Inductorless wah: JFET vs BJT as control element?

[Fancy Lime]

Hi there,

I have been doing some research into inductorless wahs based on twin- or bridged-T negative feedback. I am trying to figure out why JFETs seem so rarely used as control elements here. There are a bunch of envelope controlled variants such as the EHX Doctor Q and Bassballs that use BJTs, though. Has anyone built one of these adaped to using JFETs or maybe MOSFETs? I wonder if that would not provide a more "natural" sounding sweep. Just checking before I start breadboarding, don't want to reinvent the wheel.

[antonis]

I am trying to figure out why JFETs seem so rarely used as control elements here.

Bloody unpredictable bias, perhaps..??

[Fancy Lime]

I am trying to figure out why JFETs seem so rarely used as control elements here.

Bloody unpredictable bias, perhaps..??

But that is easily solved by adjusting the Range pot that these things tend to have anyway, no? Example, Jacks improved Doctor Q variant Dr Quack:
http://www.muzique.com/schem/quack.gif

Of course the range for the gate would have to be more negative than the source, so the whole DC scheme around a JFET would be slightly different than with a BJT. But that alone can hardly be the obstacle, can it?

I am mostly wondering about the difference in control voltage vs channel resistance / emitter current.

[iainpunk]

i breadboarded an inductorless wah using two BF245's but i couldn't quite get them to keep the Q constant across the board. there was a volume dip in the middle of the sweep and squealing for the last 5%, which i couldn't iron out and/or live with.

the bridge T filter had the bridge and ground legs being capacitance, and the arms being resistors and JFET's. the gain element i used was a simple opamp.

this was all quite in the beginning of my building journey (when i still insisted on usign a +/-3v bipolar supply), and there isn't much in my note book.

good luck with your journey

cheers

[amptramp]

You might have better luck with a MOSFET in a CD4007 IC.  You can gain access to an individual FET but it will be biased with the gate above the source (if you use the n-channel FET) and it would be interesting to have both the n-channel and p-channel in operation being driven from inverted gate drives so the p-channel gate goes negative when the n-channel gate goes positive to turn both on.  This would be interesting but probably not necessary.  Voltage and channel resistance are better characterized for the CD4007 than for most JFET's.

[Fancy Lime]

Hmmm, food for thought... Thanks!

I have been meaning to explore the fabled use of CMOS inverters as variable resistors to ground some time ago but could not dig up a schematic and was too lazy to figure it out myself. Today while walking the dog I finally understood how this would work. More fiddly than a JFET, it seems, but may be worth a try anyway.

Andy

[amptramp]

This is the datasheet for the MC14007, Motorola's (now ON's) version of the CD4007:

https://www.onsemi.com/pdf/datasheet/mc14007ub-d.pdf

Figure 1 shows a diagram for a 2-channel multiplexer that would be a great bypass design that avoids using 2SK30 unobtanium devices but the IC still allows access to individual MOSFET's even though all the n substrates are connected together and all the p substrates are connected together.  You could use multiple sections to get lower "on" resistance.  A CMOS inverter with separate gate control for n- and p-channel devices could be used to enable both the n- and p-channel devices to be used simultaneously.  This could give better linearity than a single device.

[Fancy Lime]

After some experimentation, I can answer the question why we don't see JFETs or MOSFETs or CD4007s or whatnot used in this role more often: it works perfectly well with a simple old BJT. There is some audible but pleasant distortion but it's barely noticeable due to the not-exactly-hifi nature of the wah-type filter effect. I am quite pleasantly surprised how well this works and curse myself for dismissing the whole concept of bridged-T wahs and envelope filters as half-arsery and inferior to more complicated topologies for so long . Live and learn, I guess.

[mac]

I am quite pleasantly surprised how well this works and curse myself for dismissing the whole concept of bridged-T wahs and envelope filters as half-arsery and inferior to more complicated topologies for so long . Live and learn, I guess

After tweaking the Colorsound Inductorless a bit It sounds fantastic, as good as an inductor wah. I used LED/LDR. No heavy duty pot, no complicated mechanism nor magical inductor :-)
I guess I can try a BJT or 4N27 which are more predictable.

mac