[Henmyr]

Well, it didn't work perfectly now when I tried it. The led does not shine. I've measured the +15 and -15 against star ground. When I first turn the power on both show +-15V, but after a second they both start to fluctuate around +-10-12V. The output transistors seem to get very warm also. I have heatsinks on all for though.

I've measured at all points for shorts and found none (+-15V to starground and against each other, 0L, 0R and 0G against each other). All electrolytic capacitors are the right way, as are the trimpots and diods. I was really careful when I soldered everything to the boards and always inspected the soldered part for shorts.

Also, the trimpots just keep on turning when I turn them counter-clockwise. There is no "click". I have two Vishay PV36 multi-trim 0736CC trimpots.

What could be wrong?

 

[nightanole]

Quote:

Originally Posted by Henmyr /img/forum/go_quote.gif Well, it didn't work perfectly now when I tried it. The led does not shine. I've measured the +15 and -15 against star ground. When I first turn the power on both show +-15V, but after a second they both start to fluctuate around +-10-12V. The output transistors seem to get very warm also. I have heatsinks on all for though. I've measured at all points for shorts and found none (+-15V to starground and against each other, 0L, 0R and 0G against each other). All electrolytic capacitors are the right way, as are the trimpots and diods. I was really careful when I soldered everything to the boards and always inspected the soldered part for shorts. Also, the trimpots just keep on turning when I turn them counter-clockwise. There is no "click". I have two Vishay PV36 multi-trim 0736CC trimpots. What could be wrong?

You have a short somewhere, your 15v gave all the current it could and tryed to regulate, but couldnt. So its now being pulled down to 12v or less. As for your trim pots, they are multi turn, it took like 7 or more revs to get mine tuned to the 14ma at the test point. Remember that your outputs can go into thermal runaway (its one of the design flaws). The hotter they get, the more current they draw, making them hotter, which makes them draw more current...


I used the stock hammond case, and mounted the outputs directly to the case bottom useing the same thermal pads as used on the regulators. If my caps werent so tall id try to mount the regulators to the case underside too.

 

[Henmyr]

The led was soldered the wrong way. I've changed it and now it glows

I then measured the +-15 points and one showed 15V and the other ~15.3V and both were stable. I then measured R23 and adjusted it to 8mV. I then measured resistor R47 and it showed 300mV. I quickly turned it of. I've added another big heatsink temporarily and it now measures 220-250mV and turning the trimpot counter-clockwise doesn't seem to help.

Unfortunately now, the R23 also show 200mV now and the +-15V points show 15V for a second when I turn it on and then ~10V. Very very strange.

EDIT: R23 is now at 0mV (the trimpot fully counter-clockwise, though I still don't hear a click). R47 still reads 300mV, and the trimpot R43 doesn't seem to do anything to change that value.

measuring OL to OG gives a varying value <20mV and average 10mV. OR to OG show a mean of 500mV, so something is very wrong here.

EDIT2: Some pictures of the build:

I have changed the led since I took the pictures.

 

[amb]

Power off and check your output transistors for shorts. You may have over-biased and damaged them. Also use your DMM's diode-check function to verify the junctions of the output transistors. Search the forum, I've written up posts on how to do this elsewhere in the past.

 

[Henmyr]

Update:
A. +V point read 14.99V now and is stable (fluctuate between 14.98V and 15.00V).
B. -V point read 15.33V now and is stable (fluctuate between 15.22V and 15.33V)
Left channel:
C. Before R1 read 0.00mV.
D. After R1 read 0.00mV (my multimeter can't measure nV, so the reading should be around 0V).
E. Between 2SK170 and R3 read 200mV at first when I power on, but decline over time to around 117mV, but might decline more.
F. There are many readings which seem to change with time, either upwards or downwards.



I can't seem to find a short.

1. The resistance between +15V and -15V pads is 5.7K ohm, which I don't think is a sign of short? 6-8K ohm between star ground and +-15V pads.

