Separate winding for bias supply?

[demonstratedspatulas]

Which is a better practice, using a separate winding for the bias supply, or deriving it from the B+ winding, and why?

[pompeiisneaks]

I think both are completely valid ways to accomplish it. The main idea probably is cost. It's easier to use a separate winding, but if you've got a transformer already that doesn't have one you can just build the type you're referring to.

~Phil

[R.G.]

What Phil said.

In addition, for "better" to mean anything except advertising, you have to have some kind of yardstick to measure how many "betters" you're getting, some kind of metric for which direction is "better".

I've always thought that better for bias supplies implied things like more stable, lower noise/ripple, and more reliable. It might even include factors like timing: comes up to voltage BEFORE the main B+ gets big enough for runaway in the power tubes. Marshall famously got this wrong by putting the bias on the switched side of B+, so that switching out of standby let the power tubes run without bias until it came back up.

You can do any of these with dedicated windings or tapped windings, or even a capacitor from the main B+ winding and no bias tap/winding at all.

[Phil_S]

Adding a distinct and separate bias winding, or any winding, for that matter, will increase cost. As RG notes, I'm not sure what you get for that in terms of value. I suspect, more often than not, the separate bias supply is a tap on the HT secondary, and not a separate winding. If the latter, then it is 6 of one, or half dozen of another with respect to the difference. The actual current load of the bias supply is so low that I can't imagine it will make much difference at all how or from where you supply it -- using a voltage divider from the HT or a tap on the HT winding.

[demonstratedspatulas]

Thank you! What about using a voltage divider from the B+ supply? Correct me if I'm wrong, but don't some amps do it that way?

[wpaulvogel]

Many amps tap off one leg of the secondary winding. Fender amps usually have a bias tap from the secondary. It’s not a separate winding. Some have a separate winding altogether. It makes no difference. I wouldn’t buy a power transformer just because it has a bias winding. I buy or spec them for VA rating and heater current.

[Phil_S]

Thank you! What about using a voltage divider from the B+ supply? Correct me if I'm wrong, but don't some amps do it that way?

Generally, this works as well as any other way. Check Marshall schematics for typical circuit configuration. I've attached a JTM45 as reference. You can replace the 15K resistor with a B25K pot to let you dial in the bias voltage. If you are making the pot permanent, get a locking type pot rather than an ordinary one. You don't want that setting to change randomly. This is probably the simplest way to derive bias voltage and it will work with just about any transformer. Actual resistor values will depend on the transformer and the target bias voltage. The 180K resistor sets the basic supply voltage and you must raise or lower that to get into the appropriate range.

[pdf64]

For good bias shift / blocking distortion mitigation, the bias feed to the grid leak resistors should be reasonably low impedance with the ability to sink / source some small amount of current. What’s not wanted is for the rectified overdriven signal voltage to charge up the bias supply caps (thereby temporarily changing the effective bias voltage). So the resistance to 0V at the bias feed should be <<lower than the grid leak resistance value.
It’s easier to achieve that bias supply design constraint with an ac source that isn’t a much higher voltage than is required for the bias Vdc. So a 50vac winding or tap can be better than a feed from a >300Vac winding.

[Kagliostro]

Only a consideration

if bias voltage fails during operation, the Power Tubes are easily damaged

if the bias voltage is feed via a separate winding that has a problem will result in Power Tubes with B+ and no bias = damage

if the bias voltage is feed via an intake on the B+ winding and there is a problem, this will result in no bias but also in no B+ and, may be, your power tubes didn't die

But I'll not consider this a big advantage of one way to do the thing or the other, the scenario isn't so common

Franco