J-FET bias Voltage Testing Circuit - help

[Rectangular]

hey all. I've got dozens of BF245C jfets that I need to comb though, in order to get specific bias voltages. they're being used for control voltage stuff and triggering in some old synthesizer circuits. now the schematics list what bias voltage ranges I require, I just have no way of actually testing for them. I've tried building R.G.'s j-fet matcher  ( http://www.geofex.com/Article_Folders/fetmatch/fetmatch.htm )  but the purpose of that circuit just seems focused on matching jfets, instead establishing what their true bias voltage is. am I wrong, or are the readings of that circuit somewhat arbitrary numbers, based around the "Rset"  10k resistor ?

I was hoping someone could help me construct a way to test for the true bias voltages in these jfets,  I do not require any matched pairs/sets though.

-rec

[db]

Probably the easiest parameter to measure is IDSS.  Just connect the source and gate to 0V and drain to +V via say a 1K resistor and measure ID by noting the volt drop across the resistor.  This will give you a rough and ready indication of where a device lies within the normal spread.

[Rectangular]

hey db

but if I need one  at, say,  2.5-3.5v ,  what's to determine where its actually at  ?  I know I can end up with a selection of jfet's that are higher or lower   in relation to each other, but I'd really like some way to get a concrete, objective number

[db]

Are you talking about VGS (off)?

[db]

Well anyway, what you could do is as I explained before but connect the gate to the wiper of a lin pot (100K or similar) with the pot connected between 0 and a negative supply (another 9V batt maybe).  Simply adjust the pot to get the current flow you decide to use as a reference point and then measure the corresponding VGS.

IDSS and VGS(off) are at the opposite extreme ends of the transfer characteristic shown in the datasheet.  i.e. IDSS is the value of ID when VGS=0 and VGS(off) is the value of VGS when ID=0 (or very close to 0).

[Rectangular]

to be honest, I dont know which of the two extremes I'm looking for. all I know is that I need a range of (for example)  "2.5-3.5v", and a method to select that. that's all it says on my debugging notes for these circuits, as to how to go about selecting the appropriate jfets, I'm not sure.

any further assistance on this would be greatly appreciated

[db]

I can't add any more without seeing the circuit you're thinking of.

[R.G.]

Welcome to the frustrations of using JFETs.

A JFET cannot pass more current than Idss. It does that when the gate is at the source voltage, or Vgs = 0V.

A JFET turns completely off when the Vgs is more than the cutoff value on the data sheet, Vgsoff.

In between the two extremes of Id = Idss at Vgs=0 and Id=0 at Vgs=Vgsoff, the current varies by the transconductance, gm. The units of gm are 1/resistance or amperes per volt.  The only thing keeping you from just specifying an Id by using Vgs times gm is that Idss, Vgsoff, and gm all vary by between 3:1 and 10:1 within the same type number.

So here's what you gotta do if you are replacing a JFET.
1. Calculate the current you need to flow in Id. If you have a spec of drain voltage, then convert that to an Id by using the drain resistance to figure it out.
2. Pick a JFET that has Idss bigger than the Id you need. If the JFET's Idss is smaller than the Id you need, you'll never get there. So test for Idss> Id.
3. From the JFETs with Idss big enough, set up a test circuit to measure the Vgs that gives you the correct Id. Swap in the JFETs that have big enough Idss and diddle the Vgs until you get the correct Id. Measure the Vgs. If your circuit can supply that Vgs, you're there. If not, keep looking.

[Rectangular]

hey RG,  thanks for giving it a shot. unfortunately I didn't understand a word of that...  I've tested for simple things like voltage drop and leakage on diodes,  hfe on transistors, but whatever's going on with these JFETs is clearly over my head.

in the notes for these schematics, whenever there's an important, highlighted jfet,  it will say something like: " select Q46: 3.5-4.0  OR  >1.5  Adjust  R43"  where R43 will be some resistor going to Ground. so I was trying to look for a solid number, like testing for hFE, but apparently its not like that at all.

[R.G.]

Sorry - I wish I could simplify it more. About now you're beginning to understand why JFETs are not more widely used.

[db]

It's not really all that difficult once you understand the basic function of a JFET and appreciate that they suffer from a high spread of performance (unless of course they have been selected by the manufacturer e.g. BF245A, B and C are essentially the same device preselected into more narrow bands of performance - even though there is still quite a bit of variance within those bands).

As I said previously, if you can post a circuit then perhaps we can help more than we already have.  It's impossible to interpret the schematic notes you quoted wthout a schematic.

[Rectangular]

unfortunately I can't post  some of these schematics, but here's a good example from one of them: http://www.hinton-instruments.co.uk/ems/vcs3env.gif.  look at Q45 there, the modern replacement of a 2N5163 is a BF245b (or c) which is what I'm using. the rest of the jfets I need fall within that same voltage range, so if you can figure out how to get this selected device, the rest are already accounted for.

thanks again for your efforts, guys

[slacker]

I don't really understand exactly what it all means but there's a circuit for measuring the vp of a jfet on the Runoffgroove Fetzer valve page http://www.runoffgroove.com/fetzervalve.html#11 so presumably you need to measure your jfets and use ones with a vp of between 3.5 and 4 volts. Which is probably what RG and db have already explained but in idiots language 

[Rectangular]

hey slacker, that looks great, it might be just what I'm looking for ! thanks a million

[Rectangular]

what do you think, db ? does it make anymore sense now that you have something to look at ?

[db]

I haven't tried the method shown (the 1Meg resistor from source to 0V) but it should work as far as I can tell.  Vp is the same as VGS(OFF) as we were talking about previously so I think this is the measurement you want.