U47 -simple voltage doubler filament supply for EF14?

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

MagnetoSound

Well-known member
Joined
May 12, 2008
Messages
1,582
Location
Southern England
I have been pondering the fact that the U47 traditionally runs just a single unregulated power rail to both the filament and the plate of the VF14 tube ...

I am used to the idea of regulated filaments being the way to keep both a steady tube gain and optimum life expectancy for the filament - and yet this most hallowed of microphones gets off completely scot free!

So this got me wondering why the filament supply for an alternative tube (I've chosen an EF14) couldn't be just a duplicate of the HT rail, using a voltage doubler to allow similar filtering to that provided in the NG PSU, but from just a regular 6.3v winding?

I have taken as a reference Mr. Archut's recommended filament voltage of 5.05v and calculated filament current of around 380mA. Dissipation across the total R is 2.9w, total drop is around 7.55v, not too much more (and sometimes less) than you would find in a traditional discrete current regulator.

Filter cap values give cutoffs at around 7Hz, will this be good enough at three octaves below line freq (this supply will be providing cathode bias too, remember)? Ramping takes around 26s.

The big gap in my knowledge is in how voltage doublers behave with regard to ripple and switching noise. Also how much current the filament winding will need to supply at power up. Would a 1A heater winding cut it? I haven't seen doublers applied in microphone supplies as far as I am aware and I would be grateful for any insight into anything I may have overlooked. I know this must have been discussed before. I did search but didn't manage to find any definitive answer.


EF14_FIL_PSUDR.jpg


 
You get 1V ripple for 1 Amp and 1,000uFd.

You have under half-Amp and 5,000uFd. First cap runs 0.1V ripple.

Ripple-rate is actually 100/120hz here.

7Hz filters give 14:1 or 0.07 attenuation.

0.1 * 0.07 *0.07 *0.07 *0.07 = 0.000,002,4V ripple on heater line.

Unsurprisingly (since the similar plan worked in U47), this should work.

> Would a 1A heater winding cut it?

Kinda.

Multiply the 0.4A by 1.6 to get AC RMS, then by 2 for the doubler. It's a 1.2A AC load. What happens when you overload 20%? The voltage sags and the copper runs hot. However a 6VA transfo has a high surface/volume ratio. Heat escapes good. It will sag something awful without burning up. And you have heaps of spare voltage-drop you can trim to compensate the sag.

The start-up load is 3X-4X bigger, but only for a few seconds. Just more sag, not long enough for the heat to matter.

If this is a $5 fun project, you are fine.

If you have money/time invested and expect to use this mike seriously, I'd ponder a larger iron just to be very sure.
 

Latest posts

Back
Top