PFC with 1kW+

Dear friends,
I have been pondering of using a L4981A for a PFC pre-regulator to 405V and
a bridge configuration to get an isolated +75V 20A for a digital amp supply.
Are there any newer ICs which even might drive the primary FET/IGBT at
around 48kHz directly? Or are there any ideas to do these stages in a more
simple way?
Which power semiconductors would you recommend?
THX

 

Hi,
Try on the International Rectifier site, they have manny drivers and
usefull applications notes o)))

 

Ban wrote...

Have you read ST's relevant app notes, especially AN628 and AN1606?

Thanks,
- Win

whill_at_picovolt-dot-com

 

Shure I have all the apps printed out. The AN825 talkes about a downstream
PFC after a conventional transformer. Maybe this is really a good way, since
the transformer can be made a little smaller and the efficiency is higher,
OTH a 1.6kW toroid is expensive and heavy. :-(
I want to make the triangle generator current in the modulator of the amp
proportional to the supply voltage and have a rather small
capacitor(20mF/80V), so the voltage sags like an analog amplifier and the
"music power(1600W)" is a bit higher than the continuous "rms-power(1200W)".
I think this way the efficiency can be improved and since I want to put the
amp into a 2or3HU rack with integrated heatsinks, I might not need forced
cooling.

 

Most manufacturers have gone away from including a high current gate
driver in their power switchmode controllers for several reasons,
including these:

1) The gate driver components should be located VERY close to the
MOSFET, which is usually not a good place to locate the controller IC,
with its sensitive inputs.

2) Driving high currents on the same IC die that is used for feedback
control poses difficult design and IC process problems.

Gate drive for a PFC can be a very simple 2-transistor complementary
emitter follower circuit, with a component cost of less than $.50 and
a very small footprint, so why worry about it?

L4981A is a good chip, but you can also consider one of the PFC/PWM
combination chips that operate PFC in leading edge PWM and DCDC in
trailing edge or vice versa. These can reduce component count, cost,
and EMI. The following mfgrs offer the combo chips AFAIK:

Champion Micro (my favorite), Fairchild, TI/Unitrode, Linear
Technology

Some of the chips give you the option of operating the DCDC at twice
the PFC frequency.

Paul Mathews

 

Paul,
yes for the power I'm aiming at it will certainly be better to use dedicated
high- and low-side drivers. Just another isolated 12V needed. :-(
It will be difficult to make a reliable downstream switcher for the cost of
a primary transformer, because the volume of my application will not be that
big, maybe 20 to 50 units/annum only. I still have to find some funding and
marketing as well, so it will be only a few prototypes in the moment.
Now I'm using a conventional amp(6x150Wrms) with 60 pounds weight and an
annoying fan.
I didn't know the champion brand, certainly very interesting.
Thanks again for the input

 

You don't need another isolated 12V for gate drive: gate drive
transformers are small and inexpensive. You'll also find that your
actual power requirements for audio program material are a fraction of
the peak power. Even highly compressed rock and roll music has a
crest factor of 5 to 1, so you don't really need to rate your PFC and
DCDC stages for as much power as you're thinking. Consider using a 2
transistor forward converter for the DCDC, which works well with a
dual-secondary gate drive transformer. If you use current mode
feedback and a sense resistor in the lower transistor Source, note
that many app notes show the gate drive winding for this transistor
connected between the Gate and Common (i.e., the grounded end of the
sense resistor). This is incorrect, since it introduces negative
feedback to the gate circuit. The gate transformer should be
connected between Gate and Source.

If your audio amp has good power supply rejection, a better
alternative might be to use an unregulated push-pull inverter for your
DCDC, relying on the PFC stage for regulation and the DCDC transformer
turns ratio for the step-down function only.

Paul Mathews