General solutions for tube preamp grounding and shielding?

Hello,

I've recently been fighting with my two channel two-bottle NYD preamp and 50hz hum. DC heaters did nothing, and the hum was identical with AC heaters. With layouts I'm a bit of a beginner, but I also couldn't find much META information on grounding and shielding lore. Thought I'd pop in a ask instead, since I've plans to build other designs, too and would like to get it right the first time from now on.

I found these,

http://www.aikenamps.com/StarGround.html
http://www.geofex.com/Article_Folders/stargnd/stargnd.htm

and followed both bus grounding and star grounding on the beard board, but the hum pretty much stays. I'm inclided to think my troubles come from outside the preamp, and are there because of inadequate shielding and quite possible layout, too.

I was suggested to look for tried and tested vintage designs for hum removal.

Could I maybe get some pointers on where to start? I would like to learn great methods now so I don't have to fight with this for the rest of my life.

1. point-to-point layout done *right*, with correct transformer placement, wiring and cap layout solutions?

2. bus or star grounding? Both seem great if done right

3. shielding? lead/tin plates and good solutions on that.

Your thoughts are much appreciated. Unfortunately I can't really contribute much, but I'm eager to learn. :grin:

Not that there couldn't be other issues, I had a simular problem a while back with tube mic pre that was kicking my ass. No matter what I tried it wouldn't go away. Turned out to be bad tubes. Out of the 8 NIB tubes I had, 4 were bad because of hum problems, even though they would pass in a tube tester.

I also had tried putting a positive bias on the heaters which helped a little, but it was mainly bad tubes in my case.

HTH

I've rules out tubes as the cause with this kind of deduction:

Both channels produce nearly identical hum wether I change tubes, or use swap in replacements.

I would think it extremely bad luck if all the tubes were hummers in an identical way.

Hence I'm inclined to think it layout-grounding-shielding issue.

[quote author="Kingston"]I've rules out tubes as the cause with this kind of deduction:

Both channels produce nearly identical hum wether I change tubes, or use swap in replacements.

I would think it extremely bad luck if all the tubes were hummers in an identical way.

Hence I'm inclined to think it layout-grounding-shielding issue.[/quote]

I forgot to mention that this was a 2 channel unit, and 2 original input stage tubes were bad, and so were the next four out of the box! :grin:

Hope you can figure it out! But it is good to make sure you're properly grounded and sheilded. Also be careful how close the power supply transformer is to the preamp.

Build a remote pwr supply and see if that solves your problem.
If it does, then that means you need to better shield the pwr supply.

[quote author="CJ"]Build a remote pwr supply and see if that solves your problem.
If it does, then that means you need to better shield the pwr supply.[/quote]

Done that. What it did accomplish was that the channel nearest to the PSU, that had more hum is now identical to the furthest one. The real problem still lies elsewhere.

Star grounding has its place, but sometimes it's taken as an overkill measure when simple, good layout practice would have sufficed.

Here's something I uploaded a while back. You might find it helpful.

4.4Meg PDF

Sometimes there is a ground loop in/around the work bench and nothing seems to work right because of that. Do you use a lot of extention cords around the room?

Thanks NYD,

The concept of shield that the article talks about is rather new to me. I suspect the only thing defined as shield in the preamp are the transformer cases.

[quote author="tk@halmi"]Sometimes there is a ground loop in/around the work bench and nothing seems to work right because of that. Do you use a lot of extention cords around the room?[/quote]

Just one for the whole system. I've already tried the amp in other rooms in different configurations and the hum is practically the same everywhere. The rest of the studio is dead silent, humwise.


How much can the layout on the preamp influence ground problems, even if star grounded? Crossed wires, tightly fitted caps and resistors near I/O transformers or each other etc. This is area where I have least practical experience.

Short the grid out to ground on each amp stage to isolate the stage that is producing the hum.
Start at the back and work your way up to the input jack.

