Output level question

[Epameinondask]

Hello everyone. I have been desinging a single JFET booster (2n5457). My "problem" is that the output level is not as loud as I want. I'm not saying that it's lower than my clean signal, but when I engage the circuit on my breadboard the output level increase is very little,almost not increased.

So my question is who do i calculate the output level of a single JFET? I am not a complete noob, neither an effects guru    , but as i google searched for a solution, I'm beginning to think that it has something to do with the resistor values that are used to bias the transistor.If my assumption is completely wrong don't shoot me .

My desing is based on this schematic http://www.muzique.com/images/jfet1.gif. The differences are that i dont have a output pot (I want it to have a fixed output level) and that i have placed a capacitor  between the gate and the source (miller capacitance effect). Also the source resistor is a 10k and I bias the transistor with a 100k trimmer on the drain, C3 is 150NF and i dont use an R1 pulldown resistor because I still have it on my breadboard.Thanks in advance

[GibsonGM]

Hi Epa, welcome to the forum!  What i think may be wrong is simply that your values of resistance aren't correct for the circuit and the FET you're using.   You do have a voltage reference in there somewhere, right? (1/2 supply)

What voltage do you have at the drain?  I recommend replacing the drain resistor with a pot (50k, 100k...) and fine-tuning the voltage you read there to about 4.5v (1/2 supply), and see if that changes anything.  Then put in a fixed resistor.  I use 2N7000s sometimes, with take MEGs to bias correctly!   You have to treat each type of FET differently.

If you take a look at the "AMZ mosfet boost" on AMZ, you'll see how Jack has placed a pot in parallel with the source resistor, connected to source by a cap...so he has a variable source bypass, which acts as a gain control. You may want to check this approach out...set your gain level, then replace the pot with a fixed resistance     Moving a couple of wires will get you going.
Hope this helps.

http://www.muzique.com/lab/boost.htm  < click the link there for "mosfet boost"...

[induction]

... i dont use an R1 pulldown resistor because I still have it on my breadboard.Thanks in advance

R1 is not a pulldown resistor, it's used for biasing the jfet. Try putting it in and see if it helps.

[Epameinondask]

Thank you guys very much for the quick response. I'll check out your ideas and post the results asap. 

[ashcat_lt]

So my question is who do i calculate the output level of a single JFET?

Multiply the input level by the gain.  What's the input level?  Depends on the guitar, and how hard you hit it.  What's the gain?  You'll have to ask somebody smarter than me. 

[GibsonGM]

After the fact, you can tell the gain by dividing the voltage at the output by the input voltage.  This is done using the AC function of your DMM, of course.

If you input 100mV, and get 1 volt out, you have 1/.1 = 10.   

You may then assume that 1 volt in will = 10 volts out, BUT there are limits (is your power supply 9V? You can't go higher than <close to> that...plus, there is some frequency dependence to it, too).   So it's not written in stone, but makes a good rule of thumb. 

Nothing (IMO) beats just measuring it...making things adjustable via a pot etc., at least during design, makes things much easier.

[PRR]

You *need* R1 (more often than not).

For this task, for good gain, source resistor should be more like 1K (as Muzique showed), it must have a large (10+uFD) bypass cap, and drain resistor is likely to be 5K or 10K.

I suggest you leave R2 at 10K (good drive to output) and trim R3 until Drain voltage is around half of 9V.

[Epameinondask]

Haven't had the chance just yet to sit and try your ideas on my breadboard, buy my question more specifically is whether or not there is a formula or something that relates the resistor and caps that are being used with the output level and the db that the output reaches (without clipping if possible)like a resistor calculator. Biasing this schematic "by ear" so it sounds right I found that there are many combinations of resistors that can be used. But it seems, that such thing is not right. Are specific numbers used for biasing a transistor the main factor that determines the output level? If not, then what determines this? Thank you again and i hope i don't tire you with silly questions (i'm new to designing circuits and effects)

P.S. I suppose that the "Miller" capacitor doesn't affect the output, just the frequencies right?

[PRR]

> whether or not there is a formula

JFET gain computation is "possible", but difficult, and the actual devices have a broad spread of parameters so you have to try it.

Basic amplifier design.

For highest gain with reasonable system behavior:

Pick your plate (collector, drain) resistor 1/2 to 1/5 of the actual load impedance. (Yes, you need to know your load!)

Pick cathode (emitter, source) resistor to cause a current which sets plate (C, D) around half of supply voltage. (With BJTs this also means picking a way to bias the Base.)

*Bypass* the cathode (emitter, source) resistor.

