Heaters - DC vs elevated voltage

[pjd3]

Hello,

My current build is a moderately high gain tube amp so, anything I can bring to it to minimize noise is fine by me. However, I come across 2 different approaches for addressing potential 60 cycle filament hum. One is to use DC voltage for the heaters. This intuitively makes sense for lowering hum inductance. The other method I hear about is elevating the voltage of the heaters. It seems people tap from different sections of the amp for this, one of the last reads he tapped off the screen B+ and connected that to the 6.3vac filament center tap.

What I haven't seen anything about is how effective one is compared to the other, or if there are circumstances where one is the more prudent approach over the other. So for instance, since I'm buying a few boards very soon from Headfirst amplifiers, I see he has a DC filament board for 10 bucks or so. But then hooking up the Filament coil center-tap to the screens through a little divider network isn't much to ask for.

I'd be really interested in your take on how they compare effectively and what you would use and why. Thanks.

Also, I find it interesting if not suspicious that those gearing up for the DC approach talk about creating their "6.3 volts DC" for their tube filaments. If we were changing 6.3vac to a DC voltage, wouldn't we want to use the RMS of 6.3? Where that would be providing the same energy over time that the filaments were designed to see? That bewilders me a bit. Pretty much everyone talks about using 6.3 volts DC. Wouldn't that be forcing the filaments to see 30% more energy than they were designed to operate with?

Thanks for your insights and knowledge on this. I have to pick one !

Best,
Phil Donovan

[martin manning]

6.3 VAC is an RMS value. If the AC is a perfect sine wave, it is equivalent to 6.3 VDC when connected to a resistive load.

[Stevem]

I would build and then first try a tube like the Electro Harmonics 12AX7EH, or the JJ electrons ECC83S.
Both of these tubes have spiral wound filiments and are very very 60 hz hum free to start with!

[trobbins]

The insight you are after is that any noticeable, or indeed measurable, change to noise/hum can be affected by the (a) the tube you use (both sample to sample, and model), and (b) by circuit layout and whether you can neutralise any hum/noise ingress by such means as a humdinger pot, and whether the means to generate DC is itself a cause of hum/noise that gets in to the audio via other paths. So imho the only way you can answer your query is to set up a few alternatives and measure/listen. There are some articles that go in to the details more, but they don't change the issue for you.

Also note that heater powering may not be an issue at all for you, as it is only one of many possible causes of hum/noise in an amp - but there in lies the journey you are embarking on.

[pdf64]

DC elevation of the heater circuit has benefits additional to reducing hum caused by imperfect heater cathode insulation. eg the stress on that insulation in DCCF stages is reduced, the likelihood of anode to heater arcing on output valve bases / sockets may be reduced.

[rangdipkin]

there's no reason you can't do both! I always tend to elevate the AC heaters a little for the aforementioned benefits to cathode follower longevity. For super high gain stuff I like to use a small transformer (ie hammond 266J6) with its own bridge rectifier and filter cap to supply my first tube, and simply connect the ground of the filter to the elevated reference voltage.

[Stevem]

For the first gain stage which is the most sensitive to noise here is a handy way to do that right off the filiament voltage that incorporates the 100 ohm hum balance resistors if you need those also.

[Colossal]

I almost always elevate the heaters to a DC reference, especially for any amp with a cathode follower(s). It's easy to do and only takes a few parts. For high gain amps, a DC first stage can be beneficial but not always required.

[TUBEDUDE]

I'm a fan of elevated heaters. Some higher gain topologies require more though. On one occasion I used a scheme like the one SteveM has shared, I got diode switching noise on another low level stage (FX loop return). On the odd chance you experience that, just put small caps across each diode.
Be aware making D.C. for heaters draws more current. So make sure your heater winding is not maxed already.

[Stevem]

Sorry, yes I should have mentioned the potential noise issue.

A .01 600 volt ciramic cap works good for snubbing that, or better yes fast recovery type diodes.