Jfet frequency response question

[HKimball]

Can you adjust frequency response (for example a bass rolloff circuit) by varying the resistors/capacitors on the source prong of a Jfet?

I ask because I'd like to add a potentiometer in conjunction with the "Fat" control, instead of a body switch, to adjust the frequency response of a JFET boost. I've been looking at the ROG Omega and I became curious if you could add a 'fat' control (q1 gain like in the fatpants) and something like a 'body' control but lower frequency so that it acts like a bass cut (I don't know if you'd be able to boost bass or what have you, I'm just spitballing really because I just watched a youtube video on how to calculate the gain of a transistor and I was completely lost.).

Obviously this wouldn't be a true treble/mid/bass tonestack, just a way to implement a similar functionality by adjusting the input impedance ('range' on the ROG), gain ('fat' on the fatpants), and whatever affects bass response. The overall goal for the pedal would be an always-on low parts count tone shaper/conditioner/enhancer with an output buffer. Maybe as part of an OD/Dist/Fuzz pedal eventually, I don't know. I'm mostly curious as to whether this is possible.

As always, thanks for any and all replies.

[midwayfair]

It's just a type of transistor, so anything you can normally do on the emitter of a BJT works, too. FETs have pretty specific biasing requirements compared with BJTs, but that's it.

If you put a capacitor in parallel with the source resistor, it will form a high-pass filter at a frequency determined by the standard cutoff calculation (1/Tau*r*c). R is your source resistor. C is your capacitor. Increase C or R and the frequency drops. This is an active high pass filter, meaning it ADDS highs, but increasing the gain at a specific frequency. The amount of increased gain you get at that frequency varies with the size of the source resistor and with the amount of gain that can be produced by the JFET. Runoff Groove gives instructions for doubling the gain of a JFET in their FETzer Valve article.

There are any number of reasons implementing a control like what's in the Fatpants would be a bad idea in the Omega.

First, the Fatpants uses smaller source and drain resistors. Much smaller. This allows you to ground the source of the FET without either reaching cutoff by raising the bias too much at minimum or over saturation by lowering the bias too much when the control is turned up. (The Fat control is a gain control that incidentally changes the -3dB cutoff frequency of the capacitor in parallel.) So you'd have to work out what the minimum and maximum resistance you could have would be so that you don't end up with settings that produce no sound.

Second, that 47uF in parallel with the source resistor pretty much means that you're not cutting any bass without making the resistance INCREDIBLY tiny. A quick calculation puts 72Hz at 47R for the source resistance! Not only would your control be impossibly bunched up, but you'd never hear a bass cut until the capacitor is basically out of the circuit and the source is grounded.

Third, the the Fat control alters the bias -- but the entire point of the omega is to mimic the behavior of a germanium rangemaster, including the 7V bias. If you change the bias using a Fat control -- assuming you solved the problems above -- then you change the clipping behavior of the booster.

Fourth, the Omega already has a control that changes the bass rolloff at the beginning of the circuit. That's what the "range" control is doing -- it's altering the input high-pass filter formed by the gate resistance and the input capacitor. So you'd be duplicating that. Yes, the input impedance changes when you dial down the Range, but that's secondary to the function of the control.

If you want to create an Omega variant where you can change the bass response of the booster without changing the input loading (perhaps if you don't like the incidental high treble increase that comes from decreasing the loading when adding more bass with the current "Range" control), look at varying the 47uF cap. You could put the cap in series with a 50KC pot and put a 100nF across it, and you'd have a control that blends in the larger capacitor as it's turned up. Or you could do the same input cap blend as is found on most Rangemaster variants.

[HKimball]

Thank you so much for the in depth explanation.

I'm not trying to recreate any specific thing about any specific circuit, so I realize that 7v is an important number, but I was planning on (1) running the whole circuit at much higher voltages (18v) and (2) have Q1 function as a kind of a clean preamp/tone shaper for Q2. I had originally wanted to use the tone stack from the Umble right after Q1 (which in that case would have a fixed bias because that would be way too many knobs for me haha) and then experiment with a volume control after Q1 to figure out a good resistor to use to limit the current going to Q2 (which I wanted to use a J201 for).

