Switching supply for 300B filaments

[Zen Audio]

Greetings. I am new to this forum. About a year ago I bought an Antique sound labs AQ-1002 tube amp to replace a very nice B&K EX-442 Sonata. I was surprised how sweet the tubes sounded in comparison. There are still a few things the B&K does better, but for overall listening the B&K now sounds thin to me in the upper mids and highs.
Anyway now, decided to try to build my own and for about 3 weeks now I have been reading about tube amps, learning all I can and designing on paper some amplifiers. I have finally settled on a design that is a mix of a couple different designs. I want to build a 300B with 3 in parallel each side. Yes 6 total, I just had to mix something in there I hadn't seen before. Plus I will be driving 89DB efficient Polk SDA-2b tower speakers with it. I have parts on order from all over the place LOL
I am concerned with noise on the 5VAC cathode heater. I have found a 5VDC 12A regulated switching power supply that I am strongly considering for the filaments. I would return the filaments to ground through a small resistor network or hum pot off the filaments at the tube.
I have read quite a bit on this forumn, but not found the exact answer to the questions I have. Has anyone used a regulated switching supply to supply the filaments on a 300B or similliar? Also how to you control the inrush current at start-up? Since the supply I am looking at has overcurrent protection I assume it would shut down or toast a fuse on start-up when the heavy in rush current comes in to the cold filament which I understand can draw 5-10 times operating current at start up.
I would appreciate any help.
Thanx Much

[dsavitsk]

http://pmillett.com/hf_fil.htm

[Zen Audio]

Thank you for the input. That is an interesting concept and makes complete electrical sense. However since my 6 300B's will be drawing in excess of 8 amps filament current, I would need 4 such circuits.
In regards to a regulated 5V 12A supply, is there a way a choke could be used to limit the current draw during start up?

[tjl]

5V 12A regulated switching power supply for 300B 8A filaments supply seems no problem for power on inrush current status.

You must check switching power supply overload protection first, just connect 6 tube's filaments parallel or use 6 pieces 4 ohm power resistor substitute each tube's filament then turn on switching power supply to see if it can withstand overload during power on maximum inrush current condition.

Or using 10,000uf / 16V electrolyic capacitor parallel to 5V 12A output terminals this will supply additional current for inrush current suddently.

If just beyond overload protection, adjust 5V to 4.8V for less current and longer life time of tubes.

Cheers

[Zen Audio]

Thank you so much for the reply TJL. Attached is the specs on the supply. I was also looking into NTC thermistors in series with the filaments. I did some calculations and figured that in order for 6 filaments in parallel to draw 7.2 A filament current (1.2A x 6) ideal, they must be presenting a total resistance in parallel of about .69ohms to a 5 volt supply.
If I put an NTC thermistor rated .5 ohms at 25C in series with each leg of the 5 V supply I could limit the inrush which is also good for the tubes. They are rated at 16A max current draw, and at heated value of 50% of max, ballpark where the tube filaments would draw, the thermistors drop to I believe .04 ohms. I could easily fine tune any drop caused by this .04 ohms to reluate correct voltage on the actual filaments.
Does this make sense

[tjl]

NTC may be the best way for controlling inrush current, I use it series at transformer primary winding, at B+ supply, also at filaments (individual) low voltage supply.
If resistance decrease to only 0.04 ohms after several seconds, it seems suitable for 6 300B 7.2A power supply,
but if NTC resistance tolerance higher than your expect value, then the heat dispation will be another problem, may be higher than 5 w on one NTC body surface not 2.1 w you calculated.
So I use several NTC for each tubes or pairs if I need.

Cheers

[Zen Audio]

Yes, i could see dividing the lines up and using more pairs. I will have to give that some thought as well.
What value do you use on your power transformer primary windings?

Best Regards:
Brian

[tjl]

NTC thermistor using AVX KYCERO 15 mm or 20mm diameter size disc type series with transformer primary winding , for about 2A 110V steady state current mode.
model no: NF15AA0509M 15mm or NF20AA0509M 20mm

5 ohms no current resistance, 0.17 / 0.15 ohms at max.current 4.9A /5.9A. power dissipation will be 4.1 /5.22 w. If feel too hot, may be change to 2.5 ohm type.

If not working at max.current rating, it is a non linear decrease curve of resistance, I don't know the actually thermo resistance value , so I parallel & series some pieces and check heat dissipation on the NTC disc and AC current meter, not too hot but must warmer than usual on NTC disc.

It is my experience for your reference only!

Cheers