testing JFET's

[jkokura]

I've searched and not found the answer for this one.

My DMM has a space for testing transistors - I can use it for checking HFE's... I have no idea what HFE's are, but it's a function of the DMM and I've tried it out.

Anyway, I'm building a phase 45, and my question is whether I can use this space on my DMM for testing the JFETs somehow, or if I need to build a JFET tester from GEO?

Jacob

[CynicalMan]

I've searched and not found the answer for this one.

My DMM has a space for testing transistors - I can use it for checking HFE's... I have no idea what HFE's are, but it's a function of the DMM and I've tried it out.

Anyway, I'm building a phase 45, and my question is whether I can use this space on my DMM for testing the JFETs somehow, or if I need to build a JFET tester from GEO?

Jacob

No. hFE is only a property of BJTs, not FETs.

[jkokura]

So is there no way to test jFet's and match them with a DMM? Do I need to build a matcher? I'm not sure if I'm going to be able to find all the parts locally...

Jacob

[R.G.]

So is there no way to test jFet's and match them with a DMM? Do I need to build a matcher?

I think I'd say this another way. Yes! It *is* possible to test JFETs and match them with a DMM! All you need to add is this one opamp and a couple of resistors to do it!

I'm not sure if I'm going to be able to find all the parts locally...

Parts procurement is always a problem. Most opamps can be used in that position, and any series/parallel combination of resistors to get the values works. You can also use different resistor values, as the article  notes, if you are willing to do some math based on your real resistor values and the readings you get.

[jkokura]

Thanks RG. As always, I appreciate all your involvement here. I was hoping it would be you that would answer

Jacob

[jkokura]

I guess I do have the parts! I was planning something else for a piece of perf, but I may as well use it for this, and I have a socket that I can mount with 3 10k Resistors and a 10uF cap - then all I do is 'borrow' a TL071 from another project for a little while to test JFETs. Now, the only problem is that I only ordered 2 JFETs for my Phase 45, and if they don't match I'll have to order more until I find 2 that match... How many do you think it will take to have two that match?

Jacob

[gigimarga]

I guess I do have the parts! I was planning something else for a piece of perf, but I may as well use it for this, and I have a socket that I can mount with 3 10k Resistors and a 10uF cap - then all I do is 'borrow' a TL071 from another project for a little while to test JFETs. Now, the only problem is that I only ordered 2 JFETs for my Phase 45, and if they don't match I'll have to order more until I find 2 that match... How many do you think it will take to have two that match?

Jacob

I bought 6 and I had a lot of luck to find 2 very well matched...so I think that you can buy minimum 10

[PRR]

What do you want: Vto or Igss?

The Official Value, or just parts with the same value?

You can "match" with battery, resistors, and digi-voltmeter. A pushbutton adds convenience.



As shown it measures Vto, the turn-off voltage. Not the exact definition, but two parts which measure the same this way have the same Vto.

Push the button, you read Idss, the maximum zero-bias current. Well, not precisely, but again two parts which measure the same this way have the same Idss. Actually you read Idss in Amps times 10. Or since you should get a DVM reading under 200mV, divide the mV number by 10. 156mV is 15.6mA. 13mV is 1.3mA.

[yongki]

You see this http://www.geofex.com/Article_Folders/fetmatch/fetmatch.htm

[jkokura]

You see this http://www.geofex.com/Article_Folders/fetmatch/fetmatch.htm

Before I posted, I read that.

[PRR]

> Before I posted, I read that.

I hadn't.

In theory, for JFETs of similar geometry, finding the two extremes Vto and Idss "tells you everything" about the JFET.

But we often need to work an FET "mid-way between the extremes". The parameters can be computed. It is easy to get it wrong. And too darn tedious.

R.G.'s plan very cleverly measures the parameter at a MID-POINT. He picked 10K, a typical mid-point for circuits which swing semi-log from 2K to 50K, giving a nice several-octave or many-dB shift. But you can change Rset to whatever suits your particular plans.

I can think of a way to avoid the opamp and get the same numbers but its too tedious to use.

[jkokura]

Yeah, and I have 3 10k resistors and a 10uF AE cap to use. I have a TL071 I can borrow from another unfinished project. Only, if my two jFets don't match I'll have to wait till I have enough cash to buy more to test and find matching ones.

The other alternative is to post in the FS forum to see if someone already has matching 5459's to sell. $1 plus the cost of a stamp to Canada and...

Jacob

[mac]

R.G.'s plan very cleverly measures the parameter at a MID-POINT. He picked 10K, a typical mid-point for circuits which swing semi-log from 2K to 50K, giving a nice several-octave or many-dB shift. But you can change Rset to whatever suits your particular plans.

i did that, the BF245A in my P90 are almost identical in a large range.

mac

[R.G.]

In theory, for JFETs of similar geometry, finding the two extremes Vto and Idss "tells you everything" about the JFET.

But we often need to work an FET "mid-way between the extremes". The parameters can be computed. It is easy to get it wrong. And too darn tedious.

R.G.'s plan very cleverly measures the parameter at a MID-POINT. He picked 10K, a typical mid-point for circuits which swing semi-log from 2K to 50K, giving a nice several-octave or many-dB shift. But you can change Rset to whatever suits your particular plans.

You wouldn't believe the long series of posts I've had telling me I did that "wrong". 

[slideman82]

You see this http://www.geofex.com/Article_Folders/fetmatch/fetmatch.htm

Before I posted, I read that.

I've never got this one working...

[box]

Hi,
try this one:
http://www.runoffgroove.com/fetzervalve.html
Best regards,
box

[jkokura]

 I've never got this one working... [/quote]

I did! yippee!

But I will note that I don't think my jFets match. One rated at -1.896 and the other at -1.948 which, unless I'm terribly misinformed, is definately not a match.

I have 2N5952's for my Phase 45. I have a bunch of j201's for an upcoming project (Dr. Boogie) that I could check and use for this phase 45 if they matched, but will I be able to use the 5952's in the place of the j201's in the Boogie?

If not I'll have to do as suggested and order 10 jFets and hope I can find 2 of the 12 that match. Then I find a use for the other 5952's... Suggestions?

Jacob

[PRR]

> One rated at -1.896 and the other at -1.948 which, unless I'm terribly misinformed, is definately not a match.

Depends what you are doing, but.... for many-many MANY purposes, a 3% match is more than close enough.

Also for many purposes, the actual value/effect goes as Square Root of the Gate voltage. So this is a 1.3% match. I would not hope to match things so close in any audio system, not without a bushel of cash. To find this in two laying-around parts is fabulous.
> telling me I did that "wrong". 

I had to stare at this "odd" plan a while to realize that it cuts right to the meat of the matter.