Calibration of Siemens U273a, the big ones

[tommia]

After recapping 2 working Siemens U273A things became bad. Self oscillation all over the place.  Who can help me with the calibration of it.
I measured all the transistors out of circuit, they are still good. One was a little out of spec, so I replaced it. Things became different but not better . So I've put the original one back in.

I disconnected the control amp and R7 (above the diodes). The amp still doesn't react like it should.

The voltages on the schematic are the ones I measure within less than 5% margin.

Please help.

 

[mjrippe]

If they worked without oscillation before you recapped them, then it is very likely that you made an error during the recap.  Check that you have the correct values and orientation (+/-) in every position.  Did you change any values or capacitor types?

 

[tommia]

I had one unit with a bad cap, it was not working. After the recap job it works perfect. I did the same with 2 working ones. They died on me. I used the same caps as original except 25 uf becomes 22 uf,... , all electrolytic

All 3units have the same caps.
I'm convinced there is no error there.

It is more something like impedance. The working one is stable all the way. The 2 bad ones start to behave strange when I point with my finger at some parts like C2. Measuring voltages with  my FLUKE multimeter makes things happening like noises and clicks.

 

[Michael Tibes]

Measuring voltages with  my FLUKE multimeter makes things happening like noises and clicks.

Which is probably totally normal during the measurement.

Michael

 

[tommia]

@Michael Tibes: not with the working one.

Ok here beta measurements of the original transistors:

TR1: 309
TR2: 365
TR3: 100
TR4: 47
TR5: 106
TR6: 143

replacing TR5 & TR6 with same type but both a beta from 75 didn't help.
replacing bcy66 with bc 550 and sst117/sst116 with bd140 didn't help. I took BC550 with a beta of 500, so within the specs of  a bcy66.

 

[mjrippe]

Not sure what you mean by impedance.  Pointing your finger at a capacitor should have no effect on the circuit though. 

Obviously something changed when you re-capped them.  Check one more time that all caps are correct polarity.  Also make sure all ground connections are good - this can cause oscillation as well.

 

[tommia]

One unit is working at 16V powersupply except for the headroom/distortion

 

[tommia]

unit is working without oscillation if I remove feedback resistor R39, like it was done in older units (see older schematics).
With oscillation I mean the output C12 oscillates between  ground and positive with a frequency of I would say 5Hz.

I don't understand this. Negative feedback must help to make an amp more stable and to lower the impedance. Please explain, if you understand.

 

[tommia]

If I add 6K8 to resistor R34 in the feedback loop the compressor is still stable. The disadvantage is that I have 6 dB more gain resulting in 6 dB more noise. Powered by a good powersupply these baby's have a  snr of 90 dB measured in protools. Datasheet speaks of 70 dB but it is definitely much more

 

[radardoug]

What your unit is doing is called motorboating, low frequency oscillation. It points to the filter caps being lower in value than they should be. This allows feedback through the power supply to couple between stages. Negative feedback under this circumstance can exacerbate the problem.

 

[tommia]

really there is nothing wrong with those caps in value and polarity.
To prove me wrong I  took R30 out and powered the first 3 transistors up with there own powersupply of 12,2 V  like mentioned in the schematic.
Different but stil the same problem.
I know, how stupid can it be, only 6 Transistors,. Resistors and caps are  all measured several times. 
After 30 years of electronics this is killing me.

 

[tommia]

feeding TR1  5,3V at R19 with an external powersupply (instead of 6,4V). Makes everything stable and smooth.

Also in all 3 units C8 sees a negative voltage of 10 mV, couldn't be a problem, but strange design.