Humdinger pot question

[joesatch]

I have elevated heaters with the CT connected to a 65vdc voltage divider. I want to add a Humdinger pot as well. In this diagram the black wire goes to ground.  must i remove my heater CT and do something with it?

[acheld]

As the diagram states, the black wire goes to ground.

What actual problem are you trying to solve? 

I'd recommend reading Merlin's discussion of heater power supply -- http://www.valvewizard.co.uk/heater.html

Personally, my experience with hum related to heaters in guitar amps is that it is not much of a problem as long as you provide a reference to ground.

[sluckey]

You have two options... Connect the wiper to ground or connect the wiper to your 65v divider. BUT DON'T DO BOTH!!!

If you have a real wirewound humdinger from the '50s you may not have a choice. Many had the wiper connected to the case and when you mounted the pot in the chassis the wiper was automatically connected to ground. Like this...

[joesatch]

this is a very high gain 2204 build and i want to eliminate every ounce of noise i can. I may go with DC heaters but i want to try this first.  So my heater CT will remain connected to the 65vdc source with this humdinger in place?

[sluckey]

So my heater CT will remain connected to the 65vdc source with this humdinger in place?

If your PT has a center tap for the heaters, disconnect it, tape the end of the wire, and tuck it out of the way.

If you built an artificial CT using two 100Ω resistors, remove them.

The humdinger pot becomes your new artificial center tap and the wiper will connect to... Ugh, see my earlier reply.

Look at this schematic for reference...

     http://sluckeyamps.com/smoky/smoky.pdf

[joesatch]

So my heater CT will remain connected to the 65vdc source with this humdinger in place?

If your PT has a center tap for the heaters, disconnect it, tape the end of the wire, and tuck it out of the way.

If you built an artificial CT using two 100Ω resistors, remove them.

The humdinger pot becomes your new artificial center tap and the wiper will connect to... Ugh, see my earlier reply.

Look at this schematic for reference...

     http://sluckeyamps.com/smoky/smoky.pdf

well i'm gonna connect the pot wiper to the elevated 65vdc instead of ground. I will remove the heater CT

[Stuff4bikes]

How can you tell the wattage of a pot....?
I have a bunch of 100 ohm and 5k pots that are physically larger than
The CTS or alpha that I get from AES...

[sluckey]

Consult the manufacturer's data sheet if the wattage is not stamped on the pot.

[joesatch]

i bought this pot (as recommended by Robinette). I am concerned it is only 100 ohm though as i'm replacing (2) 100 ohm resisters with one (this pot)

https://www.amplifiedparts.com/products/potentiometer-100-linear-knurled-5w-wirewound-24mm

[Willabe]

i bought this pot (as recommended by Robinette). I am concerned it is only 100 ohm though as i'm replacing (2) 100 ohm resisters with one (this pot)

It will work, but it will loose a just little more heater current to ground.

This here is very important for the voltage divider in a dcv heater stand off, from Merlin;

The divider should have a fairly high resistance so as not to waste current, although the lower arm (R2) should not be excessively large or Rhk(max) may be grossly exceeded, so it is advisable not to make it greater than 100k.

So make R2 100K for least waste current and adjust dcv with R1. (See Merlin's drawing for R2/R1, in link posted below.)

Where did you take the heater dcv from? I always try and take it from the last B+ filter cap node, it's the cleanest in the amp. And there's only 2 heater wires on the 1st preamp tube, so easy to wire in the dc stand of voltage. I put my hum-dinger pot there too.   

Read this on heaters, has hum-dinger and dc stand off info;

http://www.valvewizard.co.uk/heater.html

[sluckey]

i bought this pot (as recommended by Robinette). I am concerned it is only 100 ohm though as i'm replacing (2) 100 ohm resisters with one (this pot)

Don't be concerned. 100Ω is the most commonly used humdinger pot.

[sluckey]

i bought this pot (as recommended by Robinette). I am concerned it is only 100 ohm though as i'm replacing (2) 100 ohm resisters with one (this pot)

It will work, but it will loose a little more heater current to ground.

Heater current flows from one side of the filament winding to the other side. It does not flow to ground. Ground is for reference only.

[Willabe]

i bought this pot (as recommended by Robinette). I am concerned it is only 100 ohm though as i'm replacing (2) 100 ohm resisters with one (this pot)

It will work, but it will loose a little more heater current to ground.

Heater current flows from one side of the filament winding to the other side. It does not flow to ground. Ground is for reference only.

