Ampeg B15N build .. hum

[bieworm]

What could be the reason of the hum pot acting like a hum volume pot with minimal hum on 1 side and maximum hum on the other side?
The hum is also barely audible with the volumes on zero

It looks like 1 heater wire gas all the hum and the other has none?

[ChopSauce]

I've been aware of a similar issue (min hum when humdinger on one side) but don't remember having read about any rationale for that - by the way
(sorry for hi-jacking your thread but this matter is of interest to me and maybe you too ... )

Something like that - but the "Power Cord Green Wire" bolted appart...



... maybe?

EDIT (as I look at it more carefully:) isolated speaker/output jacks, really...

Hmm. That particular grounding scheme will have issues with rectifier-induced hum to some degree. The power transformer HV CT needs to go directly to the negative side of the first filter caps to keep charging-pulse induced ground voltage shifts off the signal ground.

Do you mean without first connecting to the ground connection, which would be:

PT/CT -> minus(res.cap) -> starGround

instead of:

PT/CT -> starGround <- minus(res.cap)

[bieworm]

Would there be problems if I let it like that, meaning one side ofvthe filament wirining at 1.5V and the other at 4.8V? It appears to be quiet then...

[R.G.]

Do you mean without first connecting to the ground connection, which would be:

PT/CT -> minus(res.cap) -> starGround

instead of:

PT/CT -> starGround <- minus(res.cap)

Yes, explicitly. The PT CT should go to the negative of the first filter cap FIRST, and NOT to the star ground. It should NOT go to the star ground first.

This is because the rectifiers charge the first filter cap with large pulses of current. These high pulse currents into the (+) side of the first filter cap also mean high pulse currents out of the first filter cap negative terminal, going back to the PT CT. These currents would run through the resistance of the wire to the star ground point. The wire resistance is small, but the current pulses are large, maybe 8-10 times the average DC current out if the first filter cap. This makes for a very low impedance voltage change across that wire. Those voltage pulses are impressed on the ground of everything. If the CT is connected only to the first filter cap's negative terminal, then only the average DC current comes out of the first filter cap to the star ground point. The charging pulses still cause a pulsed voltage across the wire to the PT CT, but this is not jerking the star ground point up and down.

[ChopSauce]

Thanks a lot R.G., for confirming with explanations. I'll try that as soon as possible!

Would there be problems if I let it like that, meaning one side ofvthe filament wirining at 1.5V and the other at 4.8V? It appears to be quiet then...

That seems fine to me.
(provided the voltage between both "sides" is well 6.3V, or close enough)

[R.G.]

Would there be problems if I let it like that, meaning one side ofvthe filament wirining at 1.5V and the other at 4.8V? It appears to be quiet then...

It helps to understand what is happening when the filament voltage is offset.
AC filaments are center tapped because the AC voltage across them induces a small voltage on the cathode, which makes audible hum. The center tap forces the two ends of the heater to induce equal and opposite voltages onto the cathode, which cancel. So far, so good.
There are other hum sources in amps, including when the tubes may have mechanical setups where the heater hum is not exactly equal and opposite. Forcing the heater voltage to be bigger on one side than the other lets you "bias" the hum so it's a better cancellation. In fact, this can be used to induce a hum which cancels other hum sources. This is the origin of the pot across the heater with a grounded wiper - it lets you "tune" the hum.
Does your amp have a grounded heater center tap? Or a low value pot across the heaters with a grounded wiper? Or two low-value resistors from heater ends to ground, which is another option?

[bieworm]

Would there be problems if I let it like that, meaning one side ofvthe filament wirining at 1.5V and the other at 4.8V? It appears to be quiet then...

It helps to understand what is happening when the filament voltage is offset.
AC filaments are center tapped because the AC voltage across them induces a small voltage on the cathode, which makes audible hum. The center tap forces the two ends of the heater to induce equal and opposite voltages onto the cathode, which cancel. So far, so good.
There are other hum sources in amps, including when the tubes may have mechanical setups where the heater hum is not exactly equal and opposite. Forcing the heater voltage to be bigger on one side than the other lets you "bias" the hum so it's a better cancellation. In fact, this can be used to induce a hum which cancels other hum sources. This is the origin of the pot across the heater with a grounded wiper - it lets you "tune" the hum.
Does your amp have a grounded heater center tap? Or a low value pot across the heaters with a grounded wiper? Or two low-value resistors from heater ends to ground, which is another option?

