Bias without bias tap

[rockdrigo]

Hi,

Is it possible to implement a MarkIII bias supply using a power transformer without a bias tap as shown in the figure below?

Is it better to use a voltage multiplier fed by the heater winding?

--Rodrigo.

[kheper]

Tap off the center tap. There are simpler schematics, than below.

[rockdrigo]

Thanks for your reply.

I think I need to reformulate my question:

Can I safely include the 150K resistor to form a voltage divider with the 10k pot and 18k resistor thus restricting the bias range between (roughly) -50 and -90 volts?

Cheers!

--Rodrigo.

[rockdrigo]

Wouldn't it be better to "move" the voltage divider towards the secondary in order to reduce the voltage ratings of the 100uF cap? (see below)

[WA4SWJ]

Rodrigo,

Why not do it like the eiclone fixed bias version? Check out the eiclone instructions. On page 18 Shannon's approach is shown. Works well!

Regards,

[Geek]

Looking at your first schematic, just drop in a 22K, 10W resistor in series with your diode on the AC side.

You have 150K + 4.7K in series. Get rid of that 150K and give it a try

[rockdrigo]

Hi,

The Eiclone concept looks good! Thanks for pointing that out!

I have two questions:

1) Why you need 3W for the 27k R?
2) If I use a 10k pot instead, does it need to be >3W too?

Thanks!

--Rodrigo.

[Shannon Parks]

The 27K, 3W resistor is kind of a magic value that I came upon after much testing. I just make the upper cap larger if I need a more negative C- (or smaller...). It is stiff enough to have keep the C- regulating but small enough to not be too much of a load on a tap. You could half wave rectify up to 200VAC on it (200VAC squared divided by 27K equals 1.5W). You want some margin on the heat dissipation because you never want it to drift or fail. Just use the exact circuit from the Ike. I used it on my latest 300B amp and it works great for SE, too, though I doubled the filter caps.

Shannon

[rockdrigo]

Hi,

The Eiclone bias concept will work only if the bias supply is loaded with R15+R19 in parallel with R16+R20.

I am building a Dynaco Mark III in which the bias supply delivers almost no current.

After simulating and testing a few circuits on the bench, I converged to the design below (almost the same as my initial post). The pot allows changing the bias voltage from about -50 to -80 volts.

Interestingly, the ratio of the cost of the 600v cap to that of the 60 volt tap in the power transformer is close to unity.

Can I multiply all the resistors (including the pot) by 10 in order to reduce the current?

Thanks for your patience.

--Rodrigo.

[Shannon Parks]

At steady state, that circuit is fine, I think. But at start-up, your C- will be slow as you have all that resistance in series before your filter cap. The time constant is 15 seconds - waaayyy too slow. But the lower leg scheme with the 10K pot and 18K resistor seems fine.

Shannon

[rockdrigo]

Hi Shannon,

Thanks for your reply.

I have simulated and built the supply and indeed takes about half a minute to reach SS (See below). I find such delay more than acceptable (My reference DIY amp takes > 1 min to reach SS). The process can evidently be accelerated by lowering the caps to, say 47u (at the cost of increased ripple).

Cheers!

--Rodrigo.

[Shannon Parks]

Looks good and not bad at all.

Shannon

[Impmon]

Hi,

Is it possible to implement a MarkIII bias supply using a power transformer without a bias tap as shown in the figure below?

Is it better to use a voltage multiplier fed by the heater winding?

--Rodrigo.

If you stay in Class *1, and the bias supply doesn't have much current demand placed on it, then that approach has been used a lot. The main point of departure is the inclusion of a high pass filter to attenuate the AC before the diode to take stress off the filter capacitors. The voltage regulation is horrible, but of no consequence if the only thing that draws current are the bias resistors/pots.