The capacitors don't set gain, only bass-cut and treble-cut points. As long as these are well outside the frequency range you need, they don't matter.

[Frank_NH]

I have developed a spreadsheet for my own use which calculates gain for a common source JFET amplifier circuit (the formulas are available in many textbooks dealing with basic electronics).  However, the Fetzer valve article at Runoffgroove should answer many of your questions and at least show you one way of implementing a JFET boost circuit.

http://www.runoffgroove.com/fetzervalve.html

I should note that they key pieces of information that you need in order to determine the appropriate source and drain resistors for YOUR circuit are the JFET's pinch-off voltage (Vp) and saturation current (Idss).  While there is a range of values for these parameters for a given JFET like a 2N5457 (see the datasheet), the ** specific values must be measured for your transistor **.  I measure all my JFETs now before using them, and have been able to avoid problems with biasing as a result.   Check out the Fezter article for more information about measuring JFETs using a simple circuit and your DVM.

[hymenoptera]

> whether or not there is a formula

JFET gain computation is "possible", but difficult, and the actual devices have a broad spread of parameters so you have to try it.

Basic amplifier design.

For highest gain with reasonable system behavior:

Pick your plate (collector, drain) resistor 1/2 to 1/5 of the actual load impedance. (Yes, you need to know your load!)

Pick cathode (emitter, source) resistor to cause a current which sets plate (C, D) around half of supply voltage. (With BJTs this also means picking a way to bias the Base.)

*Bypass* the cathode (emitter, source) resistor.

The capacitors don't set gain, only bass-cut and treble-cut points. As long as these are well outside the frequency range you need, they don't matter.

This is why I love your posts, Paul. You make things make sense to the rest of us. Several key concepts just came together for me while reading this post. Thank you!

[blackieNYC]

Another question - I'd like to keep this thread alive.
I've built a few fetzers, in accordance with the runoffgroove instructions.  The pinch off voltage seems inversely proportional to the gain.  Does this mean there is no reason to chose a J201 over a MPf102?  Because the J201 will  clip sooner, will it not?  Always? And at a p-p voltage that a humbucker can easily put out.  It seems you'd have to take preventative measure, like an input attenuators.  And in the fetzer circuit at least, you can't get any gain out of the MPF102. Can you get gain out of them otherwise?  I realize the fetzer circuit is aiming to imitate a particular tube preamp stage and is not a get circuit maximized for performance.

[PRR]

> The pinch off voltage seems inversely proportional to the gain.

Partially.

The Vgs(off) is nearly the Maximum useful input swing. At zero you get maximum current (Idss); at Vgs(off) you get zero current.

The output swing is your *supply voltage* (nearly).

The ratio V(supply)/Vgs(off) is *roughly* the voltage gain possible.

It is not that simple. Working well below Idss the gain can be some higher.

> you'd have to take preventative measure, like an input attenuators.

Or an UN-bypassed Source resistor, which will about double the useful input swing without so much of the hiss issues of a pre-gain attenuator.

> you can't get any gain out of the MPF102. Can you get gain out of them otherwise?

Yes, but "not" with a puny 9V supply. The MPF102's Vgs(off) can be as high as 8V! No lower limit specified, but assume 4V to 8V. The ratio V(supply)/Vgs(off) is hardly over "2". Then when you throw in an unbypassed Source resistor, or load the output with anything similar to R(drain), the gain is about half, so about unity.

Run a 24V supply, 5K drain resistor, around 800r bypassed source resistor, hi-Z load, the MPF102 does around gain of 14 with input overload of a Volt or so and around 20V peak-peak out.

Look at the datasheets. J201 is an "amplifier". MPF102 is a "switch".

They are really nearly the same things, except sorted by "sensitivity". (Actually a "switch" part will be scaled to large area and low sensitivity; but there is a lot of overlap in both jobs and parts.)

I have several water valves. One needs 3 turns from off to on. Another needs 1/4 turn off to on.

The 1/4-turn valve "amplifies" my wrist-twist more.

So why do I have a 3-turn? Let's get on my tractor and bounce through the woods/swamp. Say the two valves have been levered: one needs 1-inch full motion, the other needs 12-inch full motion. I'm bouncing in the tractor several inches. With the 1-inch valve I'm constantly slamming. With the 12-inch valve the flow changes much less.

JFET "switch" applications have a "bouncing" signal and a wide-swing control. With a "switch" (high Vgs{off}) JFET of 8V turn-off, 1V "bounces" in the signal hardly affect the JFET's on/off-ness.

If you are crazy for gain, you probably want the "amplifier" parts.

Yes, in today's declining market you don't have a lot of choices.