matching FETs

Danny G · September 21, 2004, 10:16:51 AM

I searched around here and didn't really find anything. I searched GEO and got the article with the FET matching schem.

I need to match 2n5952's for a Phase 45 and Phase 90. Is there an easy way? My DMM has a tranny tester for hfe. I'm not knowledgable enough to turn the FET matching schem into a useable layout, and much of the GEO article was Greek to me.

Suggestions?

Paul Marossy · September 21, 2004, 10:27:10 AM

There really isn't a better way to test FETs than the GEO method. The hfe thing on your DMM is only for regular transistors. I built the tester, and it works great. Have you built anything on perfboard yet? This is about as simple of a circuit as you can get...

cd · September 21, 2004, 10:30:19 AM

You can build a Phase45/90 but can't read a schematic?

Paul Marossy · September 21, 2004, 10:41:53 AM

cd's reply is much more to the point than mine...

Danny G · September 21, 2004, 10:43:23 AM

Turning a schematic into a layout is something I haven't mastered.

Anybody have a parts layout for that FET tester? What kind of op-amp do i use?

Paul Marossy · September 21, 2004, 11:12:09 AM

I sketched up a little layout on a piece of paper, but I have since thrown it out. If you do your layout like the schematic, it might help you to do a layout. I believe that the opamp can be any opamp. I think I used a 741 in mine.

Danny G · September 21, 2004, 11:14:03 AM

I'll give it a shot...

Danny G · September 21, 2004, 12:13:49 PM

Okay. I think I have it worked out, but that Op-Amp gave me a hard time. hope this thing works.

If anyone has a layout for it, PLEASE post it.

Gringo · September 21, 2004, 12:30:43 PM

Check at my site :

http://gr1ng0.tripod.com/Downloads/dl.html

http://gr1ng0.tripod.com/Downloads/fm.zip

Hope it helps.

Paul Marossy · September 21, 2004, 01:01:26 PM

Get a data sheet for the opamp you want to use so that you can determine the pins that you need to connect to. Most data sheets are available on-line and can be found using a search engine...

Michael Allen · September 21, 2004, 01:42:36 PM

You know, I think that Rob Strand's FET matcher is much easier to use. So simple and quick

Do a search for Rob Strand + FET Matcher. It might even be in the Rob Strand link off of Aron's homepage.

Uses just one transistor instead of an opamp....

cd · September 21, 2004, 02:11:46 PM

You'll do yourself better in the long run if you learn the standard op-amp pinout, how to get power to them, connecting up parts, etc. It's not that hard, get yourself a breadboard and connect up all the parts as in the schematic.

Danny G · September 22, 2004, 09:11:32 AM

Cool, thanks Gringo. I was close, but still way off... I built it last nite and it woiks!

Muchas gracias!

petemoore · September 22, 2004, 11:55:49 AM

Anyone find the Rob Strand Fet Matcher schematic?
I've rebuilt the GEO one a couple times and can't seem to make heads or tails of anything as far as the readings...they all seem the same...even when I tried like 8 J201's...I 'know they all can't be that close...so Far I've got a phaze 90 'working' but it has a tick...I used the swap till you drop method of testing fets in live phaze 90 circuit.

vdm · September 22, 2004, 08:28:59 PM

i found the easiest way to match fets in a P45 is to just throw one in at a time and you'll get some phase shift when it is biased correctly. just find a couple that bias at the same spot in the trimpots rotation and your set!

trent

edit: of course this requires using sockets.... which i think is must anyway :wink:

Paul Marossy · September 23, 2004, 10:22:22 AM

When I built my Phase 45, I just popped a couple of J201s in there, no matching at all, and it worked great. :shock:
Maybe I was just lucky?

DDD · September 25, 2004, 08:50:53 AM

Hi men,
Could somebody please advise me a widely-spread matched FET pairs with low drain current at the zero gate-source voltage?
Many grateful thanks in advance.

fretzburner · March 29, 2010, 01:16:58 AM

How to use this fet matcher? I tried to read 3 j201 and the reading is not stable it varies in voltage value around 4.1v to 5v.I dont know if i did it right.The DMM is between OPamp out and positive input.

solderman · March 29, 2010, 04:37:23 AM

A bit of topic but..
I am a curios cat. So when I built my MXR phase 90 clone I first populated it with matched FET's. I used the GEO circuit to match them. The phase 90 worked perfect.
Then I had to test so I took 4 of the FET's that were most apart and populatet the circuit with. I cant remember how far apart they were but it was just that is was plenty.
I was quite surprised to find that the thing still worked fine. I tested both with J201 and 2N5457 with the same result.
Someone who have an explanation for this

R.G. · March 29, 2010, 10:19:23 AM

Quote from: solderman on March 29, 2010, 04:37:23 AM
I am a curios cat. So when I built my MXR phase 90 clone I first populated it with matched FET's. I used the GEO circuit to match them. The phase 90 worked perfect.
Then I had to test so I took 4 of the FET's that were most apart and populatet the circuit with. I cant remember how far apart they were but it was just that is was plenty.
I was quite surprised to find that the thing still worked fine. I tested both with J201 and 2N5457 with the same result.
Someone who have an explanation for this

Sure. The "allpass" filter circuit we use for phasers contributes a phase shift even if the JFETs are replaced by fixed resistors. You'll still get the characteristic notches even if you don't change the phase resistances (which is what the JFETs are used for) at all. They just won't move around.

Matching JFETs is all about getting the notches to all be moving in the most noticeable part of the audio range at the same time. If one JFET is more sensitive than the others, it's already full resistance or least resistance while the others are still moving in the middle of their range. It's maxed out it's contribution while the others move. It will still make a phasing sound, but different from the sound where they all move in concert.

There was some scholarly work back in the 70s, a paper presented at the Audio Engineering Society (AES) which purported to show that phasers with matched resistors and matched phase caps produced the most intense effect. Notice that this is at odds with the Univibe, which deliberately mismatches the phase caps.

So will mismatched JFETs work? Sure, to some degree. If they're too far mismatched, the effect gets sketchy and limited. But if you get a whole batch out of the same shipment by the manufacturer, they tend to have come from the same die lot, and may be quasi matched anyway. You can get all degrees of from "works OK" down to "not very interesting" with random JFETs. You can get a more focused effect by matching. But it does depend on your ear.

News:

matching FETs