A Simple Way To Add More Gain To This FET Preamp?

I've just built a couple of these preamps to use as buffers on my acoustic guitar pickups. One is being driven from a Sunrise magnetic soundhole pickup, the other from 3 piezo bugs glued under the bridge plate of the guitar. going to a seperate outboard blender. The output from the piezo bugs is quite a lot higher than the magnetic soundhole pickup so i need to give one pre a bit of gain, i read somewhere if you shunt the bottom resistor (R2)with a high value capacitor it will increase the gain or with a small cap it will increase the treble. I've tried 100uf and 47 uf but it kills the output any idea what value cap to add? Do i need to alter the input resistor to suit the impediance for each pickup? or is there another way of increasing the gain of one pre?
http://tinypic.com/view/?pic=ipz7me
Thanks
Vikki(uk) :grin:

Hi Vikki,

It shall be ok to "capacitor bypass" the source to ground, to get maximum gain in the FET-transistor.

But you must check some small thing, to be sure that the FET work with right bias.

Adjust the source 2,2k resistor (with the bypass capacitor connected) so you measure the half of the supply voltage on the drain. (the FET output)

Alternative, with a 1 kHz sine tone connected to the FET input, and a ocilloscope on the FET output, and the real load connected. (volume and tone control)
Turn up the 1 kHz tone on FET input just before clipp on the output, adjust the source 2,2k resistor so the sine curve on the scope start to clipp fairly equal in upper and lower edge in same time.

If there are some dc-component on the input, you must have a capacitor in serial with the FET input. (approx. 0,1 uF)

If the FET output will drive a volume potentiometer, you can disconnect the 51k resistor after the output capacitor, to minimize the drive load.

--Bo

Thanks Bo
I was thinking of putting a 10k preset in place of the 2k2 so i had some adjustment. I was thinking of using perhaps a 10uf to boost the gain a bit does that sound about right? The preamp does sound quite nice considering how simple the circuit is, i don't think there will be any dc from the guitar pu's? But i may be wrong there.
Vikki(uk) :grin:

there is always the trick of paralleling another stage with it.. :thumb:

The 201 has an IDSS spec of .2ma to 1ma

http://www.till.com/articles/GuitarPreamp/ looks like the article it came from. The circuit as drawn operates at .44ma (drain at 6V so a 3V drop across the 6.8K) so you will want a fet with a higher IDSS than .44ma. .7ma to 1ma might be a good spot.

Connect the gate to the source source to ground, insert a DMM set for current in series with a 9V battery to the drain that is you IDSS for 9V supply like in the circuit.

.44ma across 2.2K is about .968V this is you input headroom something people seem to forget about. (.44ma X 2.2K) you can go lower with the mag I would think.

Most piezo can give a higher voltage out than 1V P to P when lightly loaded if what I have read is true this might be why the 10meg was dropped to 3meg in the article

I would think a source follower to a BJT EF or gain stage might be a better circuit, or even a charge amp setup like the km84 when using a Piezo pickup

Now back to the mag pickup check you bias points if they are good 4V to 6V at the drain. A cap across the source R should help with the gain if the gain is to much a pot in series with the bypass cap can act as a gain control Also like Bo posted you can get rid of the 51K if the output goes to a volume pot.

Yes, a bypass cap should increase the gain, and 10uF would be a reasonable value in parallel with R2. And no, I wouldn't mess with the input resistor; piezos basically like to see as high an impedance as possible.

The bad news: many FETs will produce distortion, including high-frequency intermodulation distortion, when driven by a high impedance source such as a piezo. Might be less if you make the FET a source follower.

In my experience the only thing that can begin to tame a piezo pickup is a vacuum tube. I've heard Brad Sarno's "Black Box", basically just a 12AU7 and a power supply, make a piezo pickup sound like real music, where it sounded like the usual piezo squack before. Very nice gadget -- there's a review in the new Tape Op, but they don't talk a lot about using it with piezos.

Peace,
Paul

A bootstrapped drain would help the distortion, which is probably mostly due to input capacitance variation with signal.

Thanks all.
I was trying to overcome the effect of the 20foot cables i have running from the guitar to the main dual channel preamp, i was hoping i could make this unit in a small box attached to my belt or guitar strap. I run this through a dual Ridge Farms valve di into my calrec eq to the main mixer. I've seen a circuit somewhere that has a second silicon transistor after the fet, i'll dig it out to see what you think.
Vikki(uk) :thumb:

Vikki;

Build the Till preamp inside of a gutiar plug.
I have seen it done on some website.
Use surface mount parts.
The circuit configured for lower gain will drive
the cable capacitance with no problem.
So your tone does not roll off with the long cable.
You will need to build a small box to add power to the center
conductor and a nice cap to block that power on the output.
If you do this build plenty of spares as they always break while you
are on stage.

This?

http://www.till.com/articles/PreampCable/index.html

note someone patented something like it after it was on the net!

Vikki

Does it have to be 9V?

If you change the piezo one to a charge amp you can reduce its gain

Piezos in parallel one side grounded other side to the gate of the fet circuit as drawn add the bypass cap to the source R

gate to ground via a >10 meg resistor.

Here is the difference. A good cap (polystyrene etc)from drain to gate.

