JFET Shootout

[Chuck D. Bones]

I started with this:

This only a guide, it does not contain enough detail to identify whether a part marking is legit or counterfeit. For that, you have to visit the manufacturer's website for details info on the logo, part numbering and LDC marking.

I went to SSI's website and searched their inventory for 2N404. Finding those, I looked for other Ge part numbers. There are distributors that do business in the niche market of obsolete and hard-to-find parts. SSI is one of them. SSI does not manufacture semiconductor die. They buy and resell complete components. In the case of obsolete parts, they frequently buy surplus die from manufacturers, test it, store it and package it with their own logo. I'm guessing that's what happened with the 2N404s.

 

[The Gator]

Ok. Sounds like im bread boarding that circuit and testing them all again.
Maybe If I notice a correlation between the readings. This tester will still be of use. I actually haven't tried the oscilloscope part yet

 

[giovanni]

Chuck, did you use this site? I don’t see AC128 in stock…

 

[The Gator]

I breadboarded the test circuit and even used a component clamp off another tester to help things go faster once I got the hang of it.
I am seeing a correlation in readings. I am wondering if there is a simple equation that could be used to get from one to the other as far as tested numbers go. Too early to tell just yet. However, so far, jfets that tested a dead match with the digital tester, also test a dead match with the breadboard circuit.

 

[Chuck D. Bones]

If we know Idss & Vgs,off, we can predict the drain current at any given gate voltage. That's because Idss (the dain current when Vgs =0) and Vgs,off (the gate voltage when the drain current is zero) gives us two points on a curve that we know from semiconductor physics is a parabola. But, if all we know is one point on the curve (Id at some Vgs), we can't predict Idss or Vgs,off.

 

[Chuck D. Bones]

Chuck, did you use this site? I don’t see AC128 in stock…

I did and you're right.

 

[temol]

@Chuck D. Bones - do you see any use for the MMBF4391?

 

[Chuck D. Bones]

Their high Vgs,off makes them unsuitable for use as a common-source amplifier with a +9V power rail. Could be used with +18V rails or higher. The best application would be as a source follower, but you will need to screen them for Vgs,off around -4.5V if you power them from +9V.

 

[Bricksnbeatles]

@Chuck D. Bones does the accuracy of the RG keen method require that the battery be at exactly 9v? I know fresh 9v are often 9.6v, while some are only around 8v out of the box. Planning on finally putting together one of these and wondering if it would be a good idea to feed it with 9v regulated down from a 12v power supply, just for the sake of consistency across the board.

 

[Chuck D. Bones]

@Chuck D. Bones does the accuracy of the RG keen method require that the battery be at exactly 9v?

No. The Keen JFET Matcher is an impedance bridge. It automatically adjusts the JFET's gate voltage until the on-resistance is equal to Rset. You could run it on 12V or 18V and it would work the same. I don't recommend using batteries because making them and disposing of them is bad for the environment. When I have to use batteries, I use rechargeables.

 

[temol]

So... what's on your list when it comes to current production/easily available JFETs we can use in gain stages?
5457, 201, 4393, 5458, 5485 do not participate in the ranking

 

[szukalski]

Let us never talk of 5458 and my inability to procure them cost effectively.

 

[Chuck D. Bones]

201, 4393, 5485 are all available in SMD.

If we're talking about thru-hole, then J112 & J113.

 

[Smrtokvitek]

What about 2SK208 and 2SK209? I have used some for buffers, but from what I remember, the spread of parameters was quite tight. Easily obtainable. Still in production.

 

[Chuck D. Bones]

Take a look at the datasheets. Same broad spread as most JFETs.
2SK208


2SK209


2SK208 is available from Mouser for $0.20 ea in lots of 100. SMD only.
2SK209 are $0.27 ea in lots of 100. SMD only.

 

[temol]

Yeah. I forgot about 2SK208. AMT uses 2SK208R in their preamps. Very small spread of Vp and Idss. I measured several dozen pieces - Vp 0.6–0.63V Idss 0.5–0.52mA. Can't say the same about J201.

 

[Chuck D. Bones]

I find that if I buy a group of JFETs from an authorized distributor, Mouser for instance, it's extremely likely that all of the parts in that group will have come from the same die lot. Idss & Vp will be tightly clustered. On the other hand, if I buy a group of JFETs from someone further down the food chain, like Tayda, Small Bear, or Electronic Goldmine, it's likely that the parts will come from multiple die lots. Then Idss & Vp will be more spread out. OOP parts, like J201, are bound to come from multiple lots. The majority of J201s I've been getting from a reputable US supplier have specs that are tightly clustered, at least so far.

 

[temol]

Can anyone measure couple 2N3819 for me? I have the datasheet but I'd like to know Vp and Idss taken from the actual transistors. 4-5 pieces would be perfect.

 

[Chuck D. Bones]

I don't have any, just checked.

 

[lrgaraujo]

Can anyone measure couple 2N3819 for me? I have the datasheet but I'd like to know Vp and Idss taken from the actual transistors. 4-5 pieces would be perfect.

I can do it, give me a couple hours! Not sure how reliable the Idss readings will be, because it can be too high for the jfet matcher (if I remember correctly)