relay closes C10 stops charging. After the first relay closes, the second relay closes in as long as it takes for the charge build-up in C10 to drain to a level that turns off TR03.

If the relay was staying open longer than normal because of aging board components it could explain why the delay time was originally passable. Now that the timing has been fixed we're trying to find out if either C10 is not getting sufficiently charged or if it is, is that charge then dissipating too quickly by leaking to somewhere it shouldn't.

I'm trying to figure this out as I go so sorry for the twists and turns. Also, I deleted an earlier comment in post #144 about the transformer being stressed because it was simply wrong.

This post by Trnsfmr gave me some valuable insight : https://audiokarma.org/forums/threa...st-fix-for-safe-operation.733713/post-9929380

Thanks for the info on R01 and 02. The 4.7Ω resistors were in there when I purchased the amp. I don't have anything close to the original rating on hand, so that will be another order.

Replaced D19 and D20 and the second relay click shortened up to 2 seconds! Replaced the rest of the 1S2473 glass diodes with no change. I'm wondering if 1N4007 for D20 is a good replacement? Might put the original back in to see if we get back up to 4 second delay on the relay.

 

[richardpp]

Hi Keith, I veered off topic with R01/ R02. That can be looked at later if needed.

It is very unusual that replacing D19 and D20 caused the delay time to decrease. D19 is there to block current path back through the thermistor to ground.
D20 suppresses back emf when the speaker relay opens. I'm not sure why that's happened. Edit: 1N4007 is a good replacement. Good you changed the glass diodes out but sorry to put you through that work unnecessarily as it turned out.

I've set up an F-2511 board on the bench hooked up to a 20V supply to replicate your set up. The first relay clicks in almost immediately, the second relay has a 7 sec delay time.

For testing the delay circuit I tried this:

Remove TR03 from the board.
Switch on the 20V supply and measure the voltage at R18. (Here I'm getting a reading of 14V, stabilising after a few seconds.)
Turn off the supply and count how long it takes for the voltage to drop to below approximately 0.7V (I'm getting approx 7secs.)

Voltage is measured on the side of R18 that was previously connected to the base of TR03.

 

[richardpp]

Sorry Keith, I need to correct that last post.

The above test was with one of the jumpers disconnected. Try disconnecting either of the two jumpers and see if you get the same behaviour.

Then try the above test with both jumpers connected. Here I'm see the voltage at R18 rise to nearly 3V and then fall to below 0.7V in approximately 6 secs.

 

[KeithD]

Hi Richard, WOW, thanks for setting up this experiment! Hopefully we can find the faulty area.

Took TR03 out. (Just to be clear, I went back to KSC1815 for both TR04 and TR03 - took out the KSC1845 transistors I had in there.) Here are my observations.

With only one jumper in and applying 20 VDC, the relays click in instantaneously. Voltage at R18 stabilizes at 12V. Falls to 0.8V at about 7 seconds after power off.

With both jumpers connected, voltage at R18 goes to 0 V at power on, and stays at 0 V at power off. Very different from what you are seeing.

 

[richardpp]

I’m not at my bench at the moment, but that’s an interesting result.

In the first test, C10 is charging through the 33k ohm resistor at either R16 or R17 and drains through R18.

In the second test, C10 is supposed to charge through the relay coil. There should be 24V nominal on both sides of the relay coil when the contacts are open and 1V at the point indicated on the schematic, ground side of the coil, when the relay is energised.

 

[richardpp]

It’s going to be hard to measure the open voltage on the first relay because it closes so quickly, you would need to stop the relay closing by removing a component.

What is the resistance across the soft start relay coil? I'm seeing ~668Ω.

 

[richardpp]

With only one jumper in and applying 20 VDC, the relays click in instantaneously. Voltage at R18 stabilizes at 12V. Falls to 0.8V at about 7 seconds after power off.

That's a couple of volts short of where it should be, check your supply voltage?

With both jumpers connected, voltage at R18 goes to 0 V at power on, and stays at 0 V at power off. Very different from what you are seeing.

Try measuring the voltage at the junction of R15 and D17. Here I'm seeing a 5V peak.

