Hi Thomas,
Very nice with your writing of hum in your blog.
Could you say some causes of Buzz and how to treat them ?
Regards,
QA.
Hi!
Did you try, what I suggested with elevating the heater potential?
I don't think you need DC heaters.
The high heater to cathode voltage is not healthy no matter if you heat with AC or DC. You should fix that. Then try with a different set of driver tubes as yours might have developed heater cathode leakage
Thomas
Did you try, what I suggested with elevating the heater potential?
I don't think you need DC heaters.
The high heater to cathode voltage is not healthy no matter if you heat with AC or DC. You should fix that. Then try with a different set of driver tubes as yours might have developed heater cathode leakage
Thomas
Highly recommend DC heaters
1) It is the best way to reduce hum.
Not only do you eliminate the induction between heater and grid, you can also better cope with Rfg and Cfg.
AC hum can also be introduced with twisted wires, through Rfg and Cfg, and induction of the heater on the grid.
2) No need to break the PCB.
3) Only need about 3 Watts extra per channel.
4) There are many low voltage drop regulators.
5) No need to change valves.
No matter if you use AC or DC, since the input valve use SRPP topology, you must elevate the heaters about 70V.
Given the maximum operating conditions of the 12AX7 this is not strictly necessary, but better.
1) It is the best way to reduce hum.
Not only do you eliminate the induction between heater and grid, you can also better cope with Rfg and Cfg.
AC hum can also be introduced with twisted wires, through Rfg and Cfg, and induction of the heater on the grid.
2) No need to break the PCB.
3) Only need about 3 Watts extra per channel.
4) There are many low voltage drop regulators.
5) No need to change valves.
No matter if you use AC or DC, since the input valve use SRPP topology, you must elevate the heaters about 70V.
Given the maximum operating conditions of the 12AX7 this is not strictly necessary, but better.
Last edited:
Hmmm... yes, I suppose so.
Three small recommendations for using DC heating.
[1] use silicon full-wave bridge topology. This doubles the number of rectified pulses, reducing size-of-capacitor-inductor chain requirement substantially.
[2] If working with 6volt heaters, then use a 470 uF initial cap, a string of 3 silicon rectifiers in series (voltage drop without a regulator), and another 4,700 uF cap for smoothing. Or use a voltage regulator - they're pretty darn cheap, and the results will be better. But the 4 to 5 amp silicon rectifiers are clearly the cheapest.
[3] Consider having a small (0.5H) inductor in front of that string of rectifiers. They're so darn good at getting rid of the higher harmonics that come from silicon-rectification. And they're pretty cheap, too.
If you're working with 12 volt heaters, then 6 silicon rectifiers in series do the voltage-drop trick.
Do NOT use LEDs... since the power (current) is so much beyond their nominal rating. They're fine for the current fad of cathode-fixed-bias, but really ... that's all.
GoatGuy
Three small recommendations for using DC heating.
[1] use silicon full-wave bridge topology. This doubles the number of rectified pulses, reducing size-of-capacitor-inductor chain requirement substantially.
[2] If working with 6volt heaters, then use a 470 uF initial cap, a string of 3 silicon rectifiers in series (voltage drop without a regulator), and another 4,700 uF cap for smoothing. Or use a voltage regulator - they're pretty darn cheap, and the results will be better. But the 4 to 5 amp silicon rectifiers are clearly the cheapest.
[3] Consider having a small (0.5H) inductor in front of that string of rectifiers. They're so darn good at getting rid of the higher harmonics that come from silicon-rectification. And they're pretty cheap, too.
If you're working with 12 volt heaters, then 6 silicon rectifiers in series do the voltage-drop trick.
Do NOT use LEDs... since the power (current) is so much beyond their nominal rating. They're fine for the current fad of cathode-fixed-bias, but really ... that's all.
GoatGuy
Thomas, I wanted to try DC heater volatage first to check if it is heater where the problem is. And it actually was heater. With DC there was no 100 Hz hum.
