Bias transformer

[Stankfut]

A few weeks ago kagliostro posted a question about VVR for a fixed bias amp. I've read through the info a few times, and still don't quite understand how bias tracking works. One thing that I did get out of the links that was posted was that a separate bias transformer is needed for bias tracking, and may be a good idea in general, if the PT doesn't have a dedicated bias tap. So....what transformer would be good for a bias supply? I'm thinking maybe something like a Hammond 164 series, maybe in the 24-28v range?


Thoughts?

[sluckey]

Someone recently posted about using a 120V:12V step-down transformer. He wired the 12v secondary to the filament string in a Sunn amp to provide a 60VAC source. I thought that was a great idea. RadioShack has several size 12V transformers. Even their smallest 300mA transformer would be overkill for a load of a bias supply.

[EL34]

Steve,
I stock a toroidal here that may work
So if you hook up 6.3vac to the secondary, you should get around 80+ vac off the primary
115 / 9 = 12.77 step up ratio
6.3 x 12.77 = 80.45
If I did my math correctly and that's not always the case


run the primaries and secondaries in parallel

They are on this [page
http://hoffmanamps.com/MyStore/catalog/parts1.htm

[PRR]

> a separate bias transformer is needed for bias tracking

Mis-written, mis-read, or mis-remembered.

VVR of 'fixed' bias amps can be done with the stock bias supply IF you do it right (and that aint easy). I don't *think* a separate transformer helps.

However there are times you don't have a bias winding, or it is way shy, and you don't like (or can't use) the wacky trick of kicking-down a small low voltage from the High Voltage winding.

Then it may be convenient to put a 120V:12V *backward*, with the "12V" side on your 6V heater winding. This gives 60V AC on the "120V" winding. Because the bias demand is VERY small, you can use the smallest transformer you can find (bigger than your thumb-- hyper small trannies use hyper-fine wire which may corrode sooner than generic-size windings).

Or 120V:9V, backward on 6V, is indeed about 84V AC.

Traditionally the bias winding for "usual audio power pentodes" is 50V AC. This is more than enough for any guitar amplifier; we always have to trim-down from the rectified ~~65V DC to our 30V-50V DC range. 60V means more trimming. 80V is even more padding-down, though that is a very cute part. And sealed against corrosion.

[kagliostro]

(and that aint easy)

Yes, sure, that may be a pain, we are still trying to have it done (using B+ as voltage source instead of a separated source)

---

We found info about the use of a separated bias PT at London Power, where they say

The problem with these stock bias supplies are twofold. First, they are usually "high impedance", which means they are derived from the plate winding through very high value resistances (100-220k) or through capacitors. A high-impedance bias supply cannot support proper bias-set networks nor will it support a bias regulator.

source: http://www.londonpower.com/bias-supply

If your amp has a capacitively coupled bias supply and you are thinking about adding Power Scaling - or merely wish to add a tracking bias regulator for other purposes - add an auxiliary transformer to generate the bias supply. This will provide a low impedance that can support multiple bias pots, etc., and have sufficiently high voltage overall to let the regulator function properly.

source: http://www.londonpower.com/raw-bias-supplies

I think that Doug's small toroidal transformer will be perfect for the purpose to have a low impedance Bias PS that is required for automatic bias tracking circuits, like Merlin's and KOC circuits

here the Merlin's circuit

K

p.s.: The cap in red was added by Mr. Merlin as to avoid auto oscillations of the circuit


EDIT: As per PRR signal the tube on the schematic has been redraw

[PRR]

> here the Merlin's circuit

I believe that drawing shows the G1 and G2 reversed.

[Stankfut]

Thanks guys! When I tackle this, I'll probably use Doug's transformer.

Mis-written, mis-read, or mis-remembered.

With me, it could be any of the above! 

[kagliostro]

This is the original draw with a pentode

K


p.s.: I go to redraw the tube correctly, thanks PRR

[PRR]

If anybody wonders what it does:

The basic VVR varies the plate (and G2) supply from as high as possible (maybe +400V) to as low as makes sense (zero, or some slightly higher voltage like +40V).

But if we reduce the G2 voltage, and keep the same G1 ("bias") voltage, the tube current is radically reduced, and the amplifier stops working right.

We need to vary the G1 voltage in about the same proportion as the G2 voltage.

For many common audio power tubes, the proportion is *about* 1/10. +400V supply, probably about -40V G1 bias; at +200V you want more like -20V G1 bias.

We want a "gain" of -0.1. The 0.1 is easy, but the - is not, unless we use an amplifier.

This amplifier has to output anything from -1V to -55V (so we can't use the cheap chips). For simplicity, we'd like to run both inputs near Ground, though not burn-out if it goes somewhat below or above zero.

It does not need strong output or rapid response.

The ZTX transistors are just an Op-Amp.

The resistors and pot give our 1/10 ratio (adjustable 1/60 to 1/6 to dial-in the full-power bias).

The opamp's negative supply is the original full bias supply, which is more than enough.

The opamp's output's positive supply is ground (via the ~~30K) because we never want G1 higher than zero.

This opamp uses the high positive voltage as supply for the inputs. It does not *have* to be that much, but we don't have any handy lower + voltage, and this scheme works fine. While we could use one 300K resistor, real-world resistors have Max Voltage ratings, sometimes as low as 200V. Cheaper to use three 1/4W parts than one 2W part just to get a 500V rating.