Transistor Distortion

[michinger]

Generally i am more a reader then writing threads here, but this time I need some help.
Would be nice if the great people here could help me.

I try to repair a realtive simple headphone section of an Mixer:



The Signal on the HP jack is good until it reaches a middle loud volume, then you can hear distortion.
Not really bad, but not good enough either.

The signal is fine before the R31 and R32. After them the distortion starts, so it is about the 2SC2240 and 2SA970
I already changed that parts (also the resistors) but nothing changed.
The voltage there is 16,7V which should 16V but that can´t the problem I think.
Power supply is fine because everything else works, also other outputs of the mixer. So the voltages seems fine.

Would be nice if someone have ideas how to find the issue.
Thanks in advance!

 

[5v333]

If you lower the amplitude of the incoming signal, is it still distorting at half volume setting?

 

[JohnRoberts]

Troubleshooting 101... for a basic circuit like that there are two approaches.

#1 test all the transistor junctions with a VOM diode drop scale... or #2 measure DC voltages around the circuit.

Op amps will have well defines output states based on input pin voltages. Transistors should have predictable junction voltage drops.

JR

 

[Ike Zimbel]

What impedance headphones are you using? Maybe the amp is simply running out of gas...

 

[michinger]

If you lower the amplitude of the incoming signal, is it still distorting at half volume setting?

Has nothing to do with the input. Sure, I could try to lower the output of the TL072, but originally it should work how it is.

 

[michinger]

What impedance headphones are you using? Maybe the amp is simply running out of gas...

tried an AKG 600Ohm and Beyer DT990 (250Ohm) with the same result

edit: I read in the manual: Output impedance is 70 Ohms but will drive headphones as low as 8 ohms.

 

[5v333]

I didnt think about it enough...
My thoughts was that the gain of the opamp (18db) was making The signal to hit the absolute headroom.

 

[michinger]

Troubleshooting 101... for a basic circuit like that there are two approaches.

#1 test all the transistor junctions with a VOM diode drop scale... or #2 measure DC voltages around the circuit.

Op amps will have well defines output states based on input pin voltages. Transistors should have predictable junction voltage drops.

JR

#1: the transistors are all new and it sounds exactly the same as with the originals
#2: I did, the TL072 gets the right DC and also the transistors

How to measure the cleanness of DC?

 

[radardoug]

Its a pretty crap circuit. They are relying on the opamp to reduce the crossover distortion.
There is no allowance for differing VBEs on the transistors.

 

[Bo Deadly]

It could be a lot of things. Even if everything else is working, this amp is unique in that it only has 68 ohms between the rails and the load. So if there's a problem with the supply, you would probably hear it. It could be just the topology of the amp. It could be a fault in the circuit that is stressing the TL072. Realize that with a 250 ohm load, V = I * R so if you put say 20mA into 250R, that's 5V to make 100mW. Meaning the higher the impedance of the hphones, the higher the voltage, the more stress on the op amp. It should be able to handle it but maybe there's a bad solder joint somewhere and at some voltage it starts to shunt current. If you press on the PCB with a chopstick, does the distortion go away? If you blow air-duster directly on some part, does the problem go away? Do parts in the power supply like rectifiers or transistors look discolored from heat? It could be a lot of things ...

The path of least resistance to fixing this is to first measure the distortion so that you can prove that the problem truely exists and characterize the distortion and prove that some test or change has an effect. Specifically, you need to make a dummy load with a resistor on a breadboard and a cable that feeds signal into a USB audio interface (actually two resistors because you will need to attenuate for the USB interface). Then use whatever software like Wavepad to generate a tone to send into the unit and whatever FFT or spectrum analyzer software to view the output from said breadboard dummy load. Then you should be able to visualize the distortion. If you get that far, post the results here and I bet you'll figure it out.

 

[musipol]

My question is: Did it ever sound good?

Looking at that circuit, it seems to me you will always hear crossover dist due to the fact there is no static bias for those transistors. It appears to bias them only when the signal voltage drops approx. +/- 0.65Volts across that 4k7 resistor. I know FB should reduce the crossover some, but this is still a pretty cheesy setup.

 

[JohnRoberts]

My question is: Did it ever sound good?

I ASSume it was a working design

Looking at that circuit, it seems to me you will always hear crossover dist due to the fact there is no static bias for those transistors. It appears to bias them only when the signal voltage drops approx. +/- 0.65Volts across that 4k7 resistor. I know FB should reduce the crossover some, but this is still a pretty cheesy setup.