If I understand correctly

the gain will be the open loop gain of the fet gain stage in parallel with the piezo cap value / the added drain to gate cap and fet drain to gate miller cap added together.

This would be a small value cap maybe even a gimmick cap (wire wrapped together to make a small value cap) Some wirewrap wire is teflon coated.

If you look at the km84 circuit fragmant in the "microphone" PDF at neumann you will see the math and the basic charge amp circuit.

I think the ceramic cap setting on a sencore LC102 might measure the value of the piezos.

This might give you what you want within the output headroom of a 9V powered circuit

> if you shunt the bottom resistor (R2)with a high value capacitor it will increase the gain or with a small cap it will increase the treble. I've tried 100uf and 47 uf but it kills the output

That should work. And not upset bias. So the real question is: why doesn't it? Wrong polarity? Leaky cap? Something not wired as in the drawing?

As Gus implies, different FETs of the same part-number can have wildly different performance.

It should also work, for the low-output magnetic, to drop R2. Gain could almost double if the drain is now sitting at 6V-7V and you can get it down to 3V or so. Put in a pot, diddle and listen.

There is of course the bad old TL074. Get any gain you could need, hi-Z input, lo-Z output, and will run quite a while on a 9V battery.

> I wouldn't mess with the input resistor; piezos basically like to see as high an impedance as possible.

I believe Vikki is asking about the gain for the magnetic pickup. These can drive 10K on up, with somewhat different sound for different load, but she won't care about "sound" until it is loud enough to hear. 3M is not wrong, when it is loud enff I might try 47K and see what the change is.

> If there are some dc-component on the input

If either the magnetic or piezo pickups are making DC, a miracle has happened.

Sample-to-sample variation in JFETs is terrible, as PRR mentions. The schematic you linked will have unpredictable gain which will tend to be on the low side. But you can bias the JFETs in a manner which makes the circuit relatively insensitive to the device variations. Here's a booster, based on J201 JFETs, which I designed and built recently:

Schematic and Pics

The entire booster could be built in a box small enough to fit on a guitar strap. Or just the input stage could be used for a boost (inverting, of course) of about 20dB. If only the input stage is used, add a tiedown resistor of 100K or higher between the output side of C3 and common.

This circuit would work well for a piezo pickup as well, if you increased the value of R2.

Just found this variation on that circuit?

Vikki (UK) :thumb:

Vikki, today's avatar is wonderful but big. Use this:


http://tinypic.com/ivh01x.jpg

Hey, you make a cute elf!

I forgot to mention, if the input stage is used on its own, the gain can be reduced to 6dB by eliminating C2. Or C2 can be connected to a tap on R5 if some intermediate value of gain is required. Or, leave C2 in place and use an audio-taper 10K pot in place of the drain resistor, R6. This would act as a volume control. The high side of the pot goes to the FET, the low side to +9V, and the output is taken off the wiper. I generally avoid running direct current through signal pots, but the voltage and current in this case are low enough that it shouldn't be a problem.

Aaaah that's better.....!
I'll have a bit of time tomorrow to do a bit of experimenting with the circuit. Perhaps the caps i used were just way too high, ill try some lower ones.
Vikki)uk) :grin:

[quote author="Vikki"]Thanks all.
I was trying to overcome the effect of the 20foot cables i have running from the guitar to the main dual channel preamp, i was hoping i could make this unit in a small box attached to my belt or guitar strap. [/quote]

Theres's a cheesball way to do thi s too if ya don'w want to be making cables. You can use those cheap 1/4" --> MXLR and 1/4" --> FXLR. I tries this with a strat into a deluxe reverb through a 50 (or was is 100?) foot cable and I really couldn't here much diference between that setup and a 12' cable .

Kiira

I must explain my statment, because both Vikki and PRR react.

Vikki wrote:
I don't think there will be any dc from the guitar pu's? But i may be wrong there.

PRR wrote:
> If there are some dc-component on the input<
If either the magnetic or piezo pickups are making DC, a miracle has happened.

Of course there are not any dc-voltage from a passive magnetic or piezo pickup, but what I mean was if Vikky want to lab with a external oscillator to set the "symetrical clip point" and be sure so the oscillator don`t distract the bias point on the FET transistor. ( for example if the oscillator have a leaking output capacitor)

So, now I have say that, and maybe I can sleep better this night :wink:

BTW, here is a very simple FET high-Z buffer with a minimum of components, that I design for a other application here on the forum.
This is very easy to get work with mostly N-channel JFET transistors you can find.
It is ok to change the "gate bias resistor" (1 Mohm) to 10 Mohm or more, if you want a higher input impedance.

It is a unity gain buffer, so it have no gain, but it is perfect for impedance converter after piezo pickups.

--Bo

I recently stuck a couple piezos on my mutant mando-guitar and came up with this -



The FET and 10M resisistor go on a piece of double-stick foam tape on the top (silvery) side of the piezo (this helps kill the resonance of the piezo as well.) The top side is then encapsulated in epoxy. A length of skinny shielded cable goes from the drain and source of the FET to a connector mounted on the guitar body.

The cool thing about this is that the output is a fairly low impedance, instead of the hum-prone high impedance piezo output.

I'll post pics and sound clips if anyone is interested.