2. As I wrote up put in the led diode with the wrong polarity, could this have damaged the output resistors (~300mV over R47)? Before I turn anything on I turned the trimpots counter-clockwise many many turns, so I don't think the trimpots could be the reason for an over-bias.

3. I've now changed to BD140/139 and the voltage over r29 and r47 are now below 20mV, so I now dare to have the amp running while looking for errors. Though neither of them measures 0 when I turn the trimpots fully counter clockwize. They vary a lot from 0 to 20mV. The outputs also show a rather large amount of dc.

5. D1, D3, D6 and D8 all show +-20V. VR1 and VR20 both read +-20V at their inputs and 0 at their GND.

6. VR1 read 15 at output and VR2 read ~15.30, so here something seems wrong. After this, neary every reading on all places (measuring with the "operation points" page as guide) are wrong.

7. The reading between R12 and R13 is 0 so this seems right anyway.

8. Between R1 and potentiometer: Left channel 0mV, right channel 4mV <- not correct?

9. Oh how I appreciate that I didn't start with a B22 Though troubleshooting a CK2III is hard enough.

 

[Chaos4]

Henmyr,

Keep at it, you will get it working.

Starting at VR1 and VR2, you said "VR1 read 15 at output and VR2 read ~15.22". Did you mean VR1 reads ~ +15 volts and VR2 reads ~ - 15.22 volts? If so, I think you are fine here.

If you measure the voltage across R23 and R47 can you adjust the voltage to read ~ 8mV? This would correspond to 17mA of current throught the ouput transistor (8mV/0.47ohms = 17mA).

You said you have a voltage present at the output terminals, how much voltage do you measure?

Make sure the voltages out of the power supply are correct. They should read + 15 volts and - 15 volts. If they do not start there. If they do we will have to start looking at the amplifer proper. Measure the voltages on the op-amp (U1). Make sure you have the + 15 volts and - 15 volts and then tell me what voltages you have at pins 5, 6, and 7.

Also make sure you do not have any voltage present at the input. Just short pins 1 and 2 together at J1 and J2. This will make sure you do not have any voltage present at the input. This is a dc coupled amplifier and as such any voltage present at the input will be coupled completely through the amp.

Lets start with this info and go from there....

 

[Chaos4]

Henmyr,

What is the voltage across R23 and R47 and can you adjust the trimpots to make them read 8mV?

 

[Henmyr]

First connected both pins of J1 and both of J2 (shorted them).

+15V point reads 14.99V.
-15V point reads -15.31V but varies from -15.28V - -15.35V.

Over R23 reads ~14mV but varies from 9-15mV. This is with the trimpot fully counter-clockwize.

Over R47 reads at first around 0.2mV, but raises to 20mV after a few seconds and then declines to ~15mV but varies a few 0.1mV. This is with the trimpot fully counter-clockwize. EDIT: Now it suddenly dropped from 15mV to 0.9mV but still varies a bit around it. EDIT: And now it jumped to 20mV and declined to 14mV then jumped to 0.1mV then jumped to 14mV. Well, it's not very stable at all.

Left output to ground (OL to GND): First around -300mV, but then declined to 0.0xmV and varies.
Right output to ground (OR to GND): Reads -300mV at first when power is off, but it keep rising. Now at +15mV. When power on it first reads ~+2.7V then declines to -300mV and then it rises to mean ~0mV but varies both - and + some 100-200mV around it.

Also, thanks for the help

I really appreciate it.

 

[Chaos4]

The power supply voltages sound ok.

I am a little puzzled by the voltages that seem to vary around a lot. Is it possible that there is a "cold" solder joint somewhere, maybe try and reflow all the solder joints on the board. Another possiblity is a low level oscilation, this may be hard to find. It may also be some type of thermal oscilation, though it looks from the picture that you have pretty good heatsinks on the output transisitors.

