Elevating Heaters from B+ with bleeder resistors

[tubeswell]

Has anyone got a scheme for running elevated heaters off the bleeder resistors? I am building an amp with two 100uF in series for the reservoir, and it also has three CF, so I thought I'd try elevating the heaters by about 40V. But if I run another divider from the B+, it will be in parallel with the bleeders, so I was trying to think of a way to do this.

Can I just double the value of the bleeders (to say 470k each) and then have a divider in parallel with that, which on its own has a total value of 1M across R1 and R2 but with the ground leg being something like 1/10 the value of the other one (i.e. 910k for R1 and 91k for the ground leg R2) - where the B+ is around 430VDC?

(Sorry for the dumb-ass question)

Edit - I just realised that I could decouple the knee of the voltage divider for the heater with a small 100V cap, which would mean that the divider wouldn't be in parallel with the bleeder circuit anyway. (But feel free to chime in with any flippant remarks)

I also have a 50VAC winding for the bias supply on the PT, which I was going to rectify one side of for a negative supply for the bias. But could I use a bridge rectifier across the 50V winding and get a negative supply off one side for the bias and a positive supply off the other side for the heater? (Second dumb-ass question)

[loudthud]

Use a 220K on the top cap and make a voltage divider on the lower cap with a 180K and a 39K. That will be pretty close to 40V.

[tubeswell]

Thanks loudthud. Is there any benefit/disadvantage in adding a small (say 22uf-47uF 100V) decoupling cap to the knee of that (lower) voltage divider (do you think)? (Edit - I did a quick schematic of what I was asking about - see the red cap)

[black_labb]

i dont think there would be an advantage. its not signal so there shouldnt be a problem.

[loudthud]

In a high gain amp the heater wires can become an unintended feedback path causing an oscillation. I always bypass elevated heater bias with 10uF or so.

[NorCalTuna]

In a high gain amp the heater wires can become an unintended feedback path causing an oscillation. I always bypass elevated heater bias with 10uF or so.

This is good to know... do you know if the path can get set up in DC fil. as well, and how to prevent it? If it's inherent and unavoidable, there's a (another) really good reason to eschew DC heaters and do the elevated gnd....

thanks!

[Merlinb]

This is good to know... do you know if the path can get set up in DC fil. as well, and how to prevent it? If it's inherent and unavoidable, there's a (another) really good reason to eschew DC heaters and do the elevated gnd....

thanks!

DC heaters will be supplied form a very large smoothing capacitor, or even an IC regulator. Both perform the same noise-shunting funtion, so DC heaters will be quiet.

[Steve Conner]

Hola el tubeswellos! Your schematic looks fine. I also agree that it's wise to bypass the elevated heater node with a capacitor.

Merlinb: This is an issue of ground referencing, since the feedback path is common-mode between heater circuit and ground. If your DC heater circuit wasn't grounded, then it would be a feedback path all the same. The - terminal should be connected to chassis.

On this note, I always WTF when I see people use the same heater winding to power both AC and DC heaters. If you're not real careful with grounding on this setup, you can either short your heater winding, or superimpose AC back onto your DC. If the heater winding has a center tap, this should be lifted from ground, and any hum balance pots or resistors removed. The negative terminal of the DC side should then be grounded.

[Guitarist]

I'm at this point too. My Hiwatt has a center tap for the heaters. I should lift this and float things through this centertap, right? I understand the resistors and hum balance pot, but centertaps not so much.

I just did this to an Ampeg V4 and WHAT a difference!! Especially with the reverb. Even drenched, the hum dropped to about 20%.

Awesome approach/great diagram! Thanks.

[tubeswell]

I'm at this point too. My Hiwatt has a center tap for the heaters. I should lift this and float things through this centertap, right? I understand the resistors and hum balance pot, but centertaps not so much.

Yep that's right. The heater winding CT (or the wiper of the hum-balance pot, or the 'ground' side of the 100R "artificial ground" resistors, if you are using those methods instead) goes to the knee of the voltage divider.

[mirlo]

Hi guys,my english is very bad,very nice topic,a question, is any problem with conect the center tap to the midle of the 2 first capacitor,without a voltage divisor? Is too much voltage?maybe is moore simple and is not necesary another cap,maybe i am wrong,thanks

[tubeswell]

Hi guys,my english is very bad,very nice topic,a question, is any problem with conect the center tap to the midle of the 2 first capacitor,without a voltage divisor? Is too much voltage?maybe is moore simple and is not necesary another cap,maybe i am wrong,thanks

Depends how high you want the heater-to-cathode voltage. The key determinant is the maximum rating in your typical pre-amp tube, which is usually around 100 Volts (before the cathode insulation begins to break down), so you want to keep your heater-to-cathode voltages lower than that.

If you ran the heater winding ground reference from between the two sets of bleeder resistors in a typical series filter cap set-up (where the values of the capacitors is the same), you would get a voltage divider that gave you half the B+, which is generally too-high-a-voltage (e.g.; it would be ~215V in the schematic I drew earlier). So you would need to knock it down some more.

[catnine]

I have never had one problem with heaters or noise . I always wire the heaters in first and twist the pairs and make certain that each wire off the heaters goes to the same area on each tube socket and that they are against the chassis . I only used the resisters when there was no PT center tap for the heaters and have good results with that but still mark the leads and make certain they go to the same pin on each tube and are closer to the chassis than any other wire . Or use two colors of the proper gauge wire so you know without doubt that the heater wires are all on the same side of each tube and not out of phase .

I keep the power scetion and preamp and input grounded as seperate sections and keep them away from eachother and neve had one single noise of hum issue this includes the filter cap grounds keep the resevoir seperate form the other filter caps as far as grounds go. I found out with a re-cap on a fender MM bass amp where I gounded all the filter caps on one place that created hum , I moved them apart hum gone .

There are three sections so design them to be seperate sections and I bet you will end up with a quiet amp even a SE amp. Placement of mod switches does matter .

[tubeswell]

Elevating the heaters is tackling a different problem from using 2 x 100R resistors on either side of the heater winding. The heater winding CT (or the other side of the 100R resistors, or the wiper from the hum-dinger pot) is connected to the point where you are elevating the heaters, so that the whole voltage reference point for the heater circuit is different from the rest of the ground return potential (which the cathode resistors are referenced to in cathode bias). The idea is that where you have high cathode voltages, such as in CF stages, you can reduce the effective heater-to-cathode voltage (and hence the reduce the risk of destroying the heater/cathode insulation) by raising the heater to ground reference voltage.

[Steve Conner]

+1. The only time when I've needed to use elevated heaters was on an old Crown stereo amp that I restored. It had a slightly bothersome hum even with the heaters centre tapped to ground, so I tried elevating the centre tap with a 40V DC supply that I got from a spare bias winding on one of the PTs. This made the hum disappear completely.

It was a Williamson design, so it had a cathodyne phase splitter with gain stages after it. This means the cathode of the PI tube would be sensitive to leakage from the heaters, and that's probably why elevating the heater circuit helped.