I have the filament winding's CT taped off an use the 100R hum pot. That wiper goes to a 50VDC reference.
The tubes also don't have filament center taps, they're 6SL7 tubes. At the Hoffman site someone told me those will hum anyway. Also that the uneven heater voltage is no problem.. there is still 6 3V across it.

[bieworm]

Do you mean without first connecting to the ground connection, which would be:

PT/CT -> minus(res.cap) -> starGround

instead of:

PT/CT -> starGround <- minus(res.cap)

Yes, explicitly. The PT CT should go to the negative of the first filter cap FIRST, and NOT to the star ground. It should NOT go to the star ground first.

This is because the rectifiers charge the first filter cap with large pulses of current. These high pulse currents into the (+) side of the first filter cap also mean high pulse currents out of the first filter cap negative terminal, going back to the PT CT. These currents would run through the resistance of the wire to the star ground point. The wire resistance is small, but the current pulses are large, maybe 8-10 times the average DC current out if the first filter cap. This makes for a very low impedance voltage change across that wire. Those voltage pulses are impressed on the ground of everything. If the CT is connected only to the first filter cap's negative terminal, then only the average DC current comes out of the first filter cap to the star ground point. The charging pulses still cause a pulsed voltage across the wire to the PT CT, but this is not jerking the star ground point up and down.

In that case I'm confused where to put my HT CT. I added the trinity trip top DC elevation circuit on my amp. It goes before the 1st filter cap. Didn't add a standby switch. Would this mean my CT gas to be put at the - of C21? I guess so...

Trinity Trip Top Schematic v17.pdf

[R.G.]

Yes, explicitly. The PT CT should go to the negative of the first filter cap FIRST, and NOT to the star ground. It should NOT go to the star ground first.

This is because the rectifiers charge the first filter cap with large pulses of current. These high pulse currents into the (+) side of the first filter cap also mean high pulse currents out of the first filter cap negative terminal, going back to the PT CT. These currents would run through the resistance of the wire to the star ground point. The wire resistance is small, but the current pulses are large, maybe 8-10 times the average DC current out if the first filter cap. This makes for a very low impedance voltage change across that wire. Those voltage pulses are impressed on the ground of everything. If the CT is connected only to the first filter cap's negative terminal, then only the average DC current comes out of the first filter cap to the star ground point. The charging pulses still cause a pulsed voltage across the wire to the PT CT, but this is not jerking the star ground point up and down.

In that case I'm confused where to put my HT CT. I added the trinity trip top DC elevation circuit on my amp. It goes before the 1st filter cap. Didn't add a standby switch. Would this mean my CT gas to be put at the - of C21? I guess so...Trinity Trip Top Schematic v17.pdf

It took me a while to ferret through the schemo. There are ... um... different ways to denote wire connections in schematics. In this one, when wires cross and do not connect, there is a little semicircular "hop over" shown. The PT CT runs down to the ground symbol with parallel bars marked "POWER GROUND". That point then runs across under C23, C24, and R43, then takes a left-hand turn up to a four-way junction. From there, a wire leads from the four-way to the left under C22, R39, and the DC HTR ELEV 50V label, to the negative side of C21.That negative side of C21 is the negative terminal of your first filter cap. The heater elevation is done by R38 and R39.

For the circuit I described, you would cut the PT CT loose from the path to the POWER GROUND symbol and reconnect it to the negative side of C21. The series combination of C20 and C21 are your "first power supply filter capacitor". The resistor string in parallel with them are there to ensure voltage balance, and to let you get that DC heater elevation thing.

I hope that the DC elevation works for you. I never found them to be much better than doing a good job of fixing the things that cause hum in the first place.

[bieworm]

Yes, explicitly. The PT CT should go to the negative of the first filter cap FIRST, and NOT to the star ground. It should NOT go to the star ground first.

