LME49830 with (HEX)Fets

[Kay]

now we find the datasheet for the LME49830 on
www.national.com
The Eval-board use IRFP240/9240

It is a good idea to give the LME a own power supply,
maybe ±50V, and use ±40V for the HexFets-powerstage?

[SpittinLLama]

You will get more output power from the FETs this way but it might not be worth the effort depending on what you actually want to accomplish. You might just want to increase the supply voltage to get what you want. The advantage is more output power from the FETs and higher peak efficiency since there is less wasted voltage.

-SL

[Kay]

Thanks

I must explain, what I like to do.

I will use one Transformer and make a voltage-double
with stabilisation for the LME49830.

If I like to have a very good amplifier, it's maybe the
best way to use a stabilized supply, but it must be enough
to do this for the voltage-amplifier (LME49830).

sample circuit

[paulb]

Take a lookat this if you haven't already seen it:
http://www.passdiy.com/pdf/A75p1.pdf
and especially this:
http://www.passdiy.com/pdf/A75p2.pdf

[Kay]

my sample circuit

[Kay]

especially the Pass_A75p2.pdf is a must read,
thank you very much


Next questions,
on the National-Eval-Board the RQ ist located near by
the hexfets.

RQ is a resistor-load for the LME, not a termination,
so I like to place it near the LME.

I'm wrong?

[craigg4c]

If you're going to the trouble of powering your output devices differently from the driver chip, I'd suggest a higher voltage on the MOSFETs, which have a substantially greater voltage drop between supply and max output than bipolars. Why would you want to send less voltage to the outputs? Take a look at the Hafler designs.

[ionomolo]

Originally posted by craigg4c
If you're going to the trouble of powering your output devices differently from the driver chip, I'd suggest a higher voltage on the MOSFETs, which have a substantially greater voltage drop between supply and max output than bipolars. Why would you want to send less voltage to the outputs? Take a look at the Hafler designs.

Mosfets need to have the gate driven over theyr drain to fully conduct, so if they are powered from higher rails than the driver stage you will always be wasting power there.

The main advantage of powering the driver stage with a different supply is that it can have a much cleaner power and this improves the result.

[AndrewT]

most FET amplifiers perform better if the output bias is set high.
This dissipates a lot of heat.
Using the highest acceptable FET supply voltage to meet your target power, and output voltage requirements gets best efficiency and best economy from your PSU and heatsink and FETs.
Using a slightly higher and clean voltage PSU for the voltage amp stage is perfectly acceptable and does not add much in the way of volume nor cost to the project but does allow improved performance from the overall package.

Extra LV windings on the mains toroid is the cheap and easy way to get your increased voltage. Voltage doublers have their own problems, particularly if the driver stage consumes high continuous or transient current.

[ionomolo]

I'm quite surprised that national is selling a FET driver, my experience with the LME49810 shows that it's perfectly capable of driving HEXFETS and it sounds very, very good with them. I prefer the IRFP's over the toshiba's more expensive pair (2SK1530 & 2SJ201).

Kay (or anyone else), Have you tried to build an amp with the LME49830 and one with the LME49810 and noticed any difference?