NOT biasing JFETs

[Rob Strand]

Any pointers? Wampler's doesn't look any different from Runoffgroove apart from the trimpot.

For a fixed drain resistor the only real starting point is use the typical JFET parameter, enter them into ltspice, and see where the drain voltage bias points end-up.    Some circuits bias a bit less than 4.5V some more than 4.5V.    If it's biasing with VDS at less than 1V or with VD greater than 8V then it's likely something is wrong.   After that try shifting the JFET parameters a bit and see where things land.   All this does is give you some confidence that it will work but it's still impossible to know where the designers board biased or where the production units end up.

The J201 and quite a few JFETs aren't so clear on typical parameter values and some extreme values are left open ended.   

If you have a circuit with a trimpot on the drain then *you* have choose where to set the drain bias voltage.   If you choose say 5V, use LTspice or a bias program to see if trimpot value gives 5V.    Some circuits have very large trimpot values and you can pretty much put any JFET in there and find a value that works!  Again, no way to *deduce* the target.  You might think to center the pot but that's way off in most cases.

So all in all not much can be done because you are kind of imposing a sensible bias point (in both cases).

In the first case it's temping to try the extreme JFET parameter, like my spreadsheet.  In a circuit with fixed resistors it's likely the circuit won't work or simulate well as the variations are too large - that's life with JFETs.   The aim of my spreadsheet is to vary the values to compensate for different JFETs and it's pretty clear even when you do that things don't work out all like the typical circuit.

JFETs have three parameters VP, IDSS and Yfs.   These parameters are linked mathematically and my spreadsheet makes sure the three parameters are at least consistent in that it doesn't let *any* of the parameters go outside of the datasheet range.   In practice, the extremes don't occur in reality.  So from this perspective my spreadsheet values are still too extreme even though they meet the datasheet values.    What you find in practice is the VP, IDSS and Yfs value are more more tightly linked since extreme Yfs values are rare.  If you use more narrow Yfs values that will automatically reduce the range of IDSS and VP value because it forces IDSS and VP to be more closely related instead of being independent (and allowed to individually take on the datasheet extreme values).   That's where things aren't as bad as they seem.   The point of all this is when you play with different JFET parameters in spice you can use a much narrower set of parameters to see where the drain voltages end-up.

I've applied all the above to quite a large number of circuits, in some cases I even knew the answer because someone took measurements of either the JFET or the circuit voltages.    My conclusion *back engineering* JFETs circuits is very imprecise.  It is only useful as a guide and bordering on a waste of time, since without the original circuit voltages you are better of just making the drain voltage 5V!