Biasing DC heaters

[lazthegaz]

I'm in the process of building a preamplifier with ECC83 and ECC88 tubes. Both tubes work in the conventional common cathode stage, so the cathodes are only a few volts above ground. I have a completely separate regulated DC supply for the heaters, so I have the option to bias the heaters to whatever positive voltage I choose.

So the question is: what do you recommend? I know that the subject of heater biasing has already been discussed here

http://www.diyaudio.com/forums/showt...p?postid=90211

but that was for AC heaters. With DC heaters it's a bit different. First, even if I connected one end of the heater to ground, there would still be a potential gradient along the heater, so some of it would be below the cathode potential (as the cathodes will be a few volts more positive than ground), other parts above it. Second, heater hum should not be a problem with regulated DC heaters.

My gut feeling would be to bias one end of the heater to maybe +10 V above ground so that the entire heater is at a positive potential. Is this right?

[Frank Berry]

When using a d.c. heater supply, it's really not necessary to put the heaters at a positive potential.

[Sch3mat1c]

Don't sweat it. 50V, yes -- 10V? pffftthbt.
If you must be a purist, you can elevate it a bit. The only requirement really is AC grounding it so PT crud doesn't get coupled into the cathode(s). For all I care you could let it float wherever it pleases.

Tim

[The-Planet]

Tim -

Can you clarify...

"The only requirement really is AC grounding it so PT crud doesn't get coupled into the cathode(s). "

I'm not sure what you're getting at.

Personally I've found that sometimes even DC filaments are quieter if you bias them a bit positive. Less so with some tubes, but with others like the 12B4 it made a huge difference in background noise.

- Gary

Tube Stuff

[Sch3mat1c]

The heater winding is capacitively coupled to the primary or HT winding of the transformer. This in turn can be coupled to the cathodes, particularly if they are unbypassed. Adding a cap to ground shunts this noise on the winding.

Tim

[fdegrove]

Hi,

When using a d.c. heater supply, it's really not necessary to put the heaters at a positive potential.

I'm sorry but I fail to see why the use of a DC heater supply alleviates the need to bias the heater upwards if that's what's called for.

Remember the heater to cathode insulation has a certain breakdown voltage and as it forms an interelectrode capacitance noise can leak through in both directions.
You'll often be able to hear and measure some signal breaktrough on the heater lines, one more good reason to use dedicated heatersupplies in sensitive circuits and avoid crosstalk between stages and channels in this way.

The main reason to bias the heater upwards is to bring it's potential as close to the cathode potential as possible.
This avoids sticky heaters and shorts leading to either whistling (VHF noise), ticking and/or popping noises emanating from unwanted heater to cathode interactions.

Emission can occur when the potential difference is a mere 30V (DC or AC it doesn't matter). While heaters aren't really coated with emissive material they're nonetheless quite capable of emitting electrons although not for a prolonged period of time.
Moreover, as they're usually seated inside the cathode sleeve, not much harm is done but the whole situation is best avoided for best operating conditions.

Cheers,

[Frank Berry]

fdegrove

Thanks. I learned something new. I didn't realize that biasing the d.c. heater to a positive potential would reduce noise.
I knew that this worked on a.c. powered filaments.
I guess you're never too old to learn.

[13D3]

What happens if you use either side of a dual dc supply (+6v - 0 - -6v), giving 12v DC across both supplies, and apply that to ECC83 heaters? Are they negative biassed in relation to Cathode, if it is the 0 which is earthed, and not the minus side of the dual supply? I imagine this should cancel hum, by reversing heater to cathode leak; but would this be a problem for the safe Cathode to Heater relation?

Any advice gladly received.
13d3

[tomchr]

As far as I understand it, biasing the cathode-heater interface with a slight positive voltage (or 0 V), reduces the coupling from the heater to the cathode. The interface is basically like a diode. If it's forward biased (cathode voltage higher than heater voltage with respect to ground), you get more hum. With DC heaters this should be less of an issue, but I'd still avoid having that interface forward biased.

I tend to connect the heater to the cathode through a 100 kOhm resistor. It's large enough to not conduct much signal but small enough to keep the heater biased around the cathode voltage.

~Tom

[trobbins]

The cathode to heater impedance has a capacitive component - which doesn't change with dc voltage differential. It also has a dc resistance component which does change with dc voltage differential. The measured curves I've seen don't support a 'diode' description, and vary wildly between valves, but do have a characteristic minimum resistance at 0V differential. I think the 'diode' comparison has been made because the resistance increase away from 0V differential is often not symmetric with positive versus negative differential - but I've also seen curves where it is basically symmetric.

I've only been interested in 12A-7 type valves, and for that variety, the DC reistance characteristic is often negligible compared to the capacitance leakage, and hence there is little if any performance improvement in many of that type of valve when elevating to a DC level - but there can also be samples where it has an advantage. And because the capacitive part soon dominates with frequency, any 'hum' reduction impact is focused mainly on the main fundamental and 2x, rather than any higher frequency rectifier noise.