The next thing I did was to uplift CT resistor networks (center tap) from GND to divider between +B and GND (100k + 10k wth parallel 220 cap) and it works! So now I use AC heater as in original circuit but the center of the CT resisotrs is not to GND but to higher voltage potential.
I did it for both channels and it works. This is great news! 100 Hz hum is gone now.
Now I have another issue - when I tried above mentioned modification the PCB and transformers were out of the chassis. Perfectly silent - not 100 Hz or 50 Hz hum.
When I put it back to chassis there is a 50 Hz hum in one channel. Not 100 Hz as it was before but 50 Hz and only in one channel. Not sure if it is sine signal 50 Hz or some burst but it is clearly lower frequency than it was before so I guess it is 50 Hz. What is strange is that it is one channel only.
I have tried different combinations (swapping input tubes, only one input tube in left channel, only one input tube in right channel, tried both input tubes) but the 50 Hz hum is still in the same channel and still in only one.
However when I remove both input tubes the hum is gone.
Any idea what can be wrong?
The next thing I did was to uplift CT resistor networks (center tap) from GND to divider between +B and GND (100k + 10k wth parallel 220 cap) and it works! So now I use AC heater as in original circuit but the center of the CT resisotrs is not to GND but to higher voltage potential.
I did it for both channels and it works. This is great news! 100 Hz hum is gone now.
Now I have another issue - when I tried above mentioned modification the PCB and transformers were out of the chassis. Perfectly silent - not 100 Hz or 50 Hz hum.
When I put it back to chassis there is a 50 Hz hum in one channel. Not 100 Hz as it was before but 50 Hz and only in one channel. Not sure if it is sine signal 50 Hz or some burst but it is clearly lower frequency than it was before so I guess it is 50 Hz. What is strange is that it is one channel only.
I have tried different combinations (swapping input tubes, only one input tube in left channel, only one input tube in right channel, tried both input tubes) but the 50 Hz hum is still in the same channel and still in only one.
However when I remove both input tubes the hum is gone.
Any idea what can be wrong?
That's great news, Vcelkamaja! Sounds like you are on the right track!
Some thoughts while we wait for Thomas to weigh in.
1) Does your transformer have two separate 6.3VAC windings? You mentioned you put the divider into both channels. If you have only one 6.3VAC winding then you would only need a single divider. It will bias all of the filaments through the transformer, you don't need a second divider on the other channel.
2) If you do have two separate windings, and thus need two divider networks, then is it possible that a wire popped off of the divider in the channel with hum? Based on your postings, the hum was gone with the cover off but re-appeared in one channel when you installed the cover. Could be that a connection was disturbed when you put the cover on, and that the humming channel no longer has filament bias?
3) If you have only a single 6.3VAC winding, or if the bias networks are still both running fine, then it's possible that the bias is not quite sufficient to overcome the hum in one of your channels. The resistor values you provided would give you about 23V of bias (260*10/110), Thomas had earlier recommended 70-ish volts (which is well within the 200VDC Cathode-Filament rating of the 12AX7).
The goal of adding filament bias in this case is to move the filament electrically closer to the upper-triode cathode voltage. The schematic indicates that the upper-triode (in the SRPP) is at about 155V potential, and the AC on the filament is likely modulating the upper cathode to introduce the hum in your output when the filament is substantially below this cathode voltage.
Moving the filament 23-ish volts closer to the cathode helps, moving it to 60-70V should help more. It seems to be a rare case of "more-is-better"
Increasing the 10k in your divider to 39k will get you close to 70V. If you string three 10k resistors in series you will get about 60V of filament bias. Depends what you have in your junk-box. The 220uF bypass cap can stay the same.
Good luck!
Last edited:
Hi!
Excellent! Glad we made some progress!
If it works without hum with the PCB outside the chassis, but started to hum with the PCB installed back in the chassis, there are two possibilities:
- As has been suggested before: Maybe something got damaged by installing the PCB again. When you take the PCB out again, does the hum go away?
or:
- There is stray fields from the power transformer and it impacts only one channel, maybe some sensitive parts of the humming channel are closer to the transformer?
Best regards
Thomas
Excellent! Glad we made some progress!
