[Amps] a 'circuits' question

[Steve Thompson]

On Wednesday 08 September 2004 14:44, Dan Sawyer wrote:
> This is a circuits question from a 575 curve tracer being used to match
> MOSFETs for an HF amp so it is a least close to topic. The problem is
> the difference in granularity between the 'steps' circuit and the 'zero'
> adjust circuit. The bias of the power MOSFETs is high enough so that
> only higher numbered steps get displayed the granularity per step is too
> high. The result is only 2 steps per trace get displayed.
>
> The circuit that produces this is a voltage adder with the steps input
> coming through a 90K Ohm resister and the bias coming through a 600K
> ohm. My question is can I substitute a smaller resister for the bias? I
> was thinking of 150K ohm. This would effectively increase the bias to
> the point where it was turning on the MOSFET and the steps drive could
> be reduced to a low enough value where multiple steps were observed.
>
> Is this a feasible modification??
I'm not familiar enough with the 575 to properly answer your question, 
although I imagine the proposed mod should work.

Let me suggest an alternative matching method that was used when I was 
involved in manufacturing rf fets:

measure the turn on point by tying the gate and drain together and applying 
volts between that and the source - go up to a few 100s mA. There's no big 
deal about the configuration - it's just quick easy way of doing it. Use the 
curve tracer, or just a PSU and voltmeter.

Use the curve tracer to measure the gm at a higher current using the step 
function.

If you have control of the circuit design, put in separate bias adjustment for 
each device, and do away with step 1.

Steve