bias tap - relay supply

[lowell]

Thought I'd start a new thread for this. Say you have to make a relay supply in one day and all you got laying around are some 12v relays and the amp on the operating table has a bias tap and you don't have any suitable transformers around.

Could you connect the base of a high powered pnp mosfet to the bias supply filter cap, drain to a negative supply off the main rectifier via 1 diode and filter cap, and source through a suitable valued Rs, in voltage follower arrangement, then a filter cap on the source, then voltage divider to make 12v supply? That would take care of the "current" issue that comes with your typical PT bias tap right? Or do transistors function differently with DC than AC? Or maybe this is absurd...

I know there's more specifics to worry about, but is this basically accurate?

[pdf64]

I've used 48V mini telecoms type relays on a bias supply. Even then, the (2) coils pulled down the bias supply a volt or 2, which I was a bit uncomfortable with, though it's never caused a problem.
The current draw of 12V relays will be much higher, and so the knock on effect to the bias voltage is likely to be greater also.
You could mitigate by having a seperate rectifier / reservoir cap for the bias supply and the relay supply, but need to know what the relay coil resistance is and the transformer winding current rating.

[Steve Conner]

I would be a bit worried about burning out the bias tap.

On the one hand you can assume it's made of the same gauge of wire as the rest of the HT winding, so it has the same nominal current rating.

But on the other hand, it is part of the HT winding, so every other half-cycle of the HT current goes through it, which is not half, but 0.7 on a RMS basis. So, most of the current rating of the wire is already used up.

Unless you use some sort of fancy switchmode regulator, the current drawn from the bias tap will be the relay current, times the number of relays that are on at once. Your regulator will have to dissipate that current, times the voltage difference, as heat. The bias tap will give 50-70V rectified, so that's a lot of wasted power to run 12V relays.

If you rectify the positive half-cycles for use with your circuit, then its current draw won't pull down the bias supply, because that runs off the negative ones. The amp's main HT draw doesn't pull down its bias, for that same reason.

I have used the bias winding to make +30V for powering op-amps with a 15V "virtual ground", but relays would make me uneasy. I prefer to triple the heater voltage using my patented tripler circuit that I've posted before.

[Gtr_tech]

Would the relay be for signal switching? If so, I'd use Jfets instead. That way all you need to do is drop some voltage off the bias supply and use a 3 terminal regulator. Pulling any current off the bias supply is not the greatest idea......

[lowell]

That's my question though guys. If I put a voltage follower in there, it has a high input imedance and a low output impedance. Just not sure it'll work in the DC conditions I described. My thoughts are that there will be minimal current drawn from the actual bias tap.

[defaced]

Voltage doubler off the heater winding. Relays have an operation tolerance you could drive a truck through, so any voltage sag will likely be within the operating range of the relay.