When looking at the omega, I thought "hey, that's a brilliant idea" and I thought it might be cool to have a "tone stack" that alters the parameters of the circuit rather than just roll off certain frequencies. So does the range control affect both the treble and bass? If that's the case, with the "fat" control primarily affecting the midrange character, they could coexist within the same pedal if (1) I did the necessary calculations to find appropriate source and drain resistors, and (2) didn't care about the 7v bias that's characteristic of the rangemaster. I remember in the article they said a 5v bias was 'fatter' so I feel like it's an idea at least.

I wanted to use Q1 as kind of a preamp, with an MPF102 and the aforementioned 'range' and 'fat' controls, and then use a dual ganged pot in connection with Q1 similar to the gain/clean you find on the Klon, but where the "gain" control is just a variable resistor controlling how much current two clipping diodes see. So, kind of a character pedal vs. one with a traditional tone stack.

Again, thank you. Clearly I have a lot of reading (and math) to work on.

[midwayfair]

I'm not trying to recreate any specific thing about any specific circuit, so I realize that 7v is an important number, but I was planning on (1) running the whole circuit at much higher voltages (18v)

Well, you'd put it at 14V if you wanted to keep that part the same. The drain voltage changes how the waveform clips.

and (2) have Q1 function as a kind of a clean preamp/tone shaper for Q2. I had originally wanted to use the tone stack from the Umble right after Q1 (which in that case would have a fixed bias because that would be way too many knobs for me haha) and then experiment with a volume control after Q1 to figure out a good resistor to use to limit the current going to Q2 (which I wanted to use a J201 for).

What you're describing in a sense completely alters the circuit. In fact, you can't even use the devices as described and it no longer really makes any sense at all to use the omega circuit for what you've described.

A quick recap on what the Omega is doing is in order. It's mimicing a treble booster that used a germanium transistor, high gain, and heavy input loading. The circuit is DESIGNED to clip, and in a particular way, so using it as a "clean preamp" wouldn't be good. It's designed to create a heavy load on the pickups to reduce the treble from the input device, which makes the distortion created more pleasant. Together with the bass cut from the input cap, the stock circuit creates a peak at about 1KHz (coincidentally right in the "hole" in a Vox amp's tone circuit, which is why they were a match made in heaven); the range control adds more bass, up to a point. Any tone control you put after it is going to be nerfed by the overall design of the circuit, which is cutting highs and bass and creating a big midrange peak. If you correct the input loading and remove the bass cut, you might as well just use the FETzer valve, which is a simpler circuit and can be full-range.

Additionally, the J201 is a poor choice for a clean preamp because it has very low input voltage headroom (under a volt, sometimes as little as half a volt). FETs are either lots of gain and low input headroom or very little gain and lots of input headroom. It's one or the other, and it doesn't matter what your supply voltage is. If you need a clean preamp, there are better devices to use. We use FETs because they sound better when they clip than other options.

Continuing, Q2 is more than just a source follower. (As a note, you can also use an NPN BJT in that spot without any changes and it'll be lower noise.) You mentioned using a J201, but I think that device won't even work in that spot the way it's set up; the gate would be referenced to your positive supply rail and not ground. I'm not totally sure about that but I think at best it might sound strange, but it will also very definitely clip if you put a FET there after all the gain coming from Q1. You also will lose a huge amount of gain from Q1 if you change how the circuit is set up. Right now the circuit has a gain of over 30x -- if you decouple the two transistors, even running on 18V and using a J201 (the highest gain JFET) you won't get that much gain.