If it works without hum with the PCB outside the chassis, but started to hum with the PCB installed back in the chassis, there are two possibilities:
- As has been suggested before: Maybe something got damaged by installing the PCB again. When you take the PCB out again, does the hum go away?
or:
- There is stray fields from the power transformer and it impacts only one channel, maybe some sensitive parts of the humming channel are closer to the transformer?
Best regards
Thomas
Thanks for your ideas and comments.
Some more observations - I have spent some time listening to the hum and it is more likely a complex signal. I comes out from the tweeter rarher than from the woofer and seems to be 50 Hz modulated with some higher frequency noise.
One thing I tried now was to increase resisotrs in the B+ to GND divider for heater elevation from 100 K + 10 k to 100 k + 20k and funny thing happened - this hum disappeared from the original channel and it is in the other channel now So what does it mean?
I will try to elevate the heater voltage further more to 100k + 30 k and see what happens.
Actually is it possible to use different elevation volatage in left and right channels? e.g. 100 k + 10 k in one channel and 100 k + 20 k in the second channel?
Would it make sense to use a kind of hum pot but in this case the CT network is not connected to GND but the divider described above.
Some more observations - I have spent some time listening to the hum and it is more likely a complex signal. I comes out from the tweeter rarher than from the woofer and seems to be 50 Hz modulated with some higher frequency noise.
One thing I tried now was to increase resisotrs in the B+ to GND divider for heater elevation from 100 K + 10 k to 100 k + 20k and funny thing happened - this hum disappeared from the original channel and it is in the other channel now
So what does it mean?I will try to elevate the heater voltage further more to 100k + 30 k and see what happens.
Actually is it possible to use different elevation volatage in left and right channels? e.g. 100 k + 10 k in one channel and 100 k + 20 k in the second channel?
Would it make sense to use a kind of hum pot but in this case the CT network is not connected to GND but the divider described above.
Hi!
Before just blindly trying out stuff, did you check as suggested if the hum is still present when the PCB is out of the chassis? It can make sense to increase the potential of the heater voltage closer to the upper cathode potential. But I would not go to high, stay below 100V. Also make sure the resistance to ground does not get higher than 20k.
Check also all the solder joints, could be that some are bad. Especially make sure the bypass cap of the lower resistor of the voltage divider is properly soldered in.
I'd also try a different set of 12AX7. They probably developed some heater to cathode leakage from operating them with a high Ufk for some time.
Best regards
Thomas
Before just blindly trying out stuff, did you check as suggested if the hum is still present when the PCB is out of the chassis? It can make sense to increase the potential of the heater voltage closer to the upper cathode potential. But I would not go to high, stay below 100V. Also make sure the resistance to ground does not get higher than 20k.
Check also all the solder joints, could be that some are bad. Especially make sure the bypass cap of the lower resistor of the voltage divider is properly soldered in.
I'd also try a different set of 12AX7. They probably developed some heater to cathode leakage from operating them with a high Ufk for some time.
Best regards
Thomas
Well, it seems it is not the level of elevation voltage as I tried 30k and beck down to 20k and it still the same.
I tried to put the PCB out of the chassis (with respect to wires that limit this) and there was no change.
I did some experiements with input tubes and it is clearly only one channel. I re-soldered all points on the PCB but no change.
However when I was experimenting and moving the PCB at one moment the hum (or more buzz) disapeared. It was actually when I put PCB into is original position in the chassis. I thought I solve the problem but when I powered it off, fixing the PCB with screws and powered it on the buz is again there (still the same channel). Then I unscrewed the fixing screws that hold the PCB but no change, still buzz.
It seems like there a bad wire or screw terminal or something like this but I can't find it. I'm trying to move the wires while amp is running but no change.
I tried to put the PCB out of the chassis (with respect to wires that limit this) and there was no change.
I did some experiements with input tubes and it is clearly only one channel. I re-soldered all points on the PCB but no change.
However when I was experimenting and moving the PCB at one moment the hum (or more buzz) disapeared. It was actually when I put PCB into is original position in the chassis. I thought I solve the problem but when I powered it off, fixing the PCB with screws and powered it on the buz is again there (still the same channel). Then I unscrewed the fixing screws that hold the PCB but no change, still buzz.
