How many parallel fets could BOSOZ drive?

[Michael Rothacher]

I need to juice up my power follower and I'd like to run 3 or 4 MOSFETS in parallel. Wishful thinking? Also, how would an amateur determine this?

[Mad_K]

The BZLS doesn't care how many fet's you paralell. It's the Gate capacitance that gets porportionally higher when you put devices in paralell. This affects the HF response of the circuit (ie lower bandwidth). Do you have a schematic?

[Nelson Pass]

Originally posted by mrothacher
I need to juice up my power follower and I'd like to run 3 or 4 MOSFETS in parallel. Wishful thinking? Also, how would an amateur determine this?

No problem. If you really want to know, the best way is to hook
them up and observe the high frequency rolloff as you increase
the source impedance. In your case, I would expect them to
make to 100 KHz, depending on the devices.

[Michael Rothacher]

Thank you! I think I undersatnd now. I'll give it a try and let you know how it goes. I'll post a schematic too.

[Michael Rothacher]

Here's a schematic of my power follower with the BZLS driving it. You might recognize the follower as a heavy-duty "Zirclotron" (I just can't give up on this)

I haven't yet figured out exactly how to do the VGS multiplier for bias. This might make a good future tutorial for Mr. Pass. For now I'm sticking with the voltage divider bias.

Thanks again.

Mike

[Joe Berry]

Originally posted by mrothacher
I haven't yet figured out exactly how to do the VGS multiplier for bias.

Hi Mike,

First, I think you want R27-R28 connected to the left side of R31, and likewise for R29-R30 and R32.

Regarding a Vgs multiplier, I think the simplest approach with your circuit is to power it from V3 with the source tied to ground. I would set is up as a fixed Vgs doubler (using 2 10k resistors, say) with a pot in shunt with the doubler to tap off the bias voltage neeed. Then add two resistors in the 100k-1M range, one from the wiper of the pot to the right side of each coupling capacitor.

You could also add a balance pot, but I donj't think you'll need it if you use matched output devices. With this arrangement, you should also be able to drop the value of the 10uF coupling caps by 1-2 orders of magnitude. Good luck.

[Michael Rothacher]

Joe:

Thanks! You're a constant source of enlightenment and encouragement. I looked up what little info I could find on Vbe multipliers and came up with this contraption. Tell me what you think of this bias scheme:


As you suggested I could also power it from the input power supply. Do you think that would eliminate the need for a highly regulated supply for the output?

R16 is for fine tuning the balance. In this case (using MOSFETS in the multiplier) would you attach the IRF610's to the same heatsink as the outputs?

I thought it would be cool if I could figure out a little bridge circuit with a meter to adjust the balance. Still working on that.

Mike

[Michael Rothacher]

Well..I guess my nulling circuit idea is a little silly since all I have to do is measure the DC across the speaker terminals and adjust R16 for zero. Am I right on this?

Also, I had to make a couple of changes to the bias which I'll re-draw and post.