[/quote]From Q2 I wanted to mimic the parallel clean boost/OD portion of the Klon, but where you jumper two lugs of the "gain" control and use it kind of like a variable resistor, with the "clean" portion of the b100k as a simple (and subtle) treble rolloff so that there isn't too much strident high end.[/quote]

I really don't understand what you're describing here or how it applies to this circuit. Depending on which leg of the gain control you jumper, on the distortion side you either turn it into a pure gain control (it doesn't attenuate the signal into the gain section, but it's still varying the gain, and it doesn't change anything about the treble) or you turn it into a variable low-pass filter and remove all the gain from the distortion section, which probably wouldn't sound like much of anything. On the clean side, you're varying the gain of the third op amp stage rather than making it a voltage divider, which isn't really any different except that you'd remove the ability at the highest gain setting to dial out the clean completely. Any effect it has on the treble is incidental because of the reduced gain. (And the klon already sounds less trebly when you turn up the gain ... or turn down the treble control, which is doing the same thing as varying the gain coming from the clean side.) In any case, I can't see how that applies to the omega. You haven't described anything that would create a treble cut, and you'd need to add more devices to create a clean blend.

edit: I do have another question. If I dropped anything resembling a bass control, would using both the aforementioned 'fat' and 'range' controls on a single JFET be a good idea? Or is there too much overlap between their functionality?

Again, the "Fat" control is essentially unworkable in the Omega. It's doing exactly the same thing that the trimpot in the Omega is doing, but you won't get any sound throughout part of its range. And the "range" control IS a bass control. It sets the high-pass filter with the input cap. I'm also not entirely certain why you want to use the Fatpant's gain control. What specifically does it do that you're trying to achieve? You also say you want a fixed bias -- again, re-read what I wrote about the Fat control. It's a bias control. Its effect on the EQ is collateral damage.

I thought it might be cool to have a "tone stack" that alters the parameters of the circuit rather than just roll off certain frequencies.

Honestly, I don't think this is really a useful way to think of things. In a transistor circuit, for the most part you're just changing what the circuit amplifies. You can tell the circuit not to amplify all the frequencies by using negative feedback and bypassing the source, but in a circuit with only a couple devices, how different is that really from just cutting some frequencies before or after the active device? Let's take the source bypass capacitor as an example. Depending on how big the capacitor is, it either makes the transistor amplify all frequencies, or it makes it amplify all frequencies except some bass frequencies. Just because it's "technically" an active filter and "boosting" certain frequencies, how different is that really than cutting the bass before it gets to the transistor and then amplifying what's left? You still pick a frequency to set as your -3dB cutoff, and the filter is still 6dB/octave just like any passive single-pole filter. They're functionally identical in a circuit this simple.

If you were working on a circuit that didn't use a source bypass capacitor, adding one to make a high-pass filter would make a pretty substantial difference, because then you'd have a gain control that also cuts bass if you want. If you were working on a circuit with several amplifying devices and could devote each device to one EQ task (this one boosts just the bass, this one boosts just the treble, etc), then there would be more sense in discussing the benefits of active vs. passive EQ.

All I'm really trying to say is that you're asking a lot from the Omega. It's got two transistors in it, but you really need to be thinking of it as a SINGLE amplifying stage. That really limits how much cool stuff you can squeeze out of it.

[HKimball]

I'm not trying to recreate any specific thing about any specific circuit, so I realize that 7v is an important number, but I was planning on (1) running the whole circuit at much higher voltages (18v)

Well, you'd put it at 14V if you wanted to keep that part the same. The drain voltage changes how the waveform clips.

and (2) have Q1 function as a kind of a clean preamp/tone shaper for Q2. I had originally wanted to use the tone stack from the Umble right after Q1 (which in that case would have a fixed bias because that would be way too many knobs for me haha) and then experiment with a volume control after Q1 to figure out a good resistor to use to limit the current going to Q2 (which I wanted to use a J201 for).

What you're describing in a sense completely alters the circuit. In fact, you can't even use the devices as described and it no longer really makes any sense at all to use the omega circuit for what you've described.

