AA1164 Princeton Reverb - 100Hz hum and noisy reverb

[gitarzysta]

Hello the best tube amp community in the world :)

I have just completed my build of Princeton Reverb clone. I did not follow any specific instructions, but it is pretty close to Rob Robinette version (with screen resistors and PI grid stopper added) - no other mods. I used Mojotone wiring diagram, cross checking with Rob and with a schematic - I am sure I got it right.

All works BUT there is some noticeable (not bad, but noticeable) hum - 100Hz (I live in Europe, so not 120Hz :) And yes I checked the frequency - it is definitely 100Hz).

I removed preamp tubes one by one and came to a conclusion that there's something around V4 (PI and tremolo) that causes this. V1 to V3 inserted (separately or all 3) - dead quiet. Only V4 inserted - some very minor hum (I'd call it not a problem). All 4 in - hum appears.

Update: 16th Feb: I read the reply to my thread, refreshed my brain and started again: removing V3 breaks the noise. V1, V2, V4 in, V3 out - no noise. V3 in - hum. In addition reverb pot increases it plus adds "white" noise.

I swapped tubes, poked the leads inside with a chop stick, I tried to isolate all reverb and tremolo jacks from the chassis and have them grounded to a different spot - nothing really helped. I rewired V3 socket to make sure the soldering is OK, no strands, no cold joints. Separated grid and plate wires. I checked all grounding points in the amp.

It is not major hum but it is noisier than my other push pull amps. Keeping in mind it is amplified as soon as V3 is inserted, any other ideas to make it super quiet?

Pictures are here:
https://gitarzysta.com/1_3_2_princeton-reverb-clone-aa1164-quot-blackface-quot-version.html

Other observations: without reverb tank connected turning up the reverb pot gets VERY noisy (all types of hiss, lots of static etc) - is this normal? Once the tank is connected increasing reverb adds some white noise, as well as 100Hz hum increases, some static still present, but much less with the tank connected.

Any help will be appreciated!

Thanks
Gitarzysta

P.S. Yet another observation: cable going from pin 7 of V3 (V3b grid to 3M3 resistor) is also "sensitive" to tapping - you can hear the tapping through the speaker even with volume at 0. I replaced it with screened wire, no change.

[pdf64]

Without a valve in V4, no signal can make it through to the power valves. I can’t see how any scratchiness be heard in that condition?
And no hum from preceding stages could pass.

With all valves in place and no signal input, ie idle, does the power valve anode or cathode current change according to the vol control setting, ie being turned    from one side of the scratchiness to the other.

[gitarzysta]

Scratchiness is not my main concern - the main one is 100Hz hum. I should have been more precise. The scratchiness was more of an observation. Sorry for this.
100Hz humming is the problem.

Today I tried 5Y3 rectifier instead of GZ34 - the noise is there but quieter. I do not have a solid state rectifier that drops into an octal socket to test if I can get the noise down even more....

Oh I have back diodes installed on the rectifier tube socket - I have them in my Champ build, too that is dead quiet, so should not matter.

I also tested the amp in another room, where there's no routers etc - the noise is the same....

Still trying to figure it out....

[gitarzysta]

A different thread pointed to a potential solution: ground preamp B voltage filter caps separately from power tubes B voltage caps.
However in my build I use a multi cap can with just one common grounding pin...

But I changed where I ground PT center tap, heaters CT and the multi cap "-". No change, though.

I also attached another 20uF (40uF total) cap to the preamp section of power supply - no luck, either.

[gitarzysta]

Without a valve in V4, no signal can make it through to the power valves. I can’t see how any scratchiness be heard in that condition?
And no hum from preceding stages could pass.

With all valves in place and no signal input, ie idle, does the power valve anode or cathode current change according to the vol control setting, ie being turned    from one side of the scratchiness to the other.

Actually I have to apologise - I was spending so much time trying to figure it out and inhaling soldering fumes I stopped thinking.
It is V3 that breaks the noise - so reverb recovery and mix tube. I checked all wires, grounds, the V3 socket - cannot really figure it out. I tried other 12ax7, 12au7 - the hum just doesn't want to go home and adding reverb adds hum and white noise.

Removing power tubes also stops any humming.

And as far as Vol scratching, the noise is practically gone - I sprayed the pot with a pot cleaner.

Picture of current layout:

[gitarzysta]

If anyone is interested, this is what I did so far with some success:

1. V3 - even though it was a new tube and a new socket it was apparently dirty - cleaning the tube pins and the socket made all the pops and cracks disappear, with or without the tank connected. Hiss and buzz still increases a little as you add more reverb

2. 100Hz buzz - I added another 47uF cap to the B+ node that goes to preamp tubes - MUCH quieter - *** GZ34 does not like to see  more than 40uF as the first filter cap. I hope it is OK if I have 67uF total as the LAST cap *** Could anyone confirm please?

3. I soldered 6V6s' cathodes to the same ground point as the can cap and trafo's center taps (heaters and high voltage)

4. I'll swap the 68K input resistors for metal film ones - I *HOPE* this will improve the hissing

5. The wire going from the vol pot wiper to V1b grid is still a little sensitive/microphonic - I use screened wire, grounded at one end - the sensitivity is there no mater what (no DC measured on the pot)

6. The wire going from 10pF cap (the one bypassing 3M3 resistor) to V3 grid (pin 7) is VERY microphonic with reverb on - this one I also swapped for screened wire - no change. I replaced the 10pF cap to another one, no change.

