Jfet matching

[ch_scr]

I bought some MMBFJ108 from China where i wanted to match some for a project. My 34970A was begging for use. This is what happend.
I looked around online and found the nice "volt meter and voltage source" trick to V_GS(off) here: http://stompville.co.uk/?p=112
I realized i could measure R(DS) with the ohms measurement of the DMM. It's not a parameter field, but good enough for a start.
Normally the two measurements would interfere (say done with two dmm and a power supply at once), but that's what the 34970 is for

I rummaged in my pile of pcb's for a small one with a suitable footprint, kicked the old parts off, and put 3 nice solder blobs on the footprint I wanted to use. Later I filed the blobs flat on top to stop the parts from slipping away when i pressed on them to make good contact while measuring.


I wrote a Python script to do the measurement and log data. Measuring the 26 parts took about 20 minutes. This includes one dead jfet where i thought it didn't make contact and messed around. I also did a repeatability test where i put on one part, let it do the measurement, took it back off, rinse and repeat. The file is also attached and honestly it came out more stable than I expected.

[David Hess]

A simpler test is to place a chosen resistance between the gate and source which sets the operating current.  Then with voltage applied, either the voltage across the resistor or the current is measured.