Matching the MOSFETs

[naula]

I'm going to build Aussieamplifiers 200W@8ohms symmetrical amp. Construction notes says MOSFETs are recommended to match. I understand why but I would like to know how big can the difference be in current sharing without matching. In other words what will happen if I don't match the output devices.

I'm going to build stereo power amp. So I have to buy eight 2SK1530 and eight 2sj201. Is it enough if I match with those eight devices, taking four as similar as possible ? I don't have money to buy extra FETs

Link to the schematic:
http://www.aussieamplifiers.com/sym-sch.htm

 

[wuffwaff]

Naula,

it depends a bit on how (un)lucky you are If not very lucky (different batches) you can have differences of 20-50%. What you can do is build the thing and then switch fets between left and right channel (this is what I did with my crescendo) until you get a reasonable spread.

The other possibility is to measure Vgs for a given bias (see "matching" in numerous threads) for the 16 fets, match them as good as possible, put them in the circuit and then decide if you have to buy a few more.......

I think current differences of 20% (highest to lowest) are acceptable.

william

 

[Nelson Pass]

Insofar as Anthony is using low values (.22 ohm) of
Source resistance, matching becomes more critical to
insure that the devices share current in such a powerful
circuit.

Best to get some extras and match them as close as
possible; I would suggest .1 volt Vgs variation or smaller.

 

[jackinnj]

if you purchase a "sleeve"

of Toshiba, IRF or ST Micro devices, they usually match very closely. For Toshiba (lateral MOSFET's) there are 30 in a sleeve. For ST Micro there are 50.

 

[UrSv]

You might also check my posts here:

http://www.diyaudio.com/forums/showthread.php?s=&threadid=5713&postid=62126

/UrSv

 

[djdan]

Lateral MOS-FET need maching ?

 

[jackinnj]

The Vgs(t) can vary

by a few hundred millivolts