It may be usefull to try and turn the trimpots and see what the current does across R47 and R23. Does it go up rapidly and never come down? Make sure and be carefull that the output transistors do not get too hot or become overbiased too much. I would concentrate on one channel at a time here. It would still be usefull to know what the servo loop is doing, so we would need to know what the input and output voltages of the op-amp are doing.

You may also try putting a load resistor on the output just to see what happens. Something on the order of 100 ohms would be good.

I assume you have already done so, but you may want to look over the parts and values again to make sure everything is soldering in the correct spot. It sounds like things are close so I don't think you have a short.

 

[Henmyr]

Quote:

Originally Posted by Chaos4 /img/forum/go_quote.gif The power supply voltages sound ok. I am a little puzzled by the voltages that seem to vary around a lot. Is it possible that there is a "cold" solder joint somewhere, maybe try and reflow all the solder joints on the board. Another possiblity is a low level oscilation, this may be hard to find. It may also be some type of thermal oscilation, though it looks from the picture that you have pretty good heatsinks on the output transisitors. It may be usefull to try and turn the trimpots and see what the current does across R47 and R23. Does it go up rapidly and never come down? Make sure and be carefull that the output transistors do not get too hot or become overbiased too much. I would concentrate on one channel at a time here. It would still be usefull to know what the servo loop is doing, so we would need to know what the input and output voltages of the op-amp are doing. You may also try putting a load resistor on the output just to see what happens. Something on the order of 100 ohms would be good. I assume you have already done so, but you may want to look over the parts and values again to make sure everything is soldering in the correct spot. It sounds like things are close so I don't think you have a short.

Thank you for the help

I will be away for the week but will get to it when I get back

 

[Henmyr]

Measured voltage at opa2134.
Pin 1: -11.50V at first, but rising to -14.99V.
Pin 2: -1.22V, then changed to ~-0.20V
Pin 3: -1.22V
Pin 4: -15.38V
Pin 5: -8.20V
Pin 6: -7.8V
Pin 7: +2.5V
Pin 8: +14.97V

+15 measure ~+14.99 and -15 ~-15.38

All but tut the + and - 15V (both the measuring points and the op-amp V supply pins) varies a bit.

 

[amb]

Henmyr, looks like you have very severe output DC offset that cannot be corrected by the DC servo. Are your input JFETs (2SK170BL and 2SJ74BL) matched to each other?

Also, please use the troubleshooting PDF found in the "initial setup" section of the CK²III website and measure your actual board voltages and compare them against the diagram. Aside from the opamp area, where else does it show a large deviation from the listed voltages?

 

[Chaos4]

I don't think I understand how pin 6 measures -7.8 volts. How is any current being devoloped across R15? I would expect Pin 6 to set the reference ground for the op-amp.

 

[Henmyr]

Hi. Yes they are supposted to be matched. Kit from Glassjar Audio.

I've measured every voltage for the left channel and here are the readings:


Between JFET and R3 and JFET and R5 it first read what I've written there, but they later stabilized at around 160mV for Q2 and -160mV for Q3. When I turned of the amp later and turned on it again Q2 read 40mV and Q3 -3xxmV.

I've measured the op-amp like this:

5- -4
6- -3
7- -2
8- o-1

Every measurement have been with one stick at the measuring point and the other stick at ground in my powerstrip (please tell me this is correct!).

 

[Chaos4]

Henmry,

All voltages should be measured with respect to star ground. If star ground and the ground you are using are the same potential all is good.

I looks like you are measuring the op-amp correctly.

I am going to suggest you try shorting across R15 (assuming you are working with the left channel) and see if you can get the voltages to stabilize to a single value. I don't think this should hurt anything, but you will want to be carefull, also this is just for test purposes. You will not want to leave the short there permanently. Unless AMB or someone else thinks this is a bad idea, it may at least settle the voltages down a bit so we can see what went wrong.