Power trans hum heard through speaker

[bigguy12321]

Soooo... I got... ticked enough and I went to the shop to get my noise finder thingy.

Good troubleshooting technique, split the circuit.
Checked aft end of 22k first. Dead quiet.

Split the remainder.

Pre 10k. NOISE!!! BZZZZZZ!!!! But expected there. Raw DC.

Post 10k. Hmmmmmmm. Just like out of speaker. This is also at the screen of the 6V6.

Recheck post 22k. Dead quiet still.

Remove 6V6. All is good with socket.

Inspect up close (again) solder lugs of octal socket. Nice and clean.

Add another filter stage ala Tweed Princeton, all of a sudden the amp remembers the words... No more humming!
I don't know. Evidently this one needed more.

2L man wins the pasta toss for most accurate toss.
Corliss wins for goodest stumping questions.
King Fan wins for bringing pasta and beer.
Everyone else gets a certificate of participation and my undying gratitude.

I still don't know why this amp pulled this stunt but at least it was save-a-ble.

Thanks to all who played along. I hope it helps someone else in the future. Remember, search is your friend.

Aaron

 

[corliss1]

To make your life easier next time, confirm the frequency of the hum first. You would have quickly found this was a 120Hz hum, which means it's happening in the power supply portion of the amp and not the transformer itself. Tacking in an extra cap would have been a quick test after you knew the frequency.

 

[bigguy12321]

To make your life easier next time, confirm the frequency of the hum first. You would have quickly found this was a 120Hz hum, which means it's happening in the power supply portion of the amp and not the transformer itself. Tacking in an extra cap would have been a quick test after you knew the frequency.

Ok. Please and thank you, for my clarification, when you say tacking in an extra cap, would it still NEED the resistor between the next stage? I'm thinking yes... I jumpered over the first and second stage separately with a separate 20uf cap thinking I may have had a bad cap but there was no difference. I even swapped the first two caps out with new eventually.

I'm pretty sure the C-R-... was necessary. Right?

Any ideas why this time the 4th cap was needed?

Thanks again for the follow up.

Aaron

 

[reddart]

I'm pretty sure the C-R-... was necessary. Right?

Any ideas why this time the 4th cap was needed?

Thanks again for the follow up.

Aaron

On a “normal” push-pull amp, both the OT/plate nodes and screen nodes of the power tubes will cancel power supply noise picked up from the rail to a significant degree (and possibly even a LTP PI would too, I presume) whereas a single ended amp does not have this cancellation in the first 2 nodes. One of the disadvantages of a single ended amp. The extra filter stage lowered the power supply ripple going into the plate and screen of the power tube.

At least that is what I assume is happening.

 

[2L man]

Actually now I did succest to test the capacitor between B+1 and cathode!

However I have lately installed three to five filter capacitors before B+1 OT feed and resistors between them and then SE amp hum less. I have also test resistors on current return up to B01 but can't say does it improve filtering still because amps have been different.

 

[corliss1]

You may or may not have needed the extra resistor - it's impossible to say without testing. Sometimes just bumping up the capacitor of one stage is enough.

When you add an extra stage with a resistor at the beginning, as a lot of builders do with single-ended amps to help reduce the hum, all your power supply nodes after that point get a bit quieter, but you may need to adjust your resistor values to get your expected voltages at all points.

 

[2L man]

Mains based hum/brum is often double the Mains frequency and there are two significant mechanisms.

It can happen electromagnetically when Power Transformer EM stray field interact with Outout Transformer. Then current limiter resistor(s) can help when current peaks are not as strong. Rectifier Tube efficiency is poor and thats why it naturally lessen EMI. When Solid State rectifier diodes are used the current peaks come higher and then installing voltage dropper resistors to both Anode circuits simulate RT and then the PT EMI comes lower back to what RT use did produce.

Anogher mechanism is when HV current drop the voltage between rectified sine wave peaks, this change the bias current double the mains frequency rate. Then increasing filter capacitor value make brum lower but incrsease EMI so that might not help?

So using series reactance like Rectifier Tube or series resistor(s) lessen both of these hum mechanisms but they increase PS sag what changing output power produce. Using few more RC filters or RCR filters often is good answer. RCR filter has another resistor also on return current path.

Choke between rectifier and first filter capacitor make current flow significantly longer time and it lessen PT EM Stray Field. HV DC comes lower but HV regulation improve for different current. However the Choke has EM Stray Field and that can interact with OT. High current, high inductance Chokes are also expensive and heavy.

Aligning PT, OT and Choke using "headphone trick" is good practice.

Nowadays MosFet regulator regulate the HV very smooth and there does not come bias current brum and this is quite common in HiFi builds when SS diode rectifier is used. Selecting higher voltage output PT allow
higher value voltage dropper resistors and PT EM Stray Field comes lower. Often when the same PT which was meant for RT use there comes enough voltage to be lost to resistors and MosFet.

Toroidal PT EM Stray Field is much weaker than EI-core PT EMST and it is strongest straigth out of hole so mounting the Toroid flat to Chassis often make EMI insignificant low.

 

[Lowerleftcoast]

Ok. Please and thank you, for my clarification, when you say tacking in an extra cap, would it still NEED the resistor between the next stage? I'm thinking yes...

This rectified AC is smoothed with First Order Low Pass RC filters. R (resistance) C (capacitance). A First Order Filter has a 6dB per octave slope. A Second Order Filter has the same cutoff frequency (corner freq... etc.) but it has a 12dB per octave slope.


When you simply parallel one of the capacitors with another cap, the cutoff frequency is lowered so there is more filtering. If the caps in question are the same value, the cutoff frequency will be shifted down one octave so, all frequencies above the cutoff frequency will be attenuated by 6dB more when the second cap is in parallel.

When an additional RC filter is added (with the same resistance and capacitance) a Second Order Filter is created. The cutoff frequency will remain the same but the frequencies above the cutoff frequency will be attenuated at 12 dB per octave.

So either way the 120 Hz content will be reduced but there would be more filtering of the 120Hz content with the Second Order Filter all else being equivalent.

 

[bigguy12321]

Whoa! Leave for a couple days and viola!!!
There are some great ideas for future consideration here.
All of this makes sense except I've done this (not exact but REEEALLY close) circuit at least a half a dozen times prior and haven't had this problem.
Predicting the problem would be helpful but at least now I'll be aware that it could happen and know what it'll take to get rid of it.
I'm not sure if it would be prudent to just add the extra filter in the future and be done with it so I never have the problem again. I tend to like the response and sound I get without it. I'll have some people I trust play it for a while to get thier opinion as I do not have a good enough ear to hear any subtle changes.
Again, thanks to all who checked in here. I really do appreciate it.