A quick recap on what the Omega is doing is in order. It's mimicing a treble booster that used a germanium transistor, high gain, and heavy input loading. The circuit is DESIGNED to clip, and in a particular way, so using it as a "clean preamp" wouldn't be good. It's designed to create a heavy load on the pickups to reduce the treble from the input device, which makes the distortion created more pleasant. Together with the bass cut from the input cap, the stock circuit creates a peak at about 1KHz (coincidentally right in the "hole" in a Vox amp's tone circuit, which is why they were a match made in heaven); the range control adds more bass, up to a point. Any tone control you put after it is going to be nerfed by the overall design of the circuit, which is cutting highs and bass and creating a big midrange peak. If you correct the input loading and remove the bass cut, you might as well just use the FETzer valve, which is a simpler circuit and can be full-range.

Sorry, I wasn't specific enough about it - I had contemplated two alternate designs. Were I to use the umble tone stack, there would be no range control; in that case I agree something like the fetzer valve would be more appropriate.

Additionally, the J201 is a poor choice for a clean preamp because it has very low input voltage headroom (under a volt, sometimes as little as half a volt). FETs are either lots of gain and low input headroom or very little gain and lots of input headroom. It's one or the other, and it doesn't matter what your supply voltage is. If you need a clean preamp, there are better devices to use. We use FETs because they sound better when they clip than other options.

I had planned on a two transistor design - Q1 was to be a very low output transistor (MPF102) and Q2 was to be a high output (but low headroom) transistor. For the prototyping stage I had planned to use a potentiometer as a "volume" control to determine what resistor I would need to keep Q2 from overdriving too much.

Continuing, Q2 is more than just a source follower. (As a note, you can also use an NPN BJT in that spot without any changes and it'll be lower noise.) You mentioned using a J201, but I think that device won't even work in that spot the way it's set up; the gate would be referenced to your positive supply rail and not ground. I'm not totally sure about that but I think at best it might sound strange, but it will also very definitely clip if you put a FET there after all the gain coming from Q1. You also will lose a huge amount of gain from Q1 if you change how the circuit is set up. Right now the circuit has a gain of over 30x -- if you decouple the two transistors, even running on 18V and using a J201 (the highest gain JFET) you won't get that much gain.

...
I really don't understand what you're describing here or how it applies to this circuit. Depending on which leg of the gain control you jumper, on the distortion side you either turn it into a pure gain control (it doesn't attenuate the signal into the gain section, but it's still varying the gain, and it doesn't change anything about the treble) or you turn it into a variable low-pass filter and remove all the gain from the distortion section, which probably wouldn't sound like much of anything. On the clean side, you're varying the gain of the third op amp stage rather than making it a voltage divider, which isn't really any different except that you'd remove the ability at the highest gain setting to dial out the clean completely. Any effect it has on the treble is incidental because of the reduced gain. (And the klon already sounds less trebly when you turn up the gain ... or turn down the treble control, which is doing the same thing as varying the gain coming from the clean side.) In any case, I can't see how that applies to the omega. You haven't described anything that would create a treble cut, and you'd need to add more devices to create a clean blend.

I wrote up something real quick to show what I'm talking about, which I've attached. It's missing capacitors and resistors pretty much everywhere but it's to give a general idea of what I'm talking about. R1 and R3 (the two variable resistors) are supposed to be tied (they represent the two levels of a 100kb dual ganged potentiometer)

I really should upload the image but I've already spent way too much time on this post. I'll try and do it later today though.

...

Again, the "Fat" control is essentially unworkable in the Omega. It's doing exactly the same thing that the trimpot in the Omega is doing, but you won't get any sound throughout part of its range. And the "range" control IS a bass control. It sets the high-pass filter with the input cap. I'm also not entirely certain why you want to use the Fatpant's gain control. What specifically does it do that you're trying to achieve? You also say you want a fixed bias -- again, re-read what I wrote about the Fat control. It's a bias control. Its effect on the EQ is collateral damage.

Again I wasn't clear enough. Here are the two separate circuits I'm talking about:

(1) umble tone stack. No range control, no fat control. No changes from the Umble schematic on the ROG website until the "volume" potentiometer.
(2) Omega-based. Range control + limited 'fat' control.

