60hz hum in "Kingston Drive" tube preamp

[plumsolly]

Without unsoldering anything, I connected a humdinger pot across the 150Ω resistors with the wiper to my elevated node. I got the best results (and they were an improvement) when the pot was at one extreme, i.e. one the the 150Rs was shorted. The other extreme of the pot (the other 150 r was shorted) was worse than before. What do you make of that? If I go this route it might make sense to just short that 150R and be done, but does this unbalance indicate something else wrong? And should I expect a change if the value of those 150R resistors is changed?

Thanks,

Ben

 

[emrr]

I've seen it many times. In other words you maybe don't even need the ct r's, it could be best with one side to ground. That would be a symptom of some other relationship in the fil wiring.  Look at the half wave rectification in the front stage of the SA-39B; super easy to hack in for a listen before going further.  It might make a world of difference. 

 

[radardoug]

If your filament wiring is very symetrical then this could indicate a tube problem, excessive heater cathode leakage to one end of the filament. By grounding that end you are stopping the leakage.
You know that d.c. works, why are you not going that way? It will give you better results than playing with the a.c.

 

[PRR]

Heater wiring RIGHT under an XLR stands-out to my eye.

Like the heater leads, the XLR with balanced connection might benefit from keeping-twisted to as-close-as-possible to the pins.

Even so, I'd move that XLR (or tube socket) 2 to 4 times further away.

Pinching the two live leads to the XLR *tight* together might help a little (and also confirm/deny if this is the tree to be barking up).

 

[plumsolly]

I've seen it many times. In other words you maybe don't even need the ct r's, it could be best with one side to ground. That would be a symptom of some other relationship in the fil wiring.  Look at the half wave rectification in the front stage of the SA-39B; super easy to hack in for a listen before going further.  It might make a world of difference.

I went ahead and installed a 500Ω humdinger pot - I figured it would be useful to have it adjustable in case something else changes, and it did indeed make all the difference. In preliminary tests, the hum is down into the rest of the the noise floor. One strange/irritating thing is that the best position for the pot - which is not necessarily all the way to one side as I though before - depends on the position of the first volume pot.

You know that d.c. works, why are you not going that way? It will give you better results than playing with the a.c.

Mostly because of the modification I would have to do to the supply. To get clean DC for 6.3VAC I would have to loose a few volts. I wanted to see how well I could do with AC and it seems that with the humdinger pot it is good enough.

Heater wiring RIGHT under an XLR stands-out to my eye.

Like the heater leads, the XLR with balanced connection might benefit from keeping-twisted to as-close-as-possible to the pins.

Even so, I'd move that XLR (or tube socket) 2 to 4 times further away.

Pinching the two live leads to the XLR *tight* together might help a little (and also confirm/deny if this is the tree to be barking up).

This may indeed be a problem, but it is not the problem I was fighting. That is a line input XLR and the hum appears at the output even when I have it switched to mic input.

Thanks,

Ben

 

[PRR]

> humdinger pot - the best position for the pot - which is not necessarily all the way to one side as I though before - depends on the position of the first volume pot.

Then you have TWO hum sources. One before VOL, one after.

 

[Landins]

How is pin 1 connected at mic input? I see no "direct to chassis" on the pictured input xlr. Can be a source of pre-gain-hum.

Otherwise, layout is king. EMF diminishes by the cube of the distance. Hover with scope probe over areas that may be prone to interfere. Shot out the sec of the input trafo to see what gives.

I built a poormans V76 with ac heaters and psu in box. At high gain settings it hums slightly. But hey! Its high impedence!

 

[Kingston]

It doesn't cost an arm and a leg to see how that AC source of yours performs with passive BridgeRect-BigCap-FatResistor-EvenBiggerCap. Well, those potentially required 22000uF/16V caps will cost a few dollars but it's for science!

You may or may not recall I ended up with passive-dropped DC heaters, but that was with a 9VAC source. Plentiful headroom for smoothing (and house heating). But then, 9VAC toroids are cheap and on the shelf everywhere.

 

[plumsolly]

Then you have TWO hum sources. One before VOL, one after.

Ah, I see. Well, I believe the the pre volume pot hum is much less than than the post.

How is pin 1 connected at mic input? I see no "direct to chassis" on the pictured input xlr. Can be a source of pre-gain-hum.

Pin 1 on all the xlr's are tied to the chassis right at the connector.

It doesn't cost an arm and a leg to see how that AC source of yours performs with passive BridgeRect-BigCap-FatResistor-EvenBiggerCap. Well, those potentially required 22000uF/16V caps will cost a few dollars but it's for science!

At this point, I am pretty content. I did a vocal session today and it was a non-issue. Maybe when I start distant-micing classical guitar with a ribbon it will be an issue. I may explore DC at some point, though. How much voltage do think I would loose off my 6.3V with the scheme you mentioned above? Right now I want to explore some small frequency response issues (high end rise - probably output needs to be damped, and low end drop off - probably input transformer (custom A-series-type UTC) for Gates). Thanks for your efforts on the design, I am excited to start enjoying it!

Best,

Ben

 

[emrr]

UTC A-11; that's what they do. 

 

[plumsolly]

UTC A-11

That makes sense. Any idea why they made custom ones for Gates with the different pin arrangement (not the circular pattern like the stock ones)? Different shielding or something?

Ben

 

[Kingston]

How much voltage do think I would loose off my 6.3V with the scheme you mentioned above?

1.4 X 6.3VAC (= 8.82VDC) is roughly what you get right after the bridge rectifier at the first cap. That means you have 2-3 volts available for smoothing. With big caps it's probably enough. Measure the exact current the tubes eat and calculate the right size and wattage for your resistor.

It's perfectly safe to run those tubes at even 5.8VDC, but see datasheets for how much exactly they will accept. 10% allowed heater voltage deviation is common.

 

[ruffrecords]

How much voltage do think I would loose off my 6.3V with the scheme you mentioned above?

1.4 X 6.3VAC (= 8.82VDC) is roughly what you get right after the bridge rectifier at the first cap. That means you have 2-3 volts available for smoothing. With big caps it's probably enough. Measure the exact current the tubes eat and calculate the right size and wattage for your resistor.

It's perfectly safe to run those tubes at even 5.8VDC, but see datasheets for how much exactly they will accept. 10% allowed heater voltage deviation is common.

In all probability you do not need a huge amount of smoothing. A 6.3VAC supply is 17,8V peak to peak. Assuming your heaters draw 2 amps, then a 22,000uF  reservoir capacitor will have a ripple of (dV=I*dt/C) about 0.9V peak to peak. This is a reduction of nearly 26dB before you even add the second RC filter.

Cheers

Ian

 

[emrr]

That is the standard UTC pin arrangement in that era. Circular came later.

 

[Speedskater]

What would happen if you totally disconnected that nasty DI Input jack?

 

[plumsolly]

I'll evaluate it for a while and if the current level of hum proves to be an issue, I'll try DC - or maybe just switch it over for a rainy day project. It seems like I could get it pretty smooth and still have ~6V.

That is the standard UTC pin arrangement in that era. Circular came later.

I was wondering about that. When did they switch? The 1949 catalog shows the circular arrangement.

What would happen if you totally disconnected that nasty DI Input jack?

Well, with the addition of the humdinger pot, the hum is pretty well under control, and most of hum before was being injected at the grid of V2B.

Ben

 

[emrr]

Best guess on UTC is somewhere '46-'49.