Matching IRFP240...

[viktor1986]

Hi

Is it good?

I want too match gm(transconductance)

I will use those mosfet's for my amplifier...

Take a look on amplifier schematic please

[Upupa Epops]

Viktor, Rod Elliot is correct, this is not good amp...Matching of vertical fets isn't easy, but it is about long talking... Try to make something with BJT's, for PA it is better solution... Belive to old tom cat...

[amplifierguru]

Hi Victor1986,

It's a VERY ordinary amplifier, lacking refinement, wasteful of output - but it would probably work.

Matching of the IRF MOSFETs for transconductance, why? The 0R22 source resistors will balance them up.

The SF output stage is wasting about 6V of output which amounts to about 40W lost at 8 ohms or 90W lost at 4 ohms.


Cheers,
Greg

[chicco_36]

Maybe you will need this

[viktor1986]

Originally posted by chicco_36
Maybe you will need this

This schematic is for matchnig Vgs!

[viktor1986]

http://www.diyaudio.com/forums/attac...amp=1132763406

Here is new schematic of my amplifier.I will use four pairs of IRFP

[Upupa Epops]

Viktor, I recoment to read something about constructions with Mosfets... Pretty nice papers have Erno Borbelly ( The Audio Amateur, Wireless World ), Robert Cordell ( The Journal of Audio Engineering Societty ) and Malcolm Hawksford ( dtto )...Don't try rescue America...

[ilimzn]

Originally posted by viktor1986
This schematic is for matchnig Vgs!

Really? I think you should look again, and check the definition of transconductance!

It is defined as delta(Id)/delta(Vgs). The MOSFET is connected so that it attempts to keep it's Vgs constant using it's own transconductance. In general, Vgs is about 3-4V for the given currents, which pretty much makes the voltage drop on the resistors constant, because it is small in comparison with Vdd (48V). This makes the resistor string behave roughly as a current source (The whole thing would work better if it was a real current source but it won't be that bad even like this). You are essentially measuring Vds at two currents, approx. 44/366=120mA and 44/330=133.33mA, which makes your delta Id = 13.5mA roughly, taking into account expected Vgs variation. The only thing left is to measure delta Vgs and divide the two! Also, for your transistors, it may be feasible to use a higher delta Id, use 100 and 200 mA as the two currents, possibly even more.

Secondly, I think you should take a look at the MOSFET data sheets. If you buy 100 each IRF 240 and 9240, you may find ONE badly matched pair, and that will be one really good 9240 and one really bad 240. Although regarding semiconductor geometry, 9240 and 240 are complementary, by definition P and N transconductance for a given geometry are different, by about 50-70% more for the N part, which has to do with carrier mobility. Transconductance of 9240 and 240 is approximately 3:5 as specified. If you want a better matched pair, you have to use 9240 and 340, deliberately downgrading the N part to get closer to the lagging P part.

FInally, your bias generator is a Vbe multiplier wihich will provide a too high temperature coefficient, meaning the bias current will drop too rapidly with temperature increase. In other words, it will be very difficult to keep the outputs optimally biassed. You can either degenerate the Vbe multiplier transistor (this can require a lot of tweaking), or use a Vgs multiplier instead (i.e. replace the BD140 with a suitable small MOSFET (IRF510, 610), and of course adjust the Rgs and Rgd resistors (they will be quite close in value to one another).