JFET that ain't no JFET...

[mateusborges]

Heyas!

Sooo, what would you do once after almost collapsing of anger you discover that your latest, non-working, build using a supposedly 2N5458 Fairchild JFET is actually a common BJT...

I exhausted all testings, measures and re-measures I could, and then, the single 2N5458 in my RAT circuit (probably) isn't working due to a freaking false BJT disguised as JFET!

><

I can't wait another 2 months for buying another replacement, wich common ones I could use here?

Cheers!

[Steben]

BF245A, MPF102, 2N5457 and 2N3819 according to R.G.

[antonis]

Wire that freaking false BJT as an Emirrer follower and earn peace of mind..!!

[Rob Strand]

Wire that freaking false BJT as an Emirrer follower and earn peace of mind..!!

Another option is plugging it into a hFE tester (multimeter, or dedicated device).  With the correct connections the gain will measure hFE = 200 (or whatever) then reverse C&E and it should measure hFE = 1 to 10, some low value.

JFETs should measure as silicon diodes between G and D and G and S but the thing that makes them different to BJT's is when you measure the ohms between D and S.  It should end-up being some low value resistance, and it should measure that resistance with the multimeter leads in both directions.   A switch JFET might be say 50 ohms and amplfier JFET would be 200 to 600 ohm.
BJTs will measure open between C and E in both directions.     Fried BJT's will often measure as a short between C and E.

[Steben]

Wire that freaking false BJT as an Emirrer follower and earn peace of mind..!!

Very keen.
All it takes is wiring the 1M to 1/2Vs = 4.5V point. And it will be louder than the original jFET buffer

[antonis]

All it takes is wiring the 1M to 1/2Vs = 4.5V point. And it will be louder than the original jFET buffer

So long as h_FE is equal to or higher than 100 or so..
For BJTs of lower DC gain, 1M resistor value should be lowered according to formula: [h_FE x 10k x (4.5 - V_BE)] / V_Emitter

[mateusborges]

BF245A, MPF102, 2N5457 and 2N3819 according to R.G.

Great, I have some MPF102, gonna try ´em.

Wire that freaking false BJT as an Emirrer follower and earn peace of mind..!!

Cool, will it work better than the JFET? How do I do it?

Cheers to replies!
Boos at disguised BJTs!

[antonis]

Cool, will it work better than the JFET? How do I do it?

All it takes is wiring the 1M to 1/2Vs = 4.5V point. And it will be louder than the original jFET buffer

[mateusborges]

I mean, I have to get the 1M resistor, the one in the circuit, and solder one of the ends in a place I get a 4.5v reading?

Sorry =/

[PRR]

I have pointed out before that the final buffer of the RAT could be replaced with a BJT with some improvement in repeatability.

Users have shouted me down because apparently they *like* the FET distortion. Even if some folks get more than others due to FET variability.

In these trying times, do the BJT until your FET ship comes in.

[287m]

NPN RAT

[PRR]

Link in 287m's post:

[mateusborges]

Thanks guys! Gonna study this a little and see what I can do =)

Cheers!

[mateusborges]

@PRR

Since there's some JFETs that performs better than others, is there any method for choosing "good ones"?

Cheers!

[PRR]

JFETs are going out of style. The current parts crisis does not help. The shipping crunch too.

I have no useful advice.

[mateusborges]

Woot! Working =D

[Steben]

All it takes is wiring the 1M to 1/2Vs = 4.5V point. And it will be louder than the original jFET buffer

So long as h_FE is equal to or higher than 100 or so..
For BJTs of lower DC gain, 1M resistor value should be lowered according to formula: [h_FE x 10k x (4.5 - V_BE)] / V_Emitter

Funny that in fact a small power silicon BJT with hFE < 100 is less cheap than just any regular one (which will be like 150 - 450).
But again, this a specific case since we don't know what transistor it actually is.

[Steben]

Woot! Working =D

BJT's !


jFET's are like the new germanium transistors: they have some qualities, but most are out of spec and the things they offer are arguably not "that" unique.
Even a ECC83 / 12AX7 is a better and predictable investment compared.

[fowl]

I guess I'll mention this, I've built three or four Rats so far, obviously I like them.  But recently I built a copy of Catalinbread's Katzenkonig, and I think it's a Rat killer.  Maybe that's where the "Cat King" name comes from?  But anyway, I can get pretty much all the good Rat sounds with it, but much more.  And it uses four BJTs, no JFETs or LM308s.  There are a few odd cap values, but nothing too hard to find.  For Qs I used BC550Bs with gains in the mid-300s, I'm not sure but I've read that Catalinbread uses those too.

[antonis]

There are a few odd cap values

There is no oddness in E6 series values..

https://www.docdroid.net/joL2PWr/catalinbread-katzenkonig-pdf