Constant current DC heater supply circuit design.

[FranciscoPerez]

Hi guys,
I've been reading about constant current heater supplies lately.
Would this work as a heater supply for 5 12ax7 heaters wired in series? Are my calculations correct?


Sorry about the quick handmade schematic
Thanks!!!

[Paul1453]

Your circuit is more advanced than I can decipher quickly.   

I haven't tried anything constant current for DC heaters.

I just tried using a DC PS and step up convertor on an amp.
It worked very well for my simple circuit.
The DC heaters were definitely quieter than any AC heaters I've used.
I used this website to calculate the dropping resistor for my heaters.

http://www.gtsparkplugs.com/Dropping_Resistor_Calc.html

Add up the heater current from the data sheets.
Input your starting DC voltage, and ending voltage 6.3V, and heater current in Amps.
It will calculate the resistance you need, and the minimum wattage of the resistor.
This seems to work well for me, maybe you can use it to check your work?   

[PRR]

> constant current heater supplies

Why?

If you want to avoid the cold-start surge, using a heater as your sense resistor won't work. Cold-start surge will be as big as ever. And I am not sure it is certain to settle at 150mA, it may rise without limit.

The TL783 (I assume "738" is a mis-write) only needs 1.25V to sense. At 60V this is "nothing". An 8.33 Ohm resistor will limit to 150mA, cold or hot. Use Fig.17 in the TI datasheet.

So we need 5*12.6V= 63V on heaters, 1.25V for sensing, and (Fig.10) 7 or 8 volts dropped in the '783 at 150mA. You need raw 72V DC. Your sketch shows 75V DC; this only leaves 4% for transformer specs, utility company wobble, etc. If the raw DC is "barely enough", the dips of the AC ripple will pass-through the regulator and throw audio buzz around the heaters, even though it looks "OK" on the DC meter.

Stepping back-- I don't know any compelling reason to feed heaters constant current. Small tube heaters "do not" burn out. Not counting bad-design TV sets, I've only seen it a couple times in decades of tube abuse.

[FranciscoPerez]

Wow thank you guys!!


Yes, it's the tl783, I miswrote the name :)


The reason to use a constant current supply was to fight against heater warmup mismatch times, for I thought it might be something to be concerned about.
Used a heater as my sense resistor to replace that 8.33R resistor, which might be hard to find, and to avoid using several resistors in parallel. I didn't think about cold-start surge as a problem, though...


If I use the 8.33R resistor instead, will there be some sort of soft start to avoid that cold-start surge?
Thanks for pointing out about the need of a higher secondary voltage transformer!

[Paul1453]

The cold start surge reared it's ugly head during my recent DC PS build.

I started with a 12VDC supply with 2.5A capacity.
To power my EF86 and 6L6 heaters requires 1.1A of current.
The calculator told me to use a 5 ohm and like 6.5W resistor.
I'm not sure what the start up surge was, but the PS handled it no problem.

Then I tried to use one of my 19.5VDC Dell laptop bricks with much higher current capacity.
To get the proper voltage for my heaters with this requires a 12 ohm 14.5 W minimum resistor.
I paralleled a couple of resistors to get 12 ohms and about 30 W.
Every Dell brick I tried, even one with over 10 A capacity, all balked at the initial heater surge current.
After a few on/off cycles, that initial heater current surge calmed down enough for the Dell bricks to remain on.

I've learned that dumping 15W of power in a dropping resistor develops a lot of unwanted heat.
Even the 5 ohm 8W resistor gets very hot.  It needs it's own space.
Yo!  Don't crowd me Bro!   
For my little DC/step up convertor PS, 12VDC seems to be my best choice.   

[FranciscoPerez]

Thanks for sharing your experience Paul!

[PRR]

> avoid using several resistors in parallel.

Four 33r resistors is 8.xx Ohms exactly for any practical purpose. Compute the power: 0.05 Watts. Four 1/4W resistors is very cheap.

> If I use the 8.33R resistor instead, will there be some sort of soft start to avoid that cold-start surge?

With a fixed (not cold/hot variable) resistor it is then a Constant Current supply. It will not deliver more than 150mA ever. OTOH the cold-start current of a 150mA heater is likely to be 500mA.

> heater warmup mismatch times

Why?

All common tubes since 1959 have "Controlled warm-up", exactly because they began to use them in series strings. A 37V 20 Watt HO tube and a 4V 2W tuner tube in series, the size differences could over-volt one or the other at cold-start; the controlled warmup ensured safe series operation.

This is more true when all your tubes are "the same", 12A_7 or similar (they all use the same heater/cathode part).

And really, 75V and 80 Ohms of added resistance will give about the same operation withOUT any silly Silicon. (Unless you need stable DC balance and "need" heater regulation to reduce DC drift.)

[trobbins]

The secondary CT is grounded.  That means the heater voltage is partly floating and partly being jolted around compared to 0V.  All that noise on the heater may well be worse than AC powering the heaters.

Just ground the heater DC itself.  Your amp may benefit from some heater elevation for certain stages to reduce heater-cathode voltage, which may mean grounding or mixing the heater string up a bit.  Preferably keep ground at least 10-15V away from any cathode voltage, to maximise heater cathode parasitic resistance.

[FranciscoPerez]

Understood!! 


Thank you both.


Cheers,
Fran