You might be able to capture a reading on the other side of D17 if your DMM has a peak hold function, ~4.7V

The resistance at junction of R15 and D17 to ground, ~200KΩ

The capacitance at junction of R15 and D17 to ground, ~15µF (22µF)

 

[KeithD]

I’m not at my bench at the moment, but that’s an interesting result.

In the first test, C10 is charging through the 33k ohm resistor at either R16 or R17 and drains through R18.

In the second test, C10 is supposed to charge through the relay coil. There should be 24V nominal on both sides of the relay coil when the contacts are open and 1V at the point indicated on the schematic, ground side of the coil, when the relay is energised.

Measured the PS voltage = 20.7 VDC I'm getting a voltage across the relay of 20 V and 0.83 V where the schematic indicates 1 V.

It’s going to be hard to measure the open voltage on the first relay because it closes so quickly, you would need to stop the relay closing by removing a component.

What is the resistance across the soft start relay coil? I'm seeing ~668Ω.

Resistance at RL2 measured across D16 is 663 Ω.

That's a couple of volts short of where it should be, check your supply voltage?


Try measuring the voltage at the junction of R15 and D17. Here I'm seeing a 5V peak.

You might be able to capture a reading on the other side of D17 if your DMM has a peak hold function, ~4.7V

The resistance at junction of R15 and D17 to ground, ~200KΩ

The capacitance at junction of R15 and D17 to ground, ~15µF (22µF)

I do have a peak hold function o my meter. Peak at R15/D17 junction is 17 V. Peak on other side of D17 = 7.8 V.

Resistance at R15/D17 to ground = 590 kΩ and rising. Capacitance at R15/D17 to ground = 1 µF.

 

[richardpp]

Measured the PS voltage = 20.7 VDC I'm getting a voltage across the relay of 20 V and 0.83 V where the schematic indicates 1 V.

It looks like these voltages are taken at the speaker relay, not the soft start relay, is that right?

Resistance at RL2 measured across D16 is 663 Ω.

That's good.

I do have a peak hold function o my meter. Peak at R15/D17 junction is 17 V. Peak on other side of D17 = 7.8 V.

Resistance at R15/D17 to ground = 590 kΩ and rising. Capacitance at R15/D17 to ground = 1 µF.

Let me just check, TR03 has been removed and your resistance and capacitance checks were done with the power supply switched off.

Resistance at R15/D17 to ground is made up of R15 +R18, 100KΩ + 100KΩ = 200KΩ.

Capacitance should measure at whatever value you put on the board at C10. I think I would remove the capacitor(s) you have at C10 and test them on your meter.

Edit: ... and check the traces and pads around C10.

 

[KeithD]

It looks like these voltages are taken at the speaker relay, not the soft start relay, is that right?

That's good.

Let me just check, TR03 has been removed and your resistance and capacitance checks were done with the power supply switched off.

Resistance at R15/D17 to ground is made up of R15 +R18, 100KΩ + 100KΩ = 200KΩ.

Capacitance should measure at whatever value you put on the board at C10. I think I would remove the capacitor(s) you have at C10 and test them on your meter.

Edit: ... and check the traces and pads around C10.

Yes, voltages at the speaker relay.

TR03 removed and PS off.

For resistance I was measuring R15/D17 to ground (pin 23), I think you meant to not include C07, but rather use the "ground side" of R18? Doing that I get 116 kΩ.

Also measuring capacitance from R15/D17 to the negative lead of C10, I get 18 µF. I've got a 22 µF cap in for C10. If I measure the capacitor directly in circuit I get 22 µF.

Sorry, I may be making measurements wrong, and I've got a busy day/week ahead, so unfortunately, I have only early morning or later at night to work on this during the week.

 

[richardpp]

No problem, Keith.

Understood about the busy schedule, will try and keep it to the point.

For our purposes ground is the negative of the PS.

Can you check this out, it looks like a crack? You said you had problems when replacing D19 and D20, this is D20 to TR04.

 

[Hyperion]

Very good eye Richard.

The crack seems to be collateral damage relating to the 'roughed up' joint just above it in the picture.