Adding gain pot introduced hum that goes away at full gain, why?

[Bowie]

  Hi, I just built a V72 clone (modified Drip board) and it worked wonderfully with very low noise (mic was extremely quiet and there was only a slight hum at the instrument jack).  To control volume I added a gain pot before the amplifier stage it started to hum at a moderate, but noticeable level.  The hum goes away at full gain, but only when the mic is selected.  It stays continuously noisy for an instrument.  Also, with the gain down, the hum is identical between the mic and inst channels.

  I have the input xformer's output going to a SPDT switch along with a 1/4" jack so that I can choose between the two.  After this, I added a 100k PEC pot before the amp-in.  Adding the pot is when teh trouble started.  The pot is wired correctly with the input on the 1st tab, output in center, and ground on 3rd & pot chassis.  Any ideas what might be going on here?
Many thanks!

 

[ruffrecords]

Exactly where did you connect the pot ground pin? Did you use shielded cable to and from the switch and instrument socket?

Cheers

Ian

 

[Bowie]

Exactly where did you connect the pot ground pin? Did you use shielded cable to and from the switch and instrument socket?

Cheers

Ian

  Thanks for the response.  The ground pin on the pot has three points of contact;
1) Braided shield covering the wire that leads to the input tube (this shield is connected to the audio ground on the other end).
2) Another is the shield covering the wire coming from the mic/inst switch (shield is not connected on the other end)
3) Connects to the shell of the pot itself.
I checked for good continuity on grounds.  Moving the shielded wires around has no affect on the noise, subtle or otherwise.  FWIW; when I kill the power, the hum completely dies and I have a very clean signal for a few secs until the tubes finally cool.

 

[JohnRoberts]

If you wired in a potentiometer as a simple attenuator, two things change between full up and as you turn it down.

1- as you turn it down the source impedance at the wiper increases as the series resistance from wiper to both ends in parallel gets added. A noise current or random noise pick-up at the point could increase as the impedance increases. Note: the impedance will drop again as the pot reaches full off.

2- A second phenomenon as the pot gets turned down the wiper output is now a voltage divider combining the signal from the top of the pot with any signal/noise at the bottom of the pot.

Which mechanism is bothering you will reveal by how it behaves at full off, where it is low impedance, but connected to a different signal (ground?).  A 100k pot at mid rotation is like adding a 25k resistor in series with the input source.

JR

 

[ruffrecords]

OK, try disconnecting the body of the pot. In other words delete the connection from the ground pin of the pot to its shell (#3 in your last post).

Cheers

Ian

 

[Bowie]

  Thank you Ian for the advice and John for the detailed info.  Unfortunately, no dice so far.  I disconnected the ground to the shell of the pot with no audible difference (even tried un-mounting the pot from the chassis).  With the pot out of the circuit, the mic pre is literally the quietest tube mic pre I've ever used.  Adding the pot creates a slight hum, even at full gain and, as mentioned, increased hum when turned down.  The DI has some hum no matter what position I have it in, though it seems that the hum increases a little with the pot in circuit.  I've tried touching a ground wire to various areas and moving every possible wire around with no audible movement in the hum.

  Any ideas as to how I can either make the pot work or how else I can get volume control before the amplifier?  This 100k pot is mentioned in the Drip electronics manual as an option, though these builds are notorious for errors.  Would a stepped attenuator resolve this?  Any info is greatly appreciated.

 

[ruffrecords]

Can you post a schematic of exactly how the transformer, switch, jack, pot and input tube are wired?

Cheers

Ian

 

[JohnRoberts]

  Any ideas as to how I can either make the pot work or how else I can get volume control before the amplifier?  This 100k pot is mentioned in the Drip electronics manual as an option, though these builds are notorious for errors.  Would a stepped attenuator resolve this?  Any info is greatly appreciated.

It would be useful to determine what the mechanism is that is injecting the hum.

Besides the two properties I posted about adding an attenuator, a third property is that the 100k pot will look like a 100k resistor connected to the input injecting signal directly.

I don't know what a V72 is but that doesn't really matter. The issue appears to be using the proper clean ground (I hate that word) audio reference.

Perhaps I need to ask specifically, "how does the hum change with pot rotation?" , "when is it loudest?".  If the hum does not get lower when you turn down the pot, the ground the bottom of the pot is connected to has hum in it.

JR

 

[emrr]

Maybe I missed it; how long are the wire runs?

 

[Bowie]

  Thanks guys, the help is truly appreciated.

Can you post a schematic of exactly how the transformer, switch, jack, pot and input tube are wired?
Cheers
Ian

Yes, I'll work on this tonight.

Perhaps I need to ask specifically, "how does the hum change with pot rotation?" , "when is it loudest?".  If the hum does not get lower when you turn down the pot, the ground the bottom of the pot is connected to has hum in it.

With the mic in selected, the hum is;
1) nil without the pot.
2) small with the pot installed and turned to full gain.
3) big when the pot is turned down.
So, to give you a direct answer; it's loudest when potted down.  When potted down, you cannot hear any difference in the hum while switching to the instrument (1/4") input.  It stays consistently loud and deep (no buzzing harmonics).  However, while the 1/4" is selected, the hum is consistent whether the pot is turned up or down.

One thing I don't understand, it seems that you are referring to different grounds.  Should I give the power (specifically heaters or AC) a different ground than the audio? As-in, no continuity between them?

Maybe I missed it; how long are the wire runs?

