K4040 distortion on right channel


#1

Hi
I have just assembled my K4040. All dc voltages on both channels check out close to the specified values, and they are virtually identical between the left and right channels. All bias adjustments are settable and the voltage on the output tube cathodes can be set very close to 0.4V.

The left channel is working properly. I checked it out with a function generator and oscilloscope (my trusty PCS500 combo) and it delivers not far from the rated 90W into 8R resistive load. Clipping is progressive and symmetrical. Frequency response measures as expected and a quick look on the PCS500 spectrum analyzer shows odd and even harmonics well down. I also have access to a Prism dScopeIII analyzer for more detailed performance measurements once I’ve sorted out the problem I have.

The problem is this: The right channel is not working properly. The output waveform degenerates from fairly low power, say a couple of volts output, resulting in distortion. There is also a very slight buzzing sound that can be heard on the output.
Here’s what I’ve checked:

  • Components are all installed as per the instructions and the component placement diagram.
  • Soldered joints are all good (I went over them just in case),
  • I have checked all resitor values in situ with an ohmmeter. All are what one would expect from an in-situ check, and resistor values between left and right channels are almost identical.
  • The tubes are OK. I established this by swapping the tubes from the left and right channels - the fault stays on the right channel.
  • I scoped the waveform at output of V10 (input amplier stage) of both channels by probing at the junction of R20/R21 (left channel) and R26/R27 (right channel). The waveform on both channels is the same at this point.
  • I scoped the waveform at the junction of R17/R84 (left channel) and R23/R87 (right channel). The faulty channel shows a much higher signal amplitude than the good channel at this point, and the waveform deteriorates markedly as input to the amplifier is increased.

This would suggest that there may be something amiss within the feedback loop, causing some kind of instability? Under ‘Troubleshooting’ on p23, the assembly instructions describe a mod of increasing R65 and R68 in value e.g. 1k or 1K5 which would reduce the negative feedback - I have not made this change yet, in case the problem is more simply solved.

Do you have any ideas on what may be causing this problem, and can you give me some pointers on where to look?

Thanks

John


#2

We’d also suspect the feedback loop or the output transformer.
You might want to swap them to make sure.


#3

I’ll try the feedback resistor increase first, as swapping the transformers is quite a tedious job. Will let you know in a couple of days what happens.
Thanks.
John


#4

Hi
I cleaned the PCB with IPA and a small brush and checked that there was no residue of any kind and no cracked tracks. This didn’t solve the problem, but the board is now immaculate. Then I removed the output transformers, checked their phasing with a function generator and scope, re-tinned their leads, and refitted them in opposite channels. Now both channels work properly. It’s annoying that I wasn’t able to pin down the exact cause of the problem, but it’s more than likely to have been a bad connection to one of the primary windings.

The next issue I need to deal with is a slight hum in the right channel. I’ve rotated the transformer as suggested in the assembly instructions, but was not able to reduce it to an imperceptible level by doing this. There are some posts on this forum and diyaudio.com which I’ll dig around, but any suggestions would be appreciated.

Cheers

John


#5

Without exact figures it is difficult to determine if the hum is excessive or not.
Tube amps are never as silent as transistor amps.
The pre-amp stage features a dual triode. One side of the triode is closer to the power transformer than the other.
Maybe this is at the cause. Added shielding could solve this problem.
Did you check the gain of both channels? Maybe there is a difference?


#6

Hello again
Fixed! I was able to eradicate the hum by the simple action of changing tube V11. The hum is now so low I can only just hear it very, very faintly if I place my ear right next to the speaker in an absolutely quiet room. Some posts about the K4040 on diyaudio.com suggested many of the ECC8x tubes can cause hum. The tube supplied in the kit was an Electro Harmonix, supposedly quite a reasonable brand. I replaced it with a NEC 12AU7 of 1970’s vintage I found in my junkbox - harvested from some radiogram when I was a teenager. I also tried a Raytheon 12AU7 I found that’s even older, inherited from my father’s junkbox. Both of these tubes worked well and did not cause hum.
Also, the midrange with the original Electro Harmonix tube was quite muddy and the overall detail unimpressive, whereas both the alternatives I tried were cleaner, with considerably more detail. The two channels are identical in both measurement and audition.

With my trusty PCS5000 I measured the output power as about 75W continuous sine per channel into 8R resistive loads, just before the onset of any visible non-linearity. The clipping is gradual and begins with a hardening of the sine crest in a kind of ‘linear interpolation’ before the onset of true ‘top-flattening’ clipping. Output is probably 90W only at the onset of true clipping, but this is not a big deal for me.

I am now very happy with my K4040 and will begin implementing a number of minor tweaks suggested by the diyaudio forum towards improved reliability and sonic improvement: change of ZD1 to better handle mains brownouts; Bourns cermet trimpots for the bias adjustment; filter cap on the LED bargraph input, and possibly JJ ECC83s for V9-11 (if they don’t hum!).

Thanks once again for your prompt response and guidance.

Cheers

John


#7

Happy to read it is solved.
Amazing how a simple tube swap can solve the problem.