JFET on resistance?

Measure one ?

Graham

Can't you get samples ? I've rarely had a problem for a genuine enquiry. The only company that gave me trouble actually was IR who said 10k devices p.a. wasn't enough.

Graham

On resistance is measured at Vgs = 0. Which gives @ 5V Id = ~ 0.07mA ( Fig 3 )

Roughly 70 k ohms.

What do want it for ? I can get you 8 ohms if you want.

Graham

1.5v/350 ua = 4.3K

But the datasheet has drain curves, and the slope of the drain curve near the origin, for Vg=0, *is* the on resistance. Looks like 1 or 2K depending on the Idss of the particular part.

John

Yup, he did make a truly bad guess, off by roughly 50:1.

Why not use a cmos analog switch like an HC4066?

Why not bottom post?

John

Well maybe you're looking at a generic JFET which is on and you must turn it off. If you can't find what your looking for, look for the pinch off voltage.

Unless we're talking about enhanced mode here?

"

OK, you can slow down the gate drive and de-click the switching, but you can't de-thump it using a simple jfet switch. Go slow enough, and you get lots of distortion during the transition.

But that jfet is a real wimp, intended as an electret mike amp. You'll need something beefier if you want to do jfet switching.

Don't people already make thumpless analog switch chips?

Led-photoresistor things could be interesting.

John

Hi;

Is there a way to figure out the R(on) of a JFET when this parameter isn't specified in the data sheet?

Thanks

So what's the matter with an MPF-102? Not available?

MOSFETs will work, too.

CMOS switches will probably provide soft switching if you ramp the 'enable' input -- the simpler ones are just a pair of MOSFETs with none-too-spectacular Rds-on figures.

--

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

Do you need to implement control loops in software?
"Applied Control Theory for Embedded Systems" was written for you.
See details at http://www.wescottdesign.com/actfes/actfes.html

isn't

I mean before I place the order.

Thanks

Thanks John. I was thinking there must be some way using Idss, but I wasnt sure which V to use. So for example from this data sheet:

it would be 1.5V/350uA = roughly 4.3 megohms.

I must be doing something wrong. Can you set me straight?

Thanks

isn't

Wow thats awful..

I'm looking for a JFET to switch a guitar signal in and out of effects pedals. I'll need 8 in total, but the signal will only ever be passing through 4 JFETS at any 1 time. I am going to bias the drain and source at

1/2 V+ (4.5v in this case) and would like to switch the JFETs on and off using a PIC. Since the PIC runs at 5v, I'm assuming it's control outputs will also be 5v, so I'll need a fairly low Vgs off to switch the JFETs completely off to isolate the effects from the signal chain when I want them bypassed. I realize the Ron doesn't need to be extremely low (although the lower the better, I suppose) but I also need the lower Vgs off. I'm trying to find a nice balance between those parameters.

Thanks

wasnt

3 )

them

the

trying

I'm trying to get the switching to be click free, and I've tried a MAX4622 CMOS switch, with no success. I tried using an MPF 102 and the switching was much quieter (only a tiny thump in one direction), and I've read that it's easier to ramp up the control voltage on a JFET than it is on CMOS. Which has led me to try ordering up a bunch of suitable JFETs to see if I can get them to behave.

Thanks

>

message

at

off

outputs

want

MAX4622

was

it's

I

Actually I breadboarded my design using an MPF 102, since it was all I had lying around. Unfortunately, it has a Vgs off of 8V, so it didnt provide great isolation when it was off. You're suggesting a MOSFET also? I've got some BS170s lying around so I'll do some work with them.

Thanks

5V is really too low a pinch-off voltage when you consider some guitar pickups can deliver volts of output.

Also, there tends to be a relationship between a higher pinch-off voltage and a lower on-resistance.

Typical jfets used for switching audio are the J09 and J175 families.

Graham

Actually I did seem I dropped a decimal place somehow. I get ~ 6k now. With Idss as low as 100uA it's always going to be a high resistance device. Note the application. Capacitor mics etc. It's designed for low current consumption.

Graham

Possibly, at insane prices.

Turn-off time is quite long.

Graham

You'll almost certainly need a seperate supply to drive the things off or the audio may simply break through by modulating the gate-source or gate-drain voltage. Beware, pinch-off voltage can vary widely too.

Graham

I've seen circuits that start with a series resistance and a switch that shorts out the signal when it's on -- these will handle higher voltages, and you can follow that with another switch that turns on/off to finish the job (although you'll then have to make sure that you've turned it on enough for that several-volt swing). Such things are fairly typical of RF circuits that use PIN diodes for switching, although I think I've seen it in audio paths.

I'd consider attenuating the guitar input to something guaranteed to be manageable by the switch, but I don't know if it'd get past the "better make sure this won't cause problems" stage of the design process.

--

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

Do you need to implement control loops in software?
"Applied Control Theory for Embedded Systems" was written for you.
See details at http://www.wescottdesign.com/actfes/actfes.html