Elevated heaters?

[Rhodesplyr]

I've seen some rather dramatic hum reduction in older guitar amps by balancing the heaters to ground with resistors or a hum-balance pot, but I've yet to find a situation where making the heaters positive with respect to the cathodes provides anything like the same degree of noise reduction.

Does this jibe with others' experience? I ask because you often see both techniques recommended by the same sources.

(I know that elevating the cathodes can help prevent heater-cathode shorts where there would otherwise be a large heater-cathode voltage difference, but that's a different problem.)

[soundguruman]

I've seen some rather dramatic hum reduction in older guitar amps by balancing the heaters to ground with resistors or a hum-balance pot, but I've yet to find a situation where making the heaters positive with respect to the cathodes provides anything like the same degree of noise reduction.

Does this jibe with others' experience? I ask because you often see both techniques recommended by the same sources.

(I know that elevating the cathodes can help prevent heater-cathode shorts where there would otherwise be a large heater-cathode voltage difference, but that's a different problem.)

Welp, then your hum is reduced for sure by making the first two preamp tubes DC heaters, the hum mainly comes from the first stages of the preamp, and then is amplified by the next stages...so concentrate on the first two preamp tubes to reduce the hum.
As I have said before, in a high gain amp this is really an issue. In a lower gain amp, the hum is not to be amplified nearly as much.
I mean making the heaters positive to the cathode...OK...but get ALL the AC ripple out of the filaments is what you really want to do.
A good DC heater circuit example is the Macintosh stereo preamp, which uses 10,000 uF caps for filters. It's pretty hard to filter DC when it is loaded by an almost short circuit condition, that is what the filaments essentially are. So, it's a bit challenging.
PV uses a diode to ground to make a half wave rectifier, with no filter cap, which can be useful in a lower gain circuit...BUT in a high gain circuit you really want a full wave rectifier and a massive capacitor. Essentially, it needs to run the filaments off a battery, no matter if it's positive to the cathode or not--what you really need is to remove the AC entirely from the preamp section.
Out of balance output tubes will also hum. Soooo, the balance circuit from like a fender twin design--will help get rid of the remaining hum, even after the preamp tube filaments are DC.
The best low noise designs use DC heaters for the preamp tubes. BUT the larger the filters, the longer the heaters take to charge up to operating voltage, so the warm up time is greatly increased. (the 10,000 uF cap slowly charges up to full voltage).
Another thought is to utilize the center tapped heater of the preamp tube. If you have 12 volts to run the preamp heaters, you can hook another filter cap to the center of the filaments.
Balance the heater winding with resistors, a pot, or a center tap...yes. But this is only a start if you are really going for low noise high gain.

[Enzo]

There are many potential sources of hum. The heaters need a reference to ground, and even elevating them does that. The reference can be balanced or not. SOme old Fenders just grounded one side of the heaters. But that is enough to kill a lot of that hum. Cathode followers aside, elevating heaters stops POTENTIAL hum from heater to cathode currents. They don't always occur, and would not be of the same magnitude as just plain floating heaters even when they do occur.

The point of ssuggesting both methods is they they stop different sources of hum. Imagine hum is cold air leaking into your home. You need to weatherstrip each window, and each door, and insulate the roof, etc. Each leak has its own cure. Taping one window has no effect on the next window. Adding six more layers of tape to the first window does nothing to solve the rest of the problem. Hum is not generic. The cathode/heater current hum is not the same as the ungrounded heater hum. Oh it may SOUND the same but that doesn;t make it the same.

It's pretty hard to filter DC when it is loaded by an almost short circuit condition

I am not sure what to make of that. A 12AX7 wired for 6v draws 0.3 amps - 300ma. And as far as I know, that is 21 ohms hot. But meanwhile, even if you have ten of them in series, the draw is only 3A. It is not difficult to filter a supply for a 3A load.


