low level hum

[ccmccull]

Hi all, I need some troubleshooting help. Thanks!

I have a low level 60hz hum which is not volume dependent. I hear it through my Grado sr325x. It occurs with or without the tubes plugged in.

I have it wired for 4 ohms. All resistance and voltage checks were within tolerance.

This seems like a ground loop to me so I checked continuity on all the grounds and paid special attention to the power transformer. All are < 0.3 ohms. I also re-seated the PT and checked for no continuity with the 8 lug strip in front of it.

I have a hypothesis. If I'm reading it properly, the wiring diagram shows that one side of each of the OTs should be connected to ground: the terminal 7 with solid {green,brown} connected. As far as I can tell, in my kit as built, this ground is established via the 0 ohm resistors attached to the black speaker binding posts. So the OT terminal 7 is only grounded when the headphone jack is not plugged in. I confirmed that continuity is broken on all three headphone jack pins when plugged in.

To confirm my hypothesis, I plugged in the amp (with tubes out) and measured VAC at OT terminal 7 vs H1, in the two conditions of whether the headphone jack was plugged in or not.

HP jack plugged in:
OT2 B side term 7 vs H1 0.38VAC at 60hz
OT2 A side term 7 vs H1 0.45VAC at 60hz

HP jack not plugged in:  
OT2 B side term 7 vs H1 0 VAC
OT2 A side term 7 vs H1 0 VAC

Then, I measured OT Term 7 vs. Term 6 to confirm these should be zero. They are zero on both sides, with HP plugged in or not.

One other piece of info: on the IEC jack, I get 0.17 VAC between N and E.

So my question is, did I miss a grounding step? Or is this an intentional part of the circuit design? I'm hesitant to test grounding the HP jack directly to the chassis without someone expert telling me it's a good idea... Or maybe this is a red herring.

Thanks for the insight!

Pictures of my build:

https://photos.app.goo.gl/xJdiB8hKFMykB4iv6

[Paul Birkeland]

One problem we see sometimes is that terminal 36 can touch the nut behind it and make noise.

Definitely give a try running a piece of wire between the bottom two lugs of the headphone jack to see if that helps.

The measurement you would want to do would be across the output terminals themselves, not necessarily the chassis plate. 

Your little bit of AC voltage between neutral and earth isn't particularly uncommon.

[ccmccull]

Thanks Paul.

On the terminal 36 vs. PT nut , I visually inspected it again and it's definitely not touching. But it only measures 0.5 ohms between that PT nut and terminal 36. In fact terminal 36 to H1 is < 0.5 ohms. Is that expected?

I also attached a wire between the bottom two lugs of the headphone jack and it does help, but not completely. It seems to remove the buzzy quality of the hum, but there remains a very faint "pure" 60 hz tone. It's only noticeable with these very sensitive headphones while in a quiet room. As expected, OT terminal 7 vs. H1 goes to 0 VAC.

I also measured VAC between the 3 headhone pins. All 3 combinations are 0 VAC, with or without the extra wire. So I suspect the stray current that I'm still hearing is actually from one (or 2 or 3?) pin through my body to the floor. When I measure VAC at any of the three HP pins while holding the other probe in my hand (not touching anything else except my feet on the floor) I get 2.5VAC.

Anything else I should try?

[Paul Birkeland]

The low resistance between the nut and terminal 36 is normal, you just don't want them physically touching because they will make a competing connection to the one that already exists and that can make a ground loop.

Is the noise that's remaining something that can be influenced by moving the volume pot around?

If you can give up the speakers switching off when headphones are plugged in, there's a mod I can show you that can knock any residual noise down even lower.

[ccmccull]

The volume pot doesn't affect it.

Sure I don't need that cutoff functionality. Excited to try a mod!

[Paul Birkeland]

You'll want a pair of 8 ohm resistors that are 2-3W rated parts and a pair of 2 ohm resistors that 1/8W or so.  Move the Cat5 wires that connect to the outside lugs of the headphone jack to the inside lugs with the other cat 5 wires.  Attach one 8 ohm resistor between the upper pair of jack lugs, another 8 ohm resistor between the middle pair of jack lugs, and a piece of wire across the bottom pair of jack lugs.

You can also try to the recipe above adding the 2 ohm resistors between the upper outer jack lug and the lower outer jack lug, as well as the middle outer jack lug and the lower outer jack lug.  This will likely be preferable to just the 8 ohm resistors on their own.

[ccmccull]

Thanks for all the help Paul! I'll buy some resistors and give it a try!

[ccmccull]

Hi Paul, would you mind checking these are good for this purpose? It's my first time buying individual resistors and there are so many choices...

2 ohm 1/8 watt:
https://www.mouser.com/ProductDetail/594-5063JD2R000F

8 ohm 3 watt:
https://www.mouser.com/ProductDetail/71-RS02B8R000FE12

Thanks!

[Paul Birkeland]

I would probably go with metal film over a wirewound for the 8 ohm resistor.  These would be OK.
https://www.mouser.com/ProductDetail/Vishay-BC-Components/PR03000208208JAC00?qs=sGAEpiMZZMvxX9Q80wbhCbIkhci%252BTqQ7i5E9Pg5jhiM%3D

[ccmccull]

great thanks I'll do that.

(I'm curious to learn about the advantages and disadvantages of wirewound vs. metal film vs. etc. Is there somewhere I can learn more about what works best where?)

[Paul Birkeland]

This was posted by PJ on our old forum and is still floating around:

"Carbon composition is the best for low inductance, which is important in grid stoppers which are effective at hundreds of MHz. They are noisy, but only if they have DC current through them, and the noise is proportional to the resistance. So as grid stoppers (no current) and as plate stoppers (small resistance, and located after the first signal amplification stage so that noise is 30dB less important) they are OK. In both applications, drift of the resistance is not important.

Plate resistors are usually metal film, which is generally very quiet even with DC current, and they are also readily available in 1% or better precision. They do have a bit of inductance, but it's not significant at audio frequencies. We also use them in the RIAA equalization network, where precision and low noise are important. We avoid them when small amounts of inductance are important.

Wirewound is the quietest, but usually has the most inductance. We us them in power supplies where the small inductance helps isolate diode switching transients and power line RFI noise.

Metal oxide is noisy and usually imprecise, but they are very sturdy (withstand voltage transients and momentary overloads without damage) so they make good power supply bleeders where the large parallel capacitance can shunt the noise to ground."

I would add that carbon composition resistors tend to be a bit delicate, so while you would find them as stoppers in our kits 10+ years ago, we moved to carbon film resistors which are less likely to break.

[ccmccull]

Nice thanks!

[ccmccull]

I soldered in the 8 ohm and 2 ohm resistors tonight and fired it up with the grados and shorted rca inputs. Silence. So silent I thought I might have made a mistake... But no! With music on, wow. Wow. Wow.

Thanks so much Paul!

[Paul Birkeland]

I'm glad that's working well for you!  The presence of the 2 ohm resistors will make this amp not work all that well on speakers, so if you decide that you'd like to use speakers with your SEX amp in the future, we can figure out some different tweaks.

-PB

[ccmccull]

Hi Paul,

I was thinking about trying out this amp with speakers. What did you have in mind? I could put in a simple rotary switch to select between the headphone jack (via the 8/2 ohm resistor setup) and speaker terminals (bypassing those resistors)