How much VAS current to drive a pair of lateral FETs?

Current is demanded by the FET gates, so the Cob of the VAS transistor is not really of any relevance to this part of the design.

I found IRFP240/9240 vertical MOSFETs pull about 3mA at 20kHz. The lateral parts have about half the capacitance, so will take about half the current. I use a rule that the source should have approx 10 times the current it is required to deliver, so in that case I would run the VAS at about 15mA. This is probably a fair bit higher than what a lot of published schematics use.
 

jcx

Member
2003-02-17 7:38 pm
..
the IRFP240 VMOSFET have way more nonlinear input C - especially near saturation so high VAS current is a good ieda

Lateral MOSFET are better behaved so aproximations based on ( output current slew demand / gm ) to get gate V slew requirement gives gate charging current - much smaller than your estimate only ~ 0.5mA required for "audio" so 5 mA VAS bias likely fine (Crss in Laterals is small enough to be ignored)
 

jcx

Member
2003-02-17 7:38 pm
..
"back of the envelope" estimate needs adjustment to exact circuit

only makes a ~25% difference in Class A - in AB you might need 2x the current unless you use a "non-switching" bias which wouldn't discharge the off Q gate C

another topological difference would be if push-pull VAS was used
 
the IRFP240 VMOSFET have way more nonlinear input C - especially near saturation so high VAS current is a good ieda

With source follower outputs the input C becomes less significant because the source pretty much follows the gate voltage.

I use 15mA VAS current with 3 pairs of IRFP240/9240 and get a good bandwidth.
 

Zero Cool

Member
Paid Member
2004-09-20 6:10 am
MN
You know this is a really interesting question and thank you to richie00boy for his explanation. I have had my own theory that Mosfets need a large VAS driver stage. All of the mosfet output stage amps I like have very large current capacity driver stages. and far too many of the early designs had wimpy driver stages and i think this had a lot to do with the early descriptions of "mosfet mist" and other anomalies . this has always just been a guess on my part. But I do find it interesting that all of the amps i like have overbuilt driver stages. either by design or error, there is something about a large driver stage ahead of mosfets that I like.
 
No real question, no real help. You haven't specified anything--what's the suppy voltage, power level, power bandwidth, exact VAS topology? Where's the schematic of your amp?

My short answer is: It's unclever to drive any power device directly from the VAS. Distortion will be unnecessarily high and there will very likely be some sort of large-signal limitation. Add an emitter follower in between.

Samuel
 

Henry8

Member
2010-07-22 1:28 pm
No real question, no real help. You haven't specified anything--what's the suppy voltage, power level, power bandwidth, exact VAS topology?

Samuel

Amp topology is ltp/vas/output
psu -/+ 50V
100W/8R
VAS is a bjt in CE mode fed by a CCS


It's unclever to drive any power device directly from the VAS. Distortion will be unnecessarily high and there will very likely be some sort of large-signal limitation. Add an emitter follower in between.

Samuel


With this statement your disrespecting many amp designs.
 
If there s enough NFB ratio , driving 3 pairs of the discussed
devices directly from the VAS is fairly possible and works quite well.
As low as a 3 to 5 mA current is generally enough.

Although adding a pair of CE coupled drivers reduce THD in a
significant ratio , it also add another pole in the transfer function,
requiring a heavier compensation, but all in all , it worth it.