It seems like there a bad wire or screw terminal or something like this but I can't find it. I'm trying to move the wires while amp is running but no change.
I have some news What I tried now is to replce input tubes. I didn''t wanted to do this because:
-Tung Sol 12AX7 I have are brand new so I thought they must be good
- I don't have any spare 12AX7
Anyway I take the risk and repaced 12AX7 with old pair of ECC82. And... no hum or buzz!
I know these tubes are different to 12AX7. ECC82 is actually 12AU7. Anyway I wanted to try it and it seems there is really no hum. It seems it works and sounds good. Maybe a bit quiter. Actuallt after a few minutes there are some crazy sounds - pops but I believe this is because of the different tube type and as I mentioned these are pretty old tubes and I don;t know their condition.
Do you think does it make sense what I tried and is it possible that with different pair of tubes there is no hum? If so I will order some new 12AX7, maybe two or three pairs to check if it helps.
What I tried now is to replce input tubes. I didn''t wanted to do this because:-Tung Sol 12AX7 I have are brand new so I thought they must be good
- I don't have any spare 12AX7
Anyway I take the risk and repaced 12AX7 with old pair of ECC82. And... no hum or buzz!
I know these tubes are different to 12AX7. ECC82 is actually 12AU7. Anyway I wanted to try it and it seems there is really no hum. It seems it works and sounds good. Maybe a bit quiter. Actuallt after a few minutes there are some crazy sounds - pops but I believe this is because of the different tube type and as I mentioned these are pretty old tubes and I don;t know their condition.
Do you think does it make sense what I tried and is it possible that with different pair of tubes there is no hum? If so I will order some new 12AX7, maybe two or three pairs to check if it helps.
Hi!
ECC82 is a totally different tube. This test tells nothing. The ECC82 has much less gain and by that will amplify any hum around it much less than a 12AX7.
It is quite possible that your tubes have some excessive leakage from heater to cathode. So I would recommend to try new ones, but 12AX7 or equivalent.
Still your description tells me that you have some problem, which should be fixed.
Thomas
ECC82 is a totally different tube. This test tells nothing. The ECC82 has much less gain and by that will amplify any hum around it much less than a 12AX7.
It is quite possible that your tubes have some excessive leakage from heater to cathode. So I would recommend to try new ones, but 12AX7 or equivalent.
Still your description tells me that you have some problem, which should be fixed.
Thomas
Thomas, I agree that ECC82 is a completely different tube. I will try to get some 12AX7/ECC83 and try it again.
I really don't know what else I should try. What is strange is that there is this strange buzz in one channel while the second one is completely silent (there is not even a low level of this buzz like in the other channel).
I was listening carefully very close to loudspeaker and actually if I unplug the input tube the strange buz is still there. It is very very low level - almost impossible to hear it but it is there. When I plud the input tube it is is much stronger and I can hear it while I'm ~ 1 m away from the speaker.
But still only in one channel. The second channel is perfectly silent.
I really don't know what else I should try. What is strange is that there is this strange buzz in one channel while the second one is completely silent (there is not even a low level of this buzz like in the other channel).
I was listening carefully very close to loudspeaker and actually if I unplug the input tube the strange buz is still there. It is very very low level - almost impossible to hear it but it is there. When I plud the input tube it is is much stronger and I can hear it while I'm ~ 1 m away from the speaker.
But still only in one channel. The second channel is perfectly silent.
Hi!
Have you tried what was suggested above? Maybe desolder the voltage divider and cap of the affected channel completly and put in a new one. What you describe looks like an intermittent failure somewhere. You really should fix this.
With new tubes the noise might be gone first, but then creep back
Best regards
Thomas
Have you tried what was suggested above? Maybe desolder the voltage divider and cap of the affected channel completly and put in a new one. What you describe looks like an intermittent failure somewhere. You really should fix this.
With new tubes the noise might be gone first, but then creep back
Best regards
Thomas
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.