From Merlin;

If the transformer doesn't have a centre tap then you can create an artificial one using a pair of resistors. The resistors should have a fairly low resistance to help shunt leakage currents from the transformer primary. Values of 100R (1/2W) or 220R (1/4W) are typical. They will of course consume a little extra current from the transformer (32mA when using 100R resistors at 6.3V) so bare this in mind.

 

Any time a voltage has a path to ground through a resistor it will bleed voltage to ground, which is bleeding current?

Using a pot, wiper (CT) has resistance on both sides, wiper (CT) goes to ground. Using 2 resistors, resistance on both sides, R's connect together forming CT (wiper) going to ground. Pretty much the same thing. ????? 

[Willabe]

i bought this pot (as recommended by Robinette). I am concerned it is only 100 ohm though as i'm replacing (2) 100 ohm resisters with one (this pot)

It will work, but it will loose a little more heater current to ground.

Heater current flows from one side of the filament winding to the other side. It does not flow to ground. Ground is for reference only.

I guess I'm thinking that would only be with a true CT winding, not with a faux CT?

[sluckey]

I guess I'm thinking that would only be with a true CT winding, not with a faux CT?

Heater current DOES NOT FLOW TO GROUND. Unless you have one of the old style amps that connects one side of the filament winding to chassis and only runs one wire to the tubes with the other tube heater pin also connected to chassis. Only then will heater current flow to chassis because the chassis acts as one conductor for the filament circuit.

But in any filament circuit that has a center tap (real or artificial) connected to chassis ground there will be no current flowing to, or through, or from chassis ground. That connection is for reference only for hum suppression.

With an artificial center tap built with two resistors there will be current flowing through the RESISTORS, but still no current flowing to, or through, or from chassis ground.

Draw it out. Think about it. Show me how you think filament current flows to chassis.

[66Strat]

I look at it in terms of AC voltage. There are +/- 3.15 volts AC from winding terminus to center tap. When one winding terminus is + 3.15, the other is -3.15. Together, they net to 0 volts at the center tap.

[Willabe]

With an artificial center tap built with two resistors there will be current flowing through the RESISTORS, but still no current flowing to, or through, or from chassis ground.

How can that be, when 1 end of the faux R's is connected to the chassis. 

I'm sorry, that makes no sense to me. I don't understand.   

Faux CT has R's connected on both sides, 1 R on each leg connected to the heater, the other end connected to ground.

You say that current flows through the resistor, so where does the current come from and go to that's flowing through the resistor?   

[pdf64]

… where does the current come from and go to that's flowing through the resistor?   

From the transformer winding.
We could reference (ie connect) one end of the winding or the other to chassis 0V common, or connect a winding CT to that chassis, or use equal value balancing resistors if no CT available.
Or reference those points to some other point, eg the heater circuit elevation under discussion.
It makes no difference to that current flow.
Unless we use the chassis as a conductor heater current, in which case balanced or elevated is not an option.

[Willabe]

We could reference (ie connect) one end of the winding or the other to chassis 0V common, or connect a winding CT to that chassis, or use equal value balancing resistors if no CT available.
Or reference those points to some other point, eg the heater circuit elevation under discussion.

I understand that.

From Merlin;

If the transformer doesn't have a centre tap then you can create an artificial one using a pair of resistors. The resistors should have a fairly low resistance to help shunt leakage currents from the transformer primary. Values of 100R (1/2W) or 220R (1/4W) are typical.

They will of course consume a little extra current from the transformer (32mA when using 100R resistors at 6.3V) so bare this in mind.

Then this is incorrect? 

[sluckey]

You say that current flows through the resistor, so where does the current come from and go to that's flowing through the resistor?

Play stop time and we stopped the clock when the top of the winding is positive and the bottom of the winding is negative. Current flows out of the bottom of the winding, through the two series connected resistors and back to the top of the winding. A simple series circuit. The fact that you connected the junction of the resistors to the earth, or the moon, or anything at all, has no bearing on the flow of electrons from the bottom of the winding back to the top of the winding. Think of it as a simple battery connected across two series resistors and the junction of the resistors is connected to Saturn.

Now fast forward 8.3mS and stop the clock again. The top of the winding will now be negative and the bottom of the winding will now be positive. Current now flows in the opposite direction, ie, from the top of the winding, through the two series connected resistors and back to the bottom of the winding. Same simple series circuit, and the fact that you connected the junction of the resistors to the earth has no bearing on where the current flows.