[quote author="CJ"]Short the grid out to ground on each amp stage to isolate the stage that is producing the hum.
Start at the back and work your way up to the input jack.[/quote]

I will probably do that just to see what happens.


By the way, I managed to get to hum to workable and reasonable level already. I had accidentally soldered output trafo screen and common. Pure carelessness (and if that's not a beginners error then I don't know what is), but hum went down 20dB at max gain, already adequate for work. Problem nearly solved.

But here's some pure oddness (to me):

1. shorted XLR input for testing.
2. min gain, pad on or off, hum is the same, as should be.
3. max gain and if I turn off the pad, hum is workable, but pad *off* and hum raises 20dB!

I'm guessing the pad resistance comes to play with the shorted inputs and input trafo. Naturally this isn't a problem with audio on the inputs and nothing shorted.

3. max gain and if I turn off the pad, hum is workable, but pad *off* and hum raises 20dB!

Click to expand...

If your pad is before the input transformer - as it should be - this points to the fact that there is hum already before the input to the pad. And as you have input shorted, the only place this hum can come from is from the input ground. Try taking input ground from a different place.

In general, there's two major "internal" sources of hum:

- Magnetic/electrostatic induced hum radiating from powersupply. You eliminate this by moving the supply. Distance is your friend - radiation drops with square of distance.

- Poor regulation of power lines, giving residual hum on top of supply voltages. This ripple voltage leads to ripple currents in your consuming-circuit PCB traces - of which the ground is a part! - which again results in ripple voltages across the length of the traces because of their (although minimal) resistance. Star grounding is your friend, as every circuit then has it's own - undisturbed - 0V reference. Note that if you had perfectly clean DC voltages, there wouldn't be any use for star grounding (unless you wanted to compensate for DC voltage offset, which we normally don't bother with in audio)

Jakob E.

[edit]

Oh dear. False alarm with success.

Grid shorting time. :?

Couple of questions on hum pickup. On point to point design, how much can these factors influence hum:

1 wires
a) b+, should the wire be separated from what others especially, twisted with common etc.
b) heaters, wether AC or DC, see a)
c) wires going inbetween XLR I/O and front panel input pad and attenuator. Twisted? avoiding what trouble spots in preamp especially?
d) wires to tubes.. do they influence each other?
e) trasformer hook up.
f) wires to ground starpoint or groundbus. Does it pick up hum by itself?

2 capacitors
do their orientation and angles influence hum to or from each other and wires

3 mu-shielded transformers
what wires, if at all shouldn't especially cross anywhere near them?

4 actual parts. How sensitive are these kind of tube circuits to the normal part value deviations, say with caps and such. if trafo zobel network or cathode bypass caps are slightly off, etc.?


Could be something I left out, but this layout design is the grey area to me, and I couldn't find an awful lot of info on it. I already know somewhat the right *theoretical* order of things within the preamp, ground etc, but the layout itself seems troubling.

here are the current measurements:

gain at "off"



gain at max



That's not quite optimal yet, is it? Or quite normal? Is it possible to completely get rid of the peak, and have just the extremely low general noise floor? The PSU is about 1 meter away from the preamp. If PSU is in the preamp the peak is exactly 25dB louder. Would it be possible to shield the PSU inside the unit so well that I wouldn't need to have it outside?

I managed to get both channels at max gain down to -65dB by some more cap orientation adjustments and rewiring.

It's all a terrible mess inside the thing now (star grounding + bad layout), but I consider this acceptable, since I won't see that noise figure unless doing ribbon recording. In general task gains the noise seems to sit at around -90-80dB, which is really good (to me)!

In fact this preamp will see a whole lot of mileage as a feebthrough processor for a lot of my DAW material, and noise is at -95dB on those gains (and pad on).

Ahhhhhh! It's finally working and done and I feel like a winner. Thank you everybody for helping me out on misc threads around the place. Oh the things I've learned.

I finally get to post pictures and sound examples early next week when I've finished the external PSU enclosure.

:thumb: