It's known how to match FETs due to generosity of The One and Only. By the luck I've got some SITs and would of liked have them matched within reason. I can't make or borrow full blown curve tracer but I can plot a curve or two per device. What should I look for?|
If you can plot three or more curves for each device in the area of your intended operating point, you can get an idea of the gain and distortion.
Gain can be calculated from the spacing between the curves along your load line. The wider the spacing, the higher the gain (larger Vds voltage swing for change in Vgs).
Second order harmonic distortion can be calculated from the differences in curve spacing. Amplitude differences between negative and positive Vds changes for constant negative/positive Vgs changes results in the distortion.
Before computers, calculations for triode vacuum tube amplifiers were performed in this manner and there is information on this on the net. For example, https://atrad-audio.co.nz/turneraudio/www.turneraudio.com.au/loadmatch-1-SE-triodes.html
Gain can be calculated from the spacing between the curves along your load line. The wider the spacing, the higher the gain (larger Vds voltage swing for change in Vgs).
Second order harmonic distortion can be calculated from the differences in curve spacing. Amplitude differences between negative and positive Vds changes for constant negative/positive Vgs changes results in the distortion.
Before computers, calculations for triode vacuum tube amplifiers were performed in this manner and there is information on this on the net. For example, https://atrad-audio.co.nz/turneraudio/www.turneraudio.com.au/loadmatch-1-SE-triodes.html
I automated the FrankenTracer with an Arduino computer so multiple curves can be plotted simultaneously. See https://www.diyaudio.com/community/threads/frankentracer-created.391816/
As you can see the Frankentracer have been used. The question is how to match devices using one or two curves. It there was a pulse mode like one in real measurement equipment you may skip mounting DUT on heatsink. Yet another cumbersome operation is curve(s) "extraction" from the scope using USB flash.
Unless you make an automated system with decent NI controlled devices or DIY "cargo cult style arduino replica" you are out of luck.
P.S. N. Pass matched FETs at one particular point voltage/current. I am curious if such a thing exists for SITs. If nothing works I will try PM him.
Unless you make an automated system with decent NI controlled devices or DIY "cargo cult style arduino replica" you are out of luck.
P.S. N. Pass matched FETs at one particular point voltage/current. I am curious if such a thing exists for SITs. If nothing works I will try PM him.
At minimum, I would use 3 curves for each device that cover the operating point and voltage swings and try to match that way.
What a surprise...
Thanks everybody.
P.S. What would be decent data acquisition board USB/Windows?
Thanks everybody.
P.S. What would be decent data acquisition board USB/Windows?
It depends on your needs. See https://www.diyaudio.com/community/...-by-vgs-and-curve-tracing.381338/post-6902977
That said, pypsucurvetrace is a nice toy:
https://pypsucurvetrace.readthedocs.io
Thanks Matthias! There's no escape from a full data aquisition. A grand or two for PSU perhaps too mach but the approach is mor or less correct. I would use resistive load and match large signal gain first and see if that suffice. Instead of mathching the whole surface. The latter may become even more complicated with complex (RLC) load (pun intended) of high order.
It may be just as quick to build a choke loaded common source amplifier test circuit and perform gain and distortion measurements on each device. All it takes is a lab power supply, a decent choke like an Hammond, some resistors, capacitors, and a heatsink. A multimeter, soundcard, and REW software would be needed for measurements.
Comparing gain and distortion at a fixed operating point at a couple of output levels would show which devices perform with similar results.
Comparing gain and distortion at a fixed operating point at a couple of output levels would show which devices perform with similar results.
You might want to stick with the heatsink and measure at operating temperatures.
Interested as I have a similar problem.
Interested as I have a similar problem.