Voltage doubler/ regulated filament questions.

[argonaut]

I know that the subject of regulated filament supplies has been discussed quite thoroughly, but I haven't seen any (or much) mention of using a spare 5V winding if available.

I'm currently working on a new build that is a perfect candidate for regulated DC. It has a stupid amount of preamp gain and distortion and the PT has a 5V @ 6A winding that's doing nothing.

I was thinking that I could just use a voltage doubler circuit (like attached) with the 5V winding and a 7806 regulator and it would be golden for two 12ax7s, right?

If that's the case, is there a general rule/guideline that should be followed when selecting the value of the capacitors in this type of circuit? I assume, from looking at other DC supply schems, that I'll need a cap after the regulator, how is its value determined?

Thanks!

[Jana]

I wouldn't even use the 7806.

I would use a 4700uf cap to filter the doubler output, then use a resistor to drop the amount of voltage you need and then another 4700uf cap to filter the supply for the filaments.

Since the filament current demand is a going to be a constant 600ma. for 2 AX7's, the resistor will do the job of dropping the voltage to a constant level.

The resistor will need to be 5 watt or so. Value? Somewhere around 0.47 ohms to a couple of ohms.

That's what I would do.

[argonaut]

Thanks for the suggestions, Jana. It's very much appreciated!

I'm not opposed to your idea of using a resistor, but I have some 7806s in hand and plenty of chassis space for heat sink. I think the closest small value, higher wattage R's I have in stock are 2ohms.

As far as the caps in the doubler circuit itself, how do you determine their value? Is the sky the limit for this type of application?

If I were to use a regulator, is this the correct layout? :

Voltage doubler circuit>filter cap>regulator>filter cap>filaments>neg. grounded after last tube

[Phil_S]

I have fooled around with a doubler a bit, but haven't actually put one into a circuit. My experience is voltage in x ~2.4 = voltage out. So 5 will give you ~12. Amps will be cut by the actual voltage multiplier factor, but at .15A per tube at 12V, you ought to be able to make this work comfortably.

I've made an assumption here...you weren't thinking about 6VDC were you?

[argonaut]

Hi Phil,

Sorry I wasn't clearer. I am thinking about 6VDC.

I was playing around with doubler circuits on a breadboard yesterday and it looks like it'll work fine. I was able to get a constant 6V (for the most part, the reg. wasn't heatsinked) feeding two bench tubes with all sorts of cap values. The problem is that although I can see that it works, I can't hear its effect at the moment. I'd like to be sure that my filtering is correct.

[David Root]

I'm using a 6.3 to 12.6V doubler with 3x1000uF 25V caps and a 12V 1.5A regulator chip. Feeding three 12AX7s, 0.9A total. ***Correction!! 0.45A total***, I was thinking at 6.3V

If I apply the 2.4 factor someone noted earlier, that turns 0.45A DC into 1.08 A AC going into the doubler off the 6.3V PT sec. tap. Is that right?

I'm using 22 ga. wire from the PT tap to the DC supply board and wondering if that should be 20 ga. Somehow I think it's OK as is. (Stranded, teflon insulation).

What think ye?

[paulster]

Here's a SPICE plot showing the circuit you posted driving 2 12AX7s using 4700uF caps (in red) and 10000uF caps (in green).

The ripple is way too high to use on its own without a regulator really, but you have plenty of voltage available there as long as you ensure that the minimum voltage it drops to is above 6.3V plus the regulator dropout voltage, plus a bit of headroom to accommodate a low wall voltage situation.

[Jana]

So it's looking like it would be about 9.7 volts with the 4700uf cap.

What would the plot look like if you added a 5.6 ohm resistor to drop 3.4 volts and then another 4700uf cap? (with 600ma current for a load).

[paulster]

Here's the same thing with an 8ohm (green) and a 9ohm (red) resistor and then another 4700uF cap, so three 4700s in total. Load is 600mA (2x 12AX7).

The ripple on top of the red plot is about 40mV p-p, so about 14mV RMS.

This is in the hypothetical world, of course, and doesn't factor in the quality of the caps, but it's certainly usable for a pair of preamp tubes.

[argonaut]

Thanks much, paulster!

I have three 10,000uf caps in hand, so I'll most likely use them. I'll check to see what R's I have, but regardless of the method I use (either resistor or regulator), should I use another filter on the output of the circuit before the tubes?

[paulster]

I have three 10,000uf caps in hand, so I'll most likely use them. I'll check to see what R's I have, but regardless of the method I use (either resistor or regulator), should I use another filter on the output of the circuit before the tubes?

You want the two caps exactly as you had them and then either the resistor and a third cap after it before the tubes (that's what I've illustrated above) or your regulator and then something like a 10uF tantalum cap between the regulator and tubes right alongside the regulator for stability (the low ESR of tantalums is great for minimising any oscillations in the regulator output). You won't need a large cap on the regulator output as it serves only to stress it at startup.

[jaysg]

Sorry I wasn't clearer. I am thinking about 6VDC.

I'd go with the 12 to 13V you'll see with it simply doubled. You can tweak the exact output by adding diodes or changing diode types -- standard vs. Schottky vs. two in series where you show one. Use 3A devices.

[argonaut]

I have three 10,000uf caps in hand, so I'll most likely use them. I'll check to see what R's I have, but regardless of the method I use (either resistor or regulator), should I use another filter on the output of the circuit before the tubes?

You want the two caps exactly as you had them and then either the resistor and a third cap after it before the tubes (that's what I've illustrated above) or your regulator and then something like a 10uF tantalum cap between the regulator and tubes right alongside the regulator for stability (the low ESR of tantalums is great for minimising any oscillations in the regulator output). You won't need a large cap on the regulator output as it serves only to stress it at startup.

Excellent, I got it now. Thanks for taking the time.