How do I match FETs?

[LektroiD]

I have 20x 2SK30 A-GR FETs, I need to find the 6 closest matches from them for a Roland 172 phaser I'm building. Anyone have a schematic or PCB/Stripboard/Breadboard layout for a FET matcher?

Thanks in advance...

[analog_backlash]

I haven't tried it myself yet, but Geofex have an article and simple circuit for doing this:

http://www.geofex.com/article_folders/fetmatch/fetmatch.htm

If you try it, let me know if it works because there are some phaser circuits that I'd like to try out.

Gary

[LektroiD]

[diablojoy]

Vmid = half the supply voltage in the example shown its 4.5v from the 9V supply
Vgs = gate to source voltage the value you wish to match between devices

[analog_backlash]

PLEASE DON'T TAKE THIS THE WRONG WAY

Looking back over your previous posts, I was a bit surprised that with your building history, that you didn't understand the terms Vmid and Vgs. I then thought about it a bit and realised that I was in the same situation when I first started DIY music electronics. I would build circuits from a schematic (or even better, from a kit) and have no idea whatsoever of how they worked.

At some stage, I started to try to understand electronics and I'm still very much low on the learning curve. There are plenty of posts on this web site which go way over my head still. What I would say though, is that the more you learn about electronics, the easier it gets to fault find when a circuit doesn't work and also allows you to improve/modify other people's circuits.

Going back to the testing FETs question, I'm having vague thoughts about how to produce an easy to use FET matcher. I suppose that it would be possible to duplicate the earlier circuit and run both the outputs into a comparator. The output of the comparator would then depend on the relative magnitudes of Vgs(FET 1) and Vgs(FET 2) and with a bit more electronics there could be LED indication when the two were very close (using a window comparator). However, the whole thing would have to be set up so that the outputs of the the 2 initial FET testing circuits give (almost) exactly the same results for a given FET, or else the result would be meaningless.

Ray Wilson has done this kind of thing for BJTs, but I've never seen a FET equivalent:

Link removed, as it didn't work for some reason, but it's on the MFOS site under Synth DIY 101/Transistor Matching 101.

This will give me something else to think about...

Gary

[frequencycentral]