If there is current flowing from heater to cathode from the filament supply, on the negative peaks of the heater current, the heater becomes less positive than the cathode - potentially - and the cathode then attracts the electrons. Current flows. Hum results. Elevating the cathode is simple and effective. As long as the elevation prevents ANY PART of the AC waveform at the heaters from becoming more negative than the cathode then those currents won;t flow. Doesn;t matter if the heaters are running AC, the magic is whether the cathodes ever become positive to the heaters. But note that only cures heater/cathode hum. Radiated 60Hz is a separate issue.

[Steve Conner]

A few years ago, I was building a tube hi-fi amp. With the heater winding centre tapped to ground, there was still a fair bit of hum. Swapping out tubes in the first stage altered the hum, but I didn't have any tubes good enough to banish it completely.

Elevating the heater supply to +40V made a huge improvement.

[trobbins]

The measured results I've come across show the heater cathode resistance is typically a minimum at 0V diferential - and hence if leakage current via the resistance path is noticeable then the aim is to keep the minimum differential somewhat greater than zero, with little to be gained in resistance increase when the minimum differential exceeds about 10V (ie. the heater AC midpoint would be about 20V from cathode for 6.3V AC).

http://dalmura.com.au/projects/Heate...erformance.pdf

http://dalmura.com.au/projects/Heater-cathode conduction plots.pdf

[Rhodesplyr]

Thanks, Enzo. The situations I'm talking about are in amps without center-tapped heater windings where only one side of the heater supply is grounded. I've found that balancing the heater supply to ground via the 100 Ohm resistor method often lowers hum quite noticeably. But I certainly agree with your point about different sources of hum.

I also appreciate the point about making sure that no portion of the AC waveform ever becomes negative with respect to the cathode. A 6.3VAC sine wave has a ~8.9V peak, so if you had a +1.5VDC cathode voltage, you'd want the heaters to be elevated *at least* 10.4VDC above ground potential.

[Rhodesplyr]

A few years ago, I was building a tube hi-fi amp. With the heater winding centre tapped to ground, there was still a fair bit of hum. Swapping out tubes in the first stage altered the hum, but I didn't have any tubes good enough to banish it completely.

Elevating the heater supply to +40V made a huge improvement.

Steve, Was this a power amp, preamp, or integrated amp?

[Rhodesplyr]

The measured results I've come across show the heater cathode resistance is typically a minimum at 0V diferential - and hence if leakage current via the resistance path is noticeable then the aim is to keep the minimum differential somewhat greater than zero, with little to be gained in resistance increase when the minimum differential exceeds about 10V (ie. the heater AC midpoint would be about 20V from cathode for 6.3V AC).

Sounds like it might be an interesting experiment to use a pot to vary the heater elevation point to see where the hum-reduction effect became significant.

[Steve Conner]

Steve, Was this a power amp, preamp, or integrated amp?

It was this thing:
scopeblog » Crown Super SXA

A 30Wpc stereo power amp.

It's a Williamson style circuit and as such the second stage is a direct-coupled cathodyne phase splitter. The high DC voltage on the cathode of this stage maybe made it respond better to a higher heater elevation voltage.

[Steve Conner]

Steve, Was this a power amp, preamp, or integrated amp?

It was this thing:
scopeblog » Crown Super SXA

A 30Wpc stereo power amp.

It's a Williamson style circuit and as such the second stage is a direct-coupled cathodyne phase splitter. The high DC voltage on the cathode of this stage maybe made it respond better to a higher heater elevation voltage.

[kleuck]

I've seen some rather dramatic hum reduction in older guitar amps by balancing the heaters to ground with resistors or a hum-balance pot, but I've yet to find a situation where making the heaters positive with respect to the cathodes provides anything like the same degree of noise reduction.

Does this jibe with others' experience? I ask because you often see both techniques recommended by the same sources.

(I know that elevating the cathodes can help prevent heater-cathode shorts where there would otherwise be a large heater-cathode voltage difference, but that's a different problem.)

Sounds strange to me, as i improved a lot of amps (already fitted with a ground reference via resistors) by elevating the reference, even using the old trick of the power tube cathode voltage (actually, the only trick i use)