6. 3nF cap, in reverb recovery stage is also a little microphonic - I changed it - no improvement

7. Bias filter cap - increased to 100uF (plus slightly cooler bias, too)

8. I found one potential ground loop around speaker jacks

9. I also used isolated speaker jacks so I decided to connect their ground directly to NFB tail resistor, not through the chassis - I do believe it helped a little

So it is much better, but still not perfect....Hum is OK, hiss still to be improved.
Maybe the sensitivity is there by design?


Gitarzysta

[DummyLoad]

3nF cap, in reverb recovery stage is also a little microphonic - I changed it - no improvement

ALL caps in the signal path are microphonic.

--pete

[pdf64]

Almost anything can be microphonic, if there’s enough gain after it!
But Class 2 ceramic caps are particularly microphonic.

Regarding Q2, any HT caps that the GZ34 charges via a resistor or choke don’t contribute to the capacitive load limit, when used in a cap input arrangement.

[DummyLoad]

Almost anything can be microphonic, if there’s enough gain after it!
But Class 2 ceramic caps are particularly microphonic.

Regarding Q2, any HT caps that the GZ34 charges via a resistor or choke don’t contribute to the capacitive load limit, when used in a cap input arrangement.

caps are microphonic as a result of their construction. it's called the piezo effect. ceramic caps by material and construction are more prone to the effect. film/foil caps made with "softer" materials have less of a piezo effect.


--pete 

[tubeswell]

A different thread pointed to a potential solution: ground preamp B voltage filter caps separately from power tubes B voltage caps.
However in my build I use a multi cap can with just one common grounding pin...

But I changed where I ground PT center tap, heaters CT and the multi cap "-". No change, though.

I also attached another 20uF (40uF total) cap to the preamp section of power supply - no luck, either.

For this to be more effective, the preamp filter cap needs to be separate from the cap can- I.e. not in parallel with the cap can’s ‘pre-amp’ cap. It’s the combined ground returns of the cap can that cause the amp to be hummy.

[gitarzysta]

Thank you all for the input - once my metal film 68k resistors arrive I'll report back if they help to cure the hiss.

I also biased the amp a little cooler, like most stock AA1164 were - this decreased 100Hz humming, too.

I also have a few, supposedly low microphonic, preamp tubes on order as well as a different set of 6V6s - we'll see.

Nevertheless now my stock Marshall Origin is noisier (100Hz hum) than my AA1164 - 

[HotBluePlates]

... I also biased the amp a little cooler, like most stock AA1164 were - this decreased 100Hz humming, too. ...

"Bias cooler" means there is a more-negative voltage on the 6V6 grid.  That more-negative voltage means the tube needs a larger input signal to hit the same power output it did when it was biased hotter.  Which also means the tube is less-sensitive.

Saying that so you're clear on the fact that biasing cooler probably didn't reduce the 100Hz hum itself, but caused less to appear in your speaker because the 6V6 didn't amplify it as much.

[gitarzysta]

Yes, clear - and tremolo is stronger, too ;)

[bmccowan]

Its great that you got the noise level down, but I cannot help thinking that there is an underlying problem. I would not think that degree of filtering is needed for a PR unless you have really boosted the gain. Did you go through the signal chain with probe and listening amp? And did you check all the caps for dc leakage? There is a recent thread here about a noisy Bassman 50 that had a small silver mica, almost new, cap leaking dc and causing similar noise issues. After all, those caps are cheap and QC is likely lacking.
Just a thought.

[gitarzysta]

Its great that you got the noise level down, but I cannot help thinking that there is an underlying problem. I would not think that degree of filtering is needed for a PR unless you have really boosted the gain. Did you go through the signal chain with probe and listening amp? And did you check all the caps for dc leakage? There is a recent thread here about a noisy Bassman 50 that had a small silver mica, almost new, cap leaking dc and causing similar noise issues. After all, those caps are cheap and QC is likely lacking.
Just a thought.

Some update:
1. 50Hz/100Hz humming all gone - I installed a humdinger pot - was it the trafo wound not symmetrically, some lead dress I could not correct or something else - humming no more
2. I **forgot** to connect a wire from speaker out to 2k7 NFB resistor - go figure! ;) - no NFB did not help to keep the amp quiet - did it? :)
3. Once all tube sockets and pins were cleaned no more static noise
4. Reverb hum - changes a lot with the tank position/orientation - from "very noisy" to "some hum" when I'm already surfing in the wetness

The only noise that still bothered me is hiss - up to 5-6 on the dial it is very well controlled, then hissing increases. Reverb - up to 3ish is very quiet, then adds to the hiss.

Now something I did not consider, being not an experienced amp builder (I built only 2 amps from the scratch) - majority of resistors I used were some variety of "carbon". I already swapped input grid resistors to metal film. Plates are 2W metal oxide. I may go farther and change some others, around V1, power tube grid stoppers (not present in the original design), some other grid leak resistors etc.

I have a bunch of 12a*7 tubes I can try, too.

However I do believe the current state of noise is low - the problem is I have no "reference" PR around to compare.

I did however compared both hum and hiss to a boutique Dumble type amp, also based on 2x6V6 with tube-driven reverb - I consider it VERY quiet - can't tell if it's on on the clean channel unless you crank master and gain.

Except reverb hiss (the "Dumble" reverb is just exceptionally noise-free) my PR clone is in the same ball park now (even quieter maxed out - "Dumble" is just louder, around 20W, so is the noise).

I do believe swapping some other resistors for metal film ones, I will do this when I have time, may improve it just a bit farther.

So maybe there is something in the way I built it that, the power trafo, the tubes etc and I just masked it with the humdinger, extra filter cap, not-too-hot bias.

I have not checked all caps - so this is also something I will for sure consider (I may have a leaking filter can - can't exclude this) - thank you!