Pete Millet's NuHybrid amp - Help diagnosing buzz/hiss in right channel

[ChetJettison]

Depends, is the noise exactly the same in both channels, like mono would be?

Or is there rather, about equal, but different, noise in the two channels? If you used headphones,
would the noise be in the middle of your head, or in each ear? Listen right between the speakers.
Does the noise center at your head, or do you hear two noise sources, at the left and right?

Yep, metal film. Pretty sure they're all KOA Speer brand.

 

[ChetJettison]

This circuit uses very high resistance values, and very low plate current.
Did you use metal film resistors?

There's no easy way to check the regulator except by replacing it,
but it is not very likely to be the problem. And there is already some filtering on it.

Can you install different tubes to see if the noise changes?

These NuTubes are rather costly so I can't exactly grab a replacement too easily. I suppose I could de-solder the NuTube from my other amp that uses one, and swap it out just to try. I have a very nice Hakko desoldering gun that always works well without damaging the PCB or the components, so I'd feel pretty confident trying that.

But let's say I tried that, and the hiss was still present... what would that tell us?

 

[ChetJettison]

Just the other one of the same type and manufacture, in another amp.
They're pretty specialized tubes, not like any other tube out there.
Check it out
Nutube – English | korgnutube.com – English

 

[rayma]

In that event I would say that the design is lacking, and operates the tubes
at excessively low plate currents (roughly 30uA).

 

[ChetJettison]

Any modification I could make to change that?

 

[ChetJettison]

In that event I would say that the design is lacking, and operates the tubes
at excessively low plate currents (roughly 30uA).

Also, for what it's worth, I understand that the NuTube has very low power requirements compared to conventional tubes.
Guide | Nutube – English

Perhaps that's why it seems low in this design? It's a very unique tube.

 

[rayma]

Yes, but we'd have to look into the tube parameters, or else get some third party advice.
The factory data sheet circuit isn't that much different from this circuit, though.

However, there is a way to see if the noise is from the bias regulator IC.
Leave out the capacitors that you removed, for now.
Connect a 10uF 35V capacitor directly from each tube grid to ground.
Make sure the positive terminal of the new capacitor is connected to the tube grid,
and the negative is connected to circuit ground.

There will be no sound when doing this, just the noise (if it is still there).
This filters the bias supply heavily, and will show whether that is the source of the noise.

 

[PRR]

...operates the tubes at excessively low plate currents (roughly 30uA).

The NuTube (originally a display device) is about the worst tube ever used in audio. I forget the curves but 30uA may be as good as it can do.

The nearest experts will be Pete Millett and Nelson Pass.

Pass's plan: https://www.firstwatt.com/pdf/art_diy_nutube_preamp.pdf -- not so different from Millett's, as far as the tube goes. (Different in/out buffers.)

 

[rayma]

First I hoped it was the 475k bias resistors on the buffer bases (should be more like 30k instead).
Then thought it could be the bias regulator IC, which still has not been ruled out, but probably
it is the tubes.

 

[PRR]

First I hoped it was the 475k bias resistors ....

Bypassed by 10k pot through 100nFd cap. About 1uV in the upper audio band. Would be quite inaudible at line level. (Even raw 475K is only 9uV, not a big hiss.)

I have not seen where either Millett or Pass comment on hiss. I wonder if there could be "good and bad" batches?

Ah,,, heater temperature matters, at least in more sophisticated cathodes.

 

[jan.didden]

OK, Test complete. If anything, that made the hiss louder. Also, perhaps obviously, no audio signal was heard.

When you say hiss, you specifically do NOT mean mains hum/buzz, right?
What 'tone' does the buzz/hiss have?

Jan

 

[Signal lost]

How about some images of your build? I have two of these HPA's and they are dead quiet. Your comment about having to ground the volume pot has made me wonder why this was needed. In my builds the volume pot is mounted to the circuit board as is the headphone connector and the rca connectors. Maybe worth a look.

 

[ChetJettison]

Here's an update for you: Pete Millett got back to me about the hiss, and recommended lowering the value of R13 and R14, starting with 220k and going as low as 100k or lower. I tried 220k and didn't hear a huge difference. Going to try 100k later and give it a listen.

When you say hiss, you specifically do NOT mean mains hum/buzz, right?
What 'tone' does the buzz/hiss have?

Jan

Higher frequency hiss. Snakelike. Doesn't sound like mains buzz to me at all, having heard that in other scenarios.

How about some images of your build? I have two of these HPA's and they are dead quiet. Your comment about having to ground the volume pot has made me wonder why this was needed. In my builds the volume pot is mounted to the circuit board as is the headphone connector and the rca connectors. Maybe worth a look.

Here's some pictures of my build. NuHybrid Build - Album on Imgur

 

[Signal lost]

Hey Chet Thanks for the pics. The build looks good. It is had to see the soldering on the volume pot, did you check for solder bridges there?

 

[ChetJettison]

Update for those interested: Lowering R13 and R14 to 100K made a palpable different. I'd say about 80%-90% reduction in hiss. Pete suggests going as low as 47K, but warns of gain falloff. I'll be trying 47K as soon as I get my hands on some.