Just enough to get around inside the chassis, maybe 8" ea.  I'll try to post a pic along with a little schematic drawing later.  Physically moving the wires has no affect.  I've even tried moving around a grounded piece of foil to see if I can block any hum between wires, xformers, etc with no change at all.

 

[PRR]

> big when the pot is turned down.

Then the groundy side of the pot goes to a bad ground.

It would normally go to the *amplifier input ground*. (As JR has said: ground follows signal.)

 

[Bowie]

> big when the pot is turned down.

Then the groundy side of the pot goes to a bad ground.

It would normally go to the *amplifier input ground*. (As JR has said: ground follows signal.)

Yes, the pot is wired to the amp-in ground.  I removed this and tried wiring the pot to other grounds with no improvement.  What creates a "bad ground" and what do you recommend i do from here?
Thanks.

 

[CJ]

what kind of input transformer?

sounds like the input is well balanced for hum rejection and the pot is screwing this up,

you might need an H attenuator or something like that,

 

[JohnRoberts]

  Thanks guys, the help is truly appreciated.

Perhaps I need to ask specifically, "how does the hum change with pot rotation?" , "when is it loudest?".  If the hum does not get lower when you turn down the pot, the ground the bottom of the pot is connected to has hum in it.

With the mic in selected, the hum is;
1) nil without the pot.
2) small with the pot installed and turned to full gain.
3) big when the pot is turned down.
So, to give you a direct answer; it's loudest when potted down.  When potted down, you cannot hear any difference in the hum while switching to the instrument (1/4") input.  It stays consistently loud and deep (no buzzing harmonics).  However, while the 1/4" is selected, the hum is consistent whether the pot is turned up or down.

One thing I don't understand, it seems that you are referring to different grounds.  Should I give the power (specifically heaters or AC) a different ground than the audio? As-in, no continuity between them?

Maybe I missed it; how long are the wire runs?

Just enough to get around inside the chassis, maybe 8" ea.  I'll try to post a pic along with a little schematic drawing later.  Physically moving the wires has no affect.  I've even tried moving around a grounded piece of foil to see if I can block any hum between wires, xformers, etc with no change at all.

OK, hum is max when pot is turned down... you can think of that as the pot has panned the input to its ground leg, so you are hearing what signal (hum) that ground terminal has on it.

Ground is a complex subject. Perhaps better to think about audio signals in general. No signal exists by itself, but is always relative  to some reference. For a singled ended circuit, like your attenuator, the reference is supposed to be a perfectly 0V ground. Since 0V only exists in theory (or at absolute zero in the lab), in the real world we must deal with noise on ground. Also ohms law predicts that current flowing in resistances (wires have resistance) will generate voltage drops, adding more voltage differences between sundry grounds.

I'm sure you will get some more specific advice when you share a schematic, while I am not a tube guy.

JR

 

[Bowie]

Thanks TJ and JR.

OK, here's a crude microsoft paint layout of the input stage essentials.  There are things not included like shields on the wires, etc, but hopefully it sets a basic picture.  Let me know what other info is needed.
Thx!

 

[JohnRoberts]

Thanks for the effort.

From observation the Sowter transformer secondary ground should be connected to amp input ground and Pot ground should also connect to amp input ground.

JR.

 

[PRR]

.

 

[Bowie]

.

Thx for the corrections, RPR.  I followed your wiring and it made a significant improvement.

OK, hum is max when pot is turned down... you can think of that as the pot has panned the input to its ground leg, so you are hearing what signal (hum) that ground terminal has on it.

Thx JR, that made a lot of sense and with this in mind I ended up finding several issues.


After staying up all night, experimenting with adding and removing different grounds, I've got the hum extremely low now.  However, two new issues have arisen;
1) There is now a small buzz when the pot is gained down.  The bassy fundamental is almost gone but some buzzy harmonics (which weren't there before) have now shown up.  Any idea what would cause it to trade noises like that?  I've been moving wires all around with no change thus far.

2) When the mic input is selected and gain pot is fully open, there is a small buzz.  The strange part is that when I was moving wires, I noticed that if I touch jacket of the wire going to EF86 pin 6 (plate) with my bare hand and ground my wrist on the chassis, this light buzz goes away completely (though the above issue is unaffected).  I have this tube mounted on the chassis instead of the board (for easier access) so the wire is leading to it is about 5".  Will sliding a braided shield over this wire help?  If so, where would I want to ground it?

Thanks again.  I'm learning a lot from you guys.

 

[My3gger]

I've had similar although not as bad problems with too long wire runs. Cure were shorter wires from trafo's secondary and tube socket's to tag board. Anode connection seems long at 5". Maybe try to shield it, ground to cathode of the same tube and see if noise improves. Also check that heater and anode wires are not too close to input connection. Do i understand you correctly, there are wires from pcb to tube socket?

 

[Bowie]

Also check that heater and anode wires are not too close to input connection. Do i understand you correctly, there are wires from pcb to tube socket?

There's a decent amount of space between the input, amp, and power sections.  Heater wires are run along the top corners, as far away from everything else as possible.
Yes, I ran wire from the PCB's designated tube socket holed to a tube socket I mounted on the back of the chassis so that I won't have to open the unit just to audition different tube (which I do a LOT of).  I also piggy-backed some wires to another socket so that I can use either an EF804 or EF86.

I'll try the shielding you recommended and see about shortening up the anode and secondary wires. Thanks.