P channel JFET for switching?

[Toney]

I would like to test switching with a P channel JFET for use in a positive ground effect. I have some 2n5460 which look OK, I think?
VGs off is given as a range of 0.75 to 6v.  No reference to "typical" which is I suppose "typical" of JFETs
So can anyone recommend any better common JFETs?

Datasheet:
http://www.futurlec.com/Transistors/2N5460.shtml

A second question:
If I cannot fully switch off a JFET set up like a Boss type switch due to high VGs off say ~6v, is there any reason I cannot alter the voltage divider and provide the drain and source with a higher reference voltage?



 

[DiscoVlad]

From a cursory look at the boss bypass flipflop, the outputs to the jfet gates are pulled up to just under the supply voltage anyway. so 6V shouldn't be a problem.

The way to read the Vgs(off) min/typ/max value is that some of them will turn off at Vgs=0.75V, and that ALL of them will be turned off once Vgs= 6V (or goes above 6V), with the "typical" (hah!) JFET turning off somewhere in between these values.

[Toney]

Yeah, what I m going to do make a voltage divider and tie the drain and source to -3v, then sort through my few 5460s and hopefully at least one of them will be able to shut the channel via the gate. Then in use, I should have plenty of room to cleanly switch at half supply.
Kinda elcheapo sorting method...

[DiscoVlad]

Yeah, what I m going to do make a voltage divider and tie the drain and source to -3v, then sort through my few 5460s and hopefully at least one of them will be able to shut the channel via the gate. Then in use, I should have plenty of room to cleanly switch at half supply.
Kinda elcheapo sorting method...

With a P-channel FET you should have the  gate at least 3V higher than the source  
FETs can be confusing if you're used to BJTs because the gate needs to be reverse biased.

[Toney]

 Yep that whats Im doing, DiscoV. As per the Boss schem but set up with a P channel JFET
Hmm mixed results so far.
I can get them to switch but not all the way off... still going.

[Toney]

 Ahh got got one switching completely off with the drain and source tied to -4.5v.
Seems the -3v idea was too close to the threshold.

These are all I have at the moment but if anyone knows a more suitable device I'll order some while they still exist.

[DiscoVlad]

The Boss schematic I'm looking at (HM-2) has the n-ch JFET drain and source biased at +4.5V, and pulls the gates down to 0V to switch off.

Boss pedals use/used 2SK30A-Y JFETs which have Vgs(off) of -0.4, to -5V. Presumably -4.5V is enough to reliably switch most of the available devices or Y(ellow) 2sk30a's have Vgs(off) closer to 0V than the other colours.

[merlinb]

I would like to test switching with a P channel JFET for use in a positive ground effect.
So can anyone recommend any better common JFETs?

I have used 2N3820. They have a big Vgs(off) spread so you need to hand select, but it's all I had.

Incidentally, you may not need a P-channel. The JFET doesn't know what sort of circuit it is in; it only knows its gate-source voltage. You can use an N-channel JFET; pull the gate diode down to -9V to turn off, and pull it up to 0V to turn on!

[Toney]

Boss pedals use/used 2SK30A-Y JFETs which have Vgs(off) of -0.4, to -5V. Presumably -4.5V is enough to reliably switch most of the available devices or Y(ellow) 2sk30a's have Vgs(off) closer to 0V than the other colours.

Yep that's what I have read mostly reading things R.G. has written (like pretty much everything I know about pedals Lol  )  The 2SK30AY is particularly suitable for switching with 9v designs due to reasonably low VGs off some others that are OK for this role include 2n5485, 2n5952, J113.

[Toney]

I would like to test switching with a P channel JFET for use in a positive ground effect.
So can anyone recommend any better common JFETs?

Incidentally, you may not need a P-channel. The JFET doesn't know what sort of circuit it is in; it only knows its gate-source voltage. You can use an N-channel JFET; pull the gate diode down to -9V to turn off, and pull it up to 0V to turn on!

Thanks! I was thinking about this last night and reasoned this, then I kinda overdid it with thinking it through and decided I was probably wrong Lol.

[DiscoVlad]

*looks at the schematic again... and it appears that the switching FETs are all decoupled from the supply. N-ch should be fine.

The Flip-flop may (or may not) need alteration - but it's probably just a case of connecting the LED to the other side at worst.