This is because the rectifiers charge the first filter cap with large pulses of current. These high pulse currents into the (+) side of the first filter cap also mean high pulse currents out of the first filter cap negative terminal, going back to the PT CT. These currents would run through the resistance of the wire to the star ground point. The wire resistance is small, but the current pulses are large, maybe 8-10 times the average DC current out if the first filter cap. This makes for a very low impedance voltage change across that wire. Those voltage pulses are impressed on the ground of everything. If the CT is connected only to the first filter cap's negative terminal, then only the average DC current comes out of the first filter cap to the star ground point. The charging pulses still cause a pulsed voltage across the wire to the PT CT, but this is not jerking the star ground point up and down.

In that case I'm confused where to put my HT CT. I added the trinity trip top DC elevation circuit on my amp. It goes before the 1st filter cap. Didn't add a standby switch. Would this mean my CT gas to be put at the - of C21? I guess so...Trinity Trip Top Schematic v17.pdf

It took me a while to ferret through the schemo. There are ... um... different ways to denote wire connections in schematics. In this one, when wires cross and do not connect, there is a little semicircular "hop over" shown. The PT CT runs down to the ground symbol with parallel bars marked "POWER GROUND". That point then runs across under C23, C24, and R43, then takes a left-hand turn up to a four-way junction. From there, a wire leads from the four-way to the left under C22, R39, and the DC HTR ELEV 50V label, to the negative side of C21.That negative side of C21 is the negative terminal of your first filter cap. The heater elevation is done by R38 and R39.

For the circuit I described, you would cut the PT CT loose from the path to the POWER GROUND symbol and reconnect it to the negative side of C21. The series combination of C20 and C21 are your "first power supply filter capacitor". The resistor string in parallel with them are there to ensure voltage balance, and to let you get that DC heater elevation thing.

I hope that the DC elevation works for you. I never found them to be much better than doing a good job of fixing the things that cause hum in the first place.

When I have the time I will connect the HT CT to the - of C21 and see if that is reducing the hum in the first place. It won't be an easy job sunce I gave it twisted along with the HT wires right towards the GZ34...
You need to know I use NOS Reflector tubes for the 6SL7 and 6L6GC .. they are too suspect to hum.. but they last forever

[pdf64]

…
I hope that the DC elevation works for you. I never found them to be much better than doing a good job of fixing the things that cause hum in the first place.

My understanding is that DC elevation of the heater circuit helps to reduce hum / buzz in the case of imperfect heater - cathode insulation in the valves used.
I’ve found that it’s particularly effective in the case of partial and unbypassed early stage cathodes, which can otherwise be rather fussy in that regard about the valves used.
Tomer recommended it back in 1960, referred to it as heater biasing, see chapter 3 http://tubebooks.org/Books/Atwood/Tomer ... 0Tubes.pdf

[Stevem]

One thing I have noticed on any amp that has a hum pot is that it has two settings so to speak.
One setting where the idle hum will be the lowest with all the amps volume controls down, and another lowest hum setting with all the volumes full up.

You pick!

I have also noticed and maybe someone can explain why this hum modification situation.

For example on a Fender amp that is two channels such as a twin reverb if you turn up both channels volume pots a certain amount the idle hum level can be made to phase shift a get lower in intensity.

[R.G.]

My understanding is that DC elevation of the heater circuit helps to reduce hum / buzz in the case of imperfect heater - cathode insulation in the valves used.
I’ve found that it’s particularly effective in the case of partial and unbypassed early stage cathodes, which can otherwise be rather fussy in that regard about the valves used.
Tomer recommended it back in 1960, referred to it as heater biasing, see chapter 3 http://tubebooks.org/Books/Atwood/Tomer ... 0Tubes.pdf

Yes; I have a copy of Tomer's book, and have read that.

I have read other sources that say that elevating the heaters is useful for heater wires contaminated by oxides that reduce the work function of the heater wire surfaces, or where heaters have been contaminated by actual cathode oxides, either before or after the tube is evacuated and sealed, and by overactive electron clouds getting some electrons onto the heater wires just by the presence of the cloud. In all of these cases, the general idea is that a modestly positive heater wire will suck any contaminating electrons out of the heater wires and back to the power supply.

