Issues with clipping in vintage Neumann microphone preamp.

[Jakobz]

Hi there, i hope i posted this in the correct forum

I have a issue with a Neumann microphone preamp .I have a feeling i have all the clues to figure this out, but i guess my understanding of how the circuit works is not adequate, so i hope some of you would take a look and point me in the right direction Here is what i know:

1. The sound from the output is thin and distorted.
2. When i run a testsignal through the the unit (1Khz sinus) the negative cycle is clipping, measured at the output. Tracing back the signal it seems the clipping occurs by Trs2/Trs3.
3. I have noticed that when the unit is powered up, for 2 seconds or so the signal looks good, and then slowly the negativ cycle flattens out. I have also noted that on power up the collector dc voltage on Trs2/Trs3 starts on 14vdc and then slowly drops to around 9,5vdc. Am i right in assuming that the collector voltage should be around half of the supply voltage, and that it being 9,5 would explain why the signal is being clipped?

I hope someone can point me in the right direction so i can getter a better understanding of these kinds of circuits. i have attached the schematics. PV76N_Schematic.pdf PV76N_Schematic.pdf PV76N_Schematic.pdf

Please let me know if you need anymore information or measurements, Thanks for looking.

[Helmholtz]

When i run a testsignal through the the unit (1Khz sinus) the negative cycle is clipping, measured at the output. Tracing back the signal it seems the clipping occurs by Trs2/Trs3.[*]I have noticed that when the unit is powered up, for 2 seconds or so the signal looks good, and then slowly the negativ cycle flattens out. I have also noted that on power up the collector dc voltage on Trs2/Trs3 starts on 14vdc and then slowly drops to around 9,5vdc. Am i right in assuming that the collector voltage should be around half of the supply voltage, and that it being 9,5 would explain why the signal is being clipped?

That makes sense.
First suspects are leaky ecaps.

Start with lifting one end of C5 and see if the collector voltage recovers/stays stable.

Actually it would make sense to replace all of these ~50 years old electrolytics.

[glebert]

Are the capacitors original? I'm thinking maybe a leaky cap somewhere around TRS1,2, but that is just a guess. I would look at base, collector, emitter voltages for TRS1,2,3 without signal to see if you can see what is going on.

It sounds like you have a scope, can you see at what point in the circuit it starts clipping?

edit: Helmholtz beat me to it.

[Jakobz]

I should have mentioned that all electrolytics, already has been replaced it seems the clipping starts at the collector on tr2/tr3. I will try to do some more measuring so I can give you some more details

[Helmholtz]

I should have mentioned that all electrolytics, already has been replaced it seems the clipping starts at the collector on tr2/tr3. I will try to do some more measuring so I can give you some more details

Not all new ecaps are good, especially when overstored.
Also check correct polarity.

[Jakobz]

Thank you so much for your input! I just checked dc voltages with Cap C5 disconnected, and the voltage still drops slowly from 13-14v to around 9,5v where it stabilizes. Checked polarity on all ecaps and they all seem to be correct.

[Helmholtz]

Are supply voltages at C6 and C7 stable? Values?

As glebert advised, please post C,B, E voltages of input transistors. No signal applied.

[Jakobz]

Voltage at c6 is stable, i am not sure abount c7 actually, will check tomorrow.
Voltages at transistors (with reference to 0v) input voltage 23v:.

tr1
B	3,38
E	2,8
C	5,94
Tr2
B	5,94
E	5,35
C	9,5
Tr3
B	5,3
E	4,7
C	9,5
Tr4
B	9,5
E	8,8
C	19,73

The supplyline settles almost instantly but the collector/emitter voltages takes several second to settle, i will see if i can take some screenshots from the scope.
I am a little confused about the dc voltage on the base of Tr1 since its only connected to the transformer, where is it coming from?

[Helmholtz]

Trs1 is biased by the voltage across C1, being fed from the Trs 3 emitter voltage via a resistor voltage divider arrangement.
This is a DC NFB path. Trs 1 bias can be adjusted via R6.

Trs 4 is an emitter follower, its collector is supplied by the voltage across C7 and zener diode Gr 1. As this is a 22V zener the voltage looks somewhat low.

[Helmholtz]

It should be possible to adjust for a symmetrical output using R6.
But don't feed too large an input signal.

[Jakobz]

Trs1 is biased by the voltage across C1, being fed from the Trs 3 emitter voltage via a resistor voltage divider arrangement.
This is a DC NFB path. Trs 1 bias can be adjusted via R6.

Of course thank you for clearing this up for me.

Trs 4 is an emitter follower, its collector is supplied by the voltage across C7 and zener diode Gr 1. As this is a 22V zener the voltage looks somewhat low.

Good point, I will look into this.

It should be possible to adjust for a symmetrical output using R6.
But don't feed too large an input signal.

I see, is that because r20 in parallel with r6,r7,r8 determines the current going through Trs3 and thereby the drop over r22?

[Helmholtz]

I see, is that because r20 in parallel with r6,r7,r8 determines the current going through Trs3 and thereby the drop over r22?

I'd simply say, the voltage divider consisting of R6+R7 and R8 determines the percentage of the Trs3 emitter voltage which is fed back and used as bias for Trs1.
R6 affects the voltages of all 4 transistors.
Once the bias is set via R6, the NFB should stabilize operating points of all directly coupled transistors.

But a leaky ecap might disturb the circuit and cause drift.
Also check R6 for a bad/intermittent wiper contact.

Did you try to adjust R6?

[Jakobz]

It seems that i was able to get rid of the clipping by adjusting r6, so far so good Thank you for your insights, this has been very helpful and educating Theres still something up with the sound though, it seems to be lacking low end. I have noticed that a 100hz test signal looks kind of "skewed" measured at the output(i attached a screenshot). The signal is getting gradually better looking as i raise the frequency. I am tracking it through the circuit and it looks good all the way up to transformer U2. Will continue tomorrow hopefully.

[Helmholtz]

Can you improve the signal (symmetry) by adjusting R40?

[Jakobz]

Sorry for the late reply, it seems that i was able to adjust for better symmetry by adjusting R40, and as far as i can tell the unit sounds good now. Thank you for helping me out