Hot running regulated dc filament supply for aikido preamp

[2004ex]

Hi, I have helped a friend to build a 6sn7-gtb/6sn7-gtb aikido stereo preamp based on the tubecad.com 2-chnl octal board (yes, paid full price). The 6sn7's are pair-connected in series for a filament volatage of 12.6V. The dc filament supply is a regulated unit - 12V rated at 1.7A (power-one.com hb12-1.7-A, adjusted to 12.6V). From the 6sn7 data sheet, the filament requirement is 0.6A at 6.3V. Since they are pairwise in series, the total filament current at 12.6V for four 6sn7s would be 2x0.6A =1.2A amp, well within the spec of the power supply (1.7A). However, the filament supply ran very hot in this case, too hot to touch after about 30-40 minutes of switching the preamp on. I am wondering if other members had similar problems when usng regulated DC filament supply modules. Power-one is quite established I think and their spec should be reliable. What kind of current ratings are people using in a similar situation? TIA.
(PS: After a couple of trial and error wiring rearrangements the preamp is quiet now and sounds good - but then it turns into a hot ...)

[ErikdeBest]

Hi 2004ex

What is the voltage before regulation? The math to calculate power dissipation is rather simple: (V in - V out) * I out. You already told V out and I out, so the equation becomes (V in - 12,6) * 1,2 = dissipated power in regulator. You sure need a heatsink, but depending on the drop across regulator you will need a bigger one. Or you can try a 24VDC fan supplied from the same 12.6VDC and see if this solves the heat problem.

It can also be that the regulator is oscillating or something like that. Poor decoupling, etc. I can't be of much help here

Erik

[mario2]

Those power supplies DO get very hot, you should have selected one that doubles the rating you need. How much input voltage does it has?

Another thing often overlooked is that for filaments designed to run in parallel like the 6SN7, only the voltage is guaranteed (6.3V) but the current is only an indicative value. So the 0.6A could well be 0.7A or 0.65. So I would go ahead and measure the current, perhaps you are close to the PS max rating.

Although the one I have gets extremely hot (too hot to touch) when taken above 1A, it's still working fine after many hours.

Hope it helps

Mario

[ErikdeBest]

Very good point by Mario

I would, just to be sure, measure if there is actually 6,3V across each tube. The filament resistance from one may be higher than from the other, so you could get 6,6VDC on one tube and 6VDC on the other. It's not hard to correct, you just have to put a resistor across the filament which has the lowest voltage. The value of the resistor? Some measures and ohm's law will tell!

[Tom Bavis]

Perhaps a 24V dc fan could be used to "breathe" on the power supply... it would be pretty quiet on 12V input. They're dirt cheap surplus parts ... in fact I would send you one for the postage...

[2004ex]

Thank you for all the suggestions.

Originally posted by ErikdeBest

What is the voltage before regulation? The math to calculate power dissipation is rather simple: (V in - V out) * I out. You already told V out and I out, so the equation becomes (V in - 12,6) * 1,2 = dissipated power in regulator.

There are 2 coils in the secondary - 7.5V and 16.5 V AC. It has a single 2N3055 pass transistor and a LM723 chip. Full wave diode rectifier. So the voltage before regulation should be around 22V DC. The formula gives about 12W. Not a lot?

It can also be that the regulator is oscillating or something like that. Poor decoupling, etc. I can't be of much help here

Good point. I should investigate that.

Originally posted by mario2

Although the one I have gets extremely hot (too hot to touch) when taken above 1A, it's still working fine after many hours.

Yes, but still I would feel much better if it is not so hot. A little worried that it could become hazzardous if one forgets to shut it off before going to sleep.

Originally posted by Tom Bavis
Perhaps a 24V dc fan could be used to "breathe" on the power supply... it would be pretty quiet on 12V input. They're dirt cheap surplus parts ... in fact I would send you one for the postage...

I think that may be what I have to do next. I do have a couple of 24V fans lying around somewhere, so thank you for your offer and I appreciate that.

[ErikdeBest]

Hi 2004

I am not really an expert with heatsinking, i.e. I don't know how big a heatsink needs to be to sink 12W, but last week I was playing around with a regulated supply for filaments and justed wanted to know how hot a heatsink of about 8x5x5 cm would get when dissipating 5W. It was quite hot...so, 12W is a lot, actually.

I would try the fan solution first. If it doesn't work or makes too much noise you could also put a resistor from input to output of regulator. I recently saw a thread where this was discussed...the idea is that the resistor passes some of the current, but still allowing regulation through the regulator.

Dit you check if there is really 6.3VDC across each 6SN7?

Erik