heater current used for bias

[Andy Le Blanc]

Every now and again, you see a schematic or get inside and old amp that uses
the heaters of the preamp tubes as a bias resistor for the power tubes.
Its a tempting notion, and its been mentioned on the forum.
I decided to play with it a bit this morning and present some observations.

So..... I replaced the cathode bias resistor in an amp with the heaters of two
12ax7, in series and wired each for 12v. I bypassed this with a 100v/100uf cap and
placed a 1 ohm in series to ground to measure bias current.

The amp uses generic 50w fender replacement tranny's and set up with a pair
of JJ EL34 which with this arrangment yielded.......

P.... 392 V
K.... 25.7 V (for the pair)

zero signal plate curent ..... .15 A....

plate voltage measured between the plate and cathode at 370V

leaveing the EL34 idle at around 28W

Amp plays fine, but seems a bit brite, Ill fiddle a bit with the bypass cap later.

The heaters look to be running a tad hot, 25.7v, thats around 12.85v each
where all the published values say 12.6v, now this is where there may be
something of intrest. The measured bias current of .15A is exactly the same
as the current requirement of a 12ax7 with its heaters wired in series.
the hot fillaments acting as a current limiter, or in this application as a
constant current source. still have to gig the amp and see how well it fairs.

[greiswig]

Very interesting! One of those "smacks self in forehead for not thinking of it himself" moments for me. Thank you for posting!

[Andy Le Blanc]

its working very well at the moment, and is a very simple amp.
one gain stage and a inverter. Ill see how well holds up.
it looks like it might eat power tubes, the resistance of the heaters is very low
until they heat up, but its a gradual turn on, with a tube rectifier so maybe not.
But it got me to thinking that cycleing with a standby switch might be an issue.
but it cuts the job of wireing heaters to nothing, and its DC to boot.
dont have to worry about heater to cathode issues in a cathodyne inverter.
And it eliminates the cathode resister and cap. For any simple 50W amp thats got
only a couple tubes in the pre it looks very appealing. Even makes
12sl7 and 12sn7's a good option. see how it does in jam , then Ill know.

[FUCHSAUDIO]

That's an old audiophile thing. Scott and Fisher (maybe) did that. Cute trick, saves on the cost of a cathode resistor anyway. I doubt 12AX7 filament current varies that much that you couldn't count on a reasonably consistent bias. It's also (FREE) very pure DC, so with a cap on it, the preamp tubes should be really quiet. One of the cooler ideas for a "least moving parts" amp design....

[Wayne]

Interesting. Surprised Leo didn't exploit this as it lessens the parts count.

Andy, does everything stay kosher when the power tubes start drawing grid current?

W

[FUCHSAUDIO]

I distinctly remember a Scott integrated amp that had this in it. I don't know that heater current changes all that much when a power tube is drawing grid current (never checked it). Maybe audiophile designers (non guitar amp guys) don't think that happens so much in the living room.....

Worth playing with imho. Might have been a cute idea in the Champ or a similar simple amp...can't speak for Leo tho....

[Andy Le Blanc]

I think I saw it first in an old kay, or maybe a supro.
Ive also seen a transistor reverb recovery in an alamo powerd by the cathode bias.
I havent noticed any change in hum or blow, with or with out a bypass.
the tone was actually better without. I found that suprising, we'll see if it becomes an issue.

[Andy Le Blanc]

The resistance of the heaters before they warm up makes a bias of around
17V or 18V, so its not absent, and then ramps up to the bias V.
So far it looks good. My only worry is standby cycle. I'd rather it all warm up
gradually once. but I think also, like a light bulb is used to limit current, the
heaters might limit and dissipate the current in-rush thru the device.
with current flow thru the tube going from the cathode to the plate.
Ill know if I start going thru power tubes, but so far it looks good.

[Wayne]

Been thinking a bit more about this - can you tell I've got a pile of paperwork on my desk I'm putting off?

Even if you wind up putting a small series R in the circuit to get the power tube's dissipation down to more reasonable levels, all the other benefits remain (free DC heaters!!!).

You could also add a 3rd preamp tube, wired for 6V, put a shunt R around the 12V tubes, and get an extra 6V of bias - would probably necessitate juggling plate voltage to get the dissipation you want.

Yep. This seems like a good deal.

As far as the standby cycling issue, I think I read somewhere about the consequences of mistakenly putting 12V on a 6V heater - if memory serves me, the "light bulb" effect did limit current to a safe level and the tube survived. Not sure if you'd want to do it over & over again, and certainly wouldn't want to try it with anything NO$

W

[j-po]

The power tubes will draw considerably more current when driven into saturation. Poor filaments...

[j-po]

You could also add a 3rd preamp tube, wired for 6V, put a shunt R around the 12V tubes, and get an extra 6V of bias - would probably necessitate juggling plate voltage to get the dissipation you want.

Yep. This seems like a good deal.

That 6V preamp tube will need 300mA current. Bit high for bias don't you think?

[Andy Le Blanc]

lets see, the way it is now the bias V at the cathode is around 25 V (E) and the
heaters are rated at 0.15A and are in series, so thats 0.075 A (I)...
so R = E/I... 25/0.075 = 333 ohms . which is why it works as a bias resistor.

now I remember that current flow in a series circuit will be the same for all
resistances so that mixing the current rating will cause either the 6v heater to
only light dimly or not at all, or for the heaters with a lower current rating than
the current available , to excede their rateings. The PT is a 022798 and only
is rated for .2 A at the primary. inserting an extra heater rated at .3A
might not light that heater at all. but if the current output of the PT is significantly
higher you could chain .3A heaters. it is a balance between haveing enough
current to light the heater, then useing the hot resistance to provide bias.
after that I imagine that paralleling strings could be done to gain some current handling.

[Wayne]

You could also add a 3rd preamp tube, wired for 6V, put a shunt R around the 12V tubes, and get an extra 6V of bias - would probably necessitate juggling plate voltage to get the dissipation you want.

Yep. This seems like a good deal.

That 6V preamp tube will need 300mA current. Bit high for bias don't you think?

OOOPS! Seems like I've done it again! (Stepped in it, that is)

In order for my idea to work, the shunt resistor would have to shunt the two 12V heaters AND the power tube. That would involve drawing quite a bit of current directly from the B+. Would still work, but quite wasteful of current/wattage.

It's a shame that this idea couldn't easily be made adjustable to accommodate the fact that 25V may not be the ideal bias point in every situation.

Thanks for letting me think out loud, guys...

W

[Andy Le Blanc]

I think youve found the real issue. Cathode bias makes a voltage with a good
deal of current, but what is it good for? With a 200ma PT,and two .15A 12v heaters,
and pair of 6l6 the numbers coincied very well, but to push the idea into other
less idea circumstance, it losses its simple attraction. I think that it might be
good to try this as part as of a combined bias schem, where you can take advantage
of the available current for the goal of quiet and very simply wired heater, but
still want control over tha bias.

[ic-racer]

Every now and again, you see a schematic or get inside and old amp that uses
the heaters of the preamp tubes as a bias resistor for the power tubes.

I'm fixing up a HiFi amp like that right now.