I really do appreciate you bearing with me. I can be all over the place sometimes.

...

Honestly, I don't think this is really a useful way to think of things. In a transistor circuit, for the most part you're just changing what the circuit amplifies. You can tell the circuit not to amplify all the frequencies by using negative feedback and bypassing the source, but in a circuit with only a couple devices, how different is that really from just cutting some frequencies before or after the active device? Let's take the source bypass capacitor as an example. Depending on how big the capacitor is, it either makes the transistor amplify all frequencies, or it makes it amplify all frequencies except some bass frequencies. Just because it's "technically" an active filter and "boosting" certain frequencies, how different is that really than cutting the bass before it gets to the transistor and then amplifying what's left? You still pick a frequency to set as your -3dB cutoff, and the filter is still 6dB/octave just like any passive single-pole filter. They're functionally identical in a circuit this simple.

This is what I needed to hear. For some reason I had gotten it into my head that there was some kind of difference in fidelity depending on where in the circuit the tone control is. Clearly there's no appreciable difference so I'll just go with the umble tone stack rather than try and figure out using the range and fat control in conjunction with one another. In my head the fat control would have functioned kind of like a "timbre" control where you can affect the girth or character of the frequencies determined by the range control.

If you were working on a circuit that didn't use a source bypass capacitor, adding one to make a high-pass filter would make a pretty substantial difference, because then you'd have a gain control that also cuts bass if you want. If you were working on a circuit with several amplifying devices and could devote each device to one EQ task (this one boosts just the bass, this one boosts just the treble, etc), then there would be more sense in discussing the benefits of active vs. passive EQ.

Again, I needed to hear this. As a conceptual matter I wanted to use the range and bias controls to adjust the circuit for different guitars and amplifiers. I don't need a 6-band EQ but I wanted something a little more versatile than a "tone" knob.

At the top of the bias/fat/timbre control I wanted to induce some very mild clipping in Q2 but the primary source of the clipping would be the diodes following Q2. Obviously in the schematic I'm missing a few capacitors that would follow the clipping stage.

All I'm really trying to say is that you're asking a lot from the Omega. It's got two transistors in it, but you really need to be thinking of it as a SINGLE amplifying stage. That really limits how much cool stuff you can squeeze out of it.

See this is where I got tripped up. I've looked at different designs - the umble, the peppermill, the omega, the King of tone, the sun king, the kingslayer, the marshall bluesbreaker, the jfet vulcan, etc etc etc and thought that it would be feasible to design a really simple, two transistor OD pedal (with diode clipping) and a kind of "familiar yet new" tone circuit. I love the sound of my fatpants with my bluesbreaker/PoT clone, but I really want to split the fatpants in two and put the buffer before the OD pedal and the boost after. Kind of like literally every other two switch distortion/od pedal out there does, but with a discrete circuit instead of using op-amps.

[HKimball]

One last bump to avoid making a new thread.

So, I'm listening to midwayfair here - I don't have a ton of time to teach myself how to properly bias a Jfet (i.e. what values you need to make a 'fat' and 'range' control work on the same transistor) so I'm just going to slap together the Omega and the Fatpants design/circuit but with 2 changes - (1) I'm using an MPF102 as a shared buffer between the high impedance Omega side and the lower impedance Fatpants side, (2) I'll use the 10ka "master" control of the omega as a sort of "gain" control to overdrive the J201 in the fatpants.

Of course, there are 3 controls that will affect the gain of the circuit, so even with the 'pre' knob maxed you can still dial it down with the 'fat' or 'range' controls. It's just that the 'pre' is the most obvious one.

It will be a true bypass pedal so I don't have to worry about how to make the buffer (which is in the middle of the circuit) switchable.

Thank you Midwayfair for your help. I know I seem silly for trying to go about designing a circuit without first learning the basics of transistors, so I thank you (and the rest of the forum) for bearing with me and taking the time to explain to me the folly of said attempts.