It's kinda like (but not exactly like) a bird sitting on a power line between two power poles. We know that current is flowing from pole A to pole B (and onward), but no current flows through the bird.

Nothing Merlin said is in dispute with what I said. If you ain't got this concept by now, maybe one of the smart guys will splain it better. And that's all I got to say about that.

[66Strat]

The current alternates through the heater elements.

[trobbins]

Joesatch, its worth appreciating that the 'elevation' aspect of your mod may have a benefit when the valve's heater-to-cathode resistance has degraded to a low level (typically through age, but also other factors).  That means valves with relatively low heater-to-cathode resistance may cause noticeable hum, but how much influence the elevation has depends on the valve's heater-cathode interface resistance and how it varies with elevation (varies between tubes).  Without that mod then tube-swapping can be a way to minimise hum through leakage from heater to cathode via the interface resistance (which is typically many megohm), as some valves may exhibit very high resistance levels (without elevation).

In contrast, the humdinger pot allows hum to be minimised caused by capacitance leakage from the heater wiring over to the input grid of the valve.  That hum leakage depends mainly on your wiring layout (ie. nothing to do with the heater-to-cathode interface).  That hum depends on the AC heater voltage waveform (as AC voltage across a capacitor causes AC current to flow) and how well the capacitance current from one heater wire is neutralised by the capacitance current from the other heater wire (ie. the pot wiper setting).  Stray capacitance from heater wiring to input grid wiring is typically very low (pF), and so this mod may only have an influence if you have a 'poor' wiring layout.

[pdf64]

…

They will of course consume a little extra current from the transformer (32mA when using 100R resistors at 6.3V) so bare this in mind.[/i][/color]

Then this is incorrect? 

The 100ohm resistors will draw 32mA regardless of what point the junction between them is connected.
No current will flow between that junction and 0V common or the heater elevation point.

[Tone Junkie]

If it still worries you put a resister between the center tap on the pot and your negative b+ Connection spot.

[Willabe]

Thanks for writing it out for me.

All I needed to know was the series part. I get it now. Thank you.

Current flows out of the bottom of the winding, through the two series connected resistors and back to the top of the winding. A simple series circuit.

Nothing Merlin said is in dispute with what I said.

So the ~34mA loss is through the 2 series 100R's, not the 100R's ground connection. 

Thanks guys. 

[sluckey]

So the ~34mA loss is through the 2 series 100R's, not the 100R's ground connection. 

Exactly. And there is zero current flowing into/out of the ground connection.

[acheld]

Interesting discussion.  So then, how does a Humdinger work? 

As I see it, the Humdinger pot varies the individual resistance of each of the heater legs to the ground reference.   The two sides of the pot are in fact connected in series via the wiper which is grounded.

But, Humdingers do work by varying the value of the "leg resistors" (so to speak).  By turning the pot, you can reduce hum.

So, if there is no current to the ground reference, how does the Humdinger pot affect hum?  The two resistors are in series, and always add up to the same total value.

[sluckey]

Supposedly minimum hum should occur when the two resistors are exactly the same value. Using a pot to simulate the two resistors allows you to precisely make the two resistors exactly equal.

And if anybody really wants to put this "current flow to ground" discussion to rest, you can simply disconnect the junction of the two resistors from ground, but leave the resistors connected to each other. Now simply connect an AC current meter between ground and the junction of the two resistors and measure the current.

[pdf64]

… how does a Humdinger work? 
….

By varying the AC balance of the heater circuit.
If the heaters in preamp valves were perfectly balanced, then the humdinger would offer no advantage over 2 equal value balancing resistors.
As the valve heaters may not be perfectly balanced, a humdinger can be handy.

[PRR]

Not only the heaters. The signal wiring is never exactly symmetrical relative to the heaters. This was particularly obvious on the PCB-based Ampeg VT-40. Heater lines made on PCB can never be perfectly balanced, and the signal line layout is also constrained in 2D. In a simple case, you can match the heater off-balance to the signal off-balance and null the hum.

In practice this does not work 100%. Ever! The coupling is capacitive, "AC" carries HF buzz, so low-hum and low-buzz happen at different settings. The coupling off-balance is different in every stage. So different at low VOL setting than at high VOL. Close-up in my lab I despaired of ever getting a good null. Out on a stage it was much less critical as I got away from the open-back cabinet.

[Willabe]

Out on a stage it was much less critical as I got away from the open-back cabinet.

Open back apposed to closed back? How does that effect heater humdinger buzz?