A TL07x can slew 13V per uSecond so it doesn't take long to slew from - one diode drop to + one diode drop all inside the NF loop.

Agreed not audiophile but serviceable for relatively slow moving audio. The popularity of that approach is that all the output devices are fully cut off when passing through zero so no mutual conduction or excess heat dissipation. (At least I would use a lower value than 4.7k resistor.)

Perhaps my advice was not understood... I don't care how new or old the transistors are, do their junctions measure good or bad? Like wise with op amp circuits the output voltage can be predicted by + and - Input voltages. Troubleshooting 101 is to confirm that the active devices are healthy and behaving as their input voltages suggest.

JR

 

[michinger]

I made some measurements.
It seems the THD is always on the Headphone Out, just not hearable on low volumes.
Anyway, here some results:

Headphone Out without Load (no headphone connected):


Same (relative low) volume but with the HP (250 ohm) connected:




Now HP Volume on max. The THD do not exceed over the -72dB in my case:



Here the Output of the TL072 which is fine as I stated above:


Maybe with that information you have some other ideas now?

 

[JohnRoberts]

I find it hard to believe that distortion products down -70dB are a huge audible issue. The difference between no load and 250 ohm suggests the class B design may be responsible. To convert that design to class AB involves adding more components and complexity.

If I could impose again, could you try more measurements with the 4.7k reduced to say 470 ohm?

[edit- For today's TMI about that circuit, to make the nonlinearity look/measure even worse reduce the input down to mV level and high audio frequency (say 20kHz). That topology is similar to active rectifiers where the op amp has to slew from - one diode drop to + one diode drop while passing through 0V. The op amp's open loop gain becomes noticeable at low input level and high frequency. [/edit]


JR

 

[5v333]

But it looks like the distortion relative reference tone gets worse with low volume. So its not - 70db distortion and downwards at lower level. Its actually around - 45db in 2:nd pic. Isnt it clearly a non linear behaivior? Not good for ears...

I guess the domination of odd harmonics is also a sign of crossover distortion.

 

[ruffrecords]

I would say those spectra are exactly what you would expect from such a simple output stage. If you want less distortion you probably need to run some small bias curent through the output transistors. Check out Doug Self's headphones amplifier.

Cheers

Ian

 

[Bo Deadly]

Same (relative low) volume but with the HP (250 ohm) connected:
<snip>
Here the Output of the TL072 which is fine as I stated above:

Is the TL072 output with the 250 ohm load?

I like JRs suggestion at this point. The 4K7 is really high. My instinct would be to use a load appropriate for the TL072 which would be more like 1K but because the transistors should kick in on when the TL072 is putting out only a few mA, a 470R is good.

Then take measurements again with the 250R load. If that improves the output significantly, that would suggest a topology issue.

If not, it could still be a supply issue or a bad solder joint in a ground wire or similar.

 

[JohnRoberts]

I like my suggestions too... This is classic "crossover" distortion and why audio power amps do not use simple class B.

In fact just like class B amps it can measure good at high levels but at low levels and high frequency it will get crunchy.

In my judgement a full class AB conversion is too much work... Try the 470 ohm resistor (or lower). If the TL072 runs out of gas A NE5532 will drop in that socket.

KISS

JR

 

[blakeyboy]

Stop.

Get a good clean variable oscillator, and a simple oscilloscope.
Try different resistive loads to start with.

Then you can begin to repair/change the circuit.
Everyone is trying to fix it from diagnostic steps 4, 5 and 6, having decided not to do diagnostic steps 1,2 and 3.

Simple stuff first. _Always._

For example:
If the waveform is clipped on just the positive peak, or just the negative peak, and would sound very similar to what you have described, and would look similar on a harmonic display,

none of you would know......

 

[Bo Deadly]

Stop.

Get a good clean variable oscillator, and a simple oscilloscope.
<snip>
If the waveform is clipped on just the positive peak, or just the negative peak, and would sound very similar to what you have described, and would look similar on a harmonic display,

none of you would know......

I don't think we would learn anything from looking at the waveform in an oscilloscope in this case because the distortion is 40 dB down. Even on a 10 bit scope you probably wouldn't be able to see that.

And I think we've already figured it's crossover distortion so you're not going to see clipping of peaks anyway.