Good idea. The effectiveness depends on the impedance of the positive voltage "suck" and on the degree of imbalance of the heater wires around the elevating voltage, as one half-cycle of heater voltage is more positive than the other half cycle, so the positive voltage source has to be enough positive to ensure sucking them all out, even on more-negative alternations. A confounding factor is that the heater voltage is dropped across the heater. On any given half cycle, one end of the heater wire is more positive than the other end, so the more-active surface half of the heater wire emits more when it's more negative. The most-positive part of the wire gets its emitted electrons sucked away more effectively. The elevation voltage then needs to be way more than the heater voltage. Of course, 50 volts or so should be good.

... er, I think.

I loved the ideas, and launched into heater elevation in some of my early amps. I had modest success, as I noted. I got better results tracking down other sources of hum. Well, actually, let me rephrase that.
There are many sources of hum in your typical tube amp. Off the top of my head, in no particular order and probably not a complete list; AC mains leakage, magnetic induction from the PT and any wires leading from it, induction into the OT from the PT or heater wires; heater wire magnetic induction, heater wire capacitive leakage, wiring loop pickup from the magnetic parts, AC mains common mode voltage, AC mains repeating spikes (... fluorescents...), magnetic induction from outside the chassis; rectifier turnoff spikes, rectifier current spikes on the CT return, rectifier/ripple voltage, and probably a whole lot more. It always took me a lot of time to tune the layout and grounding to get rid of all of these I could. Way down in there somewhere, imperfect tubes and tube heaters would lurk.

Sorry - I'm drinking my evening drink and I got going. You get the idea: removing hum is peeling an onion.

One complicating issue is that the various sources of hum are not necessarily in phase with one another. So you can actually get less total hum by making one of the hums increase in magnitude to more effectively cancel some other hum source you're not working on. It's maddening. I really liked one article I read on a hum canceller that worked by taking the AC mains (resistor-dropped) and a half-wave rectified version of it and feeding this to a set of filters. The filters were tight bandpass filters centered on 60Hz (gotta correct if you're in a 50Hz country...) 120Hz, 180Hz, and 240 Hz. These were fed into all-pass variable phase stages and then summed after volume controls. This composite and phase shifted hum signal was then fed into an offending amp. The volumes were set to all zero, then the 60Hz output turned up a bit. The overall hum would increase a bit, so the phase control for 60Hz was tuned for lower hum. The 60Hz volume control was then adjusted more, then the phase again, until hum was as low as practical. Next the 120Hz controls were tinkered. Then the 180 and the 240. At the end, you had pretty good cancellation of all the hum sources for (1) that amp (2) that AC mains socket (3) that magnetic environment. If you were good, you might even tun out some of the hum from a single coil. Or an unbalanced heater. Heater balance controls are a similar idea, just with only the AC mains frequency and no phase control.

One thing I have noticed on any amp that has a hum pot is that it has two settings so to speak.
One setting where the idle hum will be the lowest with all the amps volume controls down, and another lowest hum setting with all the volumes full up.
You pick!
I have also noticed and maybe someone can explain why this hum modification situation.
For example on a Fender amp that is two channels such as a twin reverb if you turn up both channels volume pots a certain amount the idle hum level can be made to phase shift a get lower in intensity.

Actually, that makes great sense to me, given the many-sources-of-hum theory above. I'd be a little surprised if it was not so.
Edit: The turning up the volumes brings up small hums in the sections before the volume controls and adds them to the mix, most likely in a different phase. So a different canceling hum level is needed from the heaters.

[pdf64]

… on a Fender amp that is two channels such as a twin reverb if you turn up both channels volume pots a certain amount the idle hum level can be made to phase shift a get lower in intensity.

Most hum will probably be due to the input stage. With a 2 channel reverb model, the channels will have opposing polarities. Hence the hum from the respective input stages will be in antiphase, and so might be cancelled out by setting each channel’s controls to produce signals of matching amplitude.