Ode to Lateral MOSFETs

[suzyj]

I get ~10% Id variation between pairs using 220mOhm source resistance, with fets from the same tube. I'm not giving up on the source resistors.

 

[AndrewT]

I get a lot wider variation than that.
With matched Lateral mosFETs and using matched Source resistors, I can get upto +30% (highest ref lowest) in paralleled sets.
I would not attempt unmatched and can't imagine how bad Id would be without Source resistors.

 

[ivanlukic]

With one pair of output Latfets source resistors are probably not needed but for multiple pairs are needed. Although Ben Duncan claims that they are not needed even with multiple pairs. High power double die Latfets are in fact two paralleled Latfets in the same package without source resisitors.

 

[ivanlukic]

I recently did a redesign of a Hafler DH220 and got very good results. I biased each output device at 200mA.

Cheers,
Bob

Bob,
Was the redesign with Vertical MOSFETs?

 

[29285]

The matching of Hitachi lat-FETs was excellent. In the early eighties I built several mono blocks with 36! pairs of 2SK135/2SJ50 without any source resistors achieving 5kW of continuous sine wave power.

 

[Bob Cordell]

Bob,
Was the redesign with Vertical MOSFETs?

The redesign used the same lateral MOSFETs in the same locations as the original amplifier. The amplifier did not use source resistors and I did not add them.

Source resistors tend to be of limited value in making up for mismatches in lateral MOSFETs, partly due to the relatively small transconductance of the laterals as compared to the conductance of typical-value source resistors.

Cheers,
Bob

 

[MiiB]

The best laterals were the TO220 pieces, low input capacitance, really nice sounding devices for preamps and inputstages, but no more..

 

[phase]

What I mentioned before about the source resistors being different when using the newer parts, should have been "gate" resistors instead, sorry about the misinformation there...