Onboard JFET Preamp question

[guidoilieff]

You should have a noticeable increase in gain if you build that 1 FET simple gain stage you posted (in Reply #8). Not a ton, but some.    If you don't, that means something is wrong with the build or the component.  Or the pinout...

The way to go about this is read the "debugging" thread, but really - what voltages are present on the 3 device pins?  Mostly these problems have to do with wrong biasing for the device, solved by a trimmer at the drain. 

Hint: If you want BIG gain, use a different circuit like the AMZ Mosfet Boost.

The volts are:
Power:8.80v
Gate: 0v
Source: 4.46v
Drain: 6v

I realized no matter how big the resistor I use for the Gate, when I probe it on the breadboard with the dmm it reads 0 ohms, when i remove it the reading is fine. Even with a pot.

[Frank_NH]

Put a 10 uF capacitor in parallel with the source resistor - that should bump up the gain quite a bit.

[GibsonGM]

Put a 10 uF capacitor in parallel with the source resistor - that should bump up the gain quite a bit.

Right...if you look at the AMZ Mosfet boost, that is exactly what is done (with a pot to create a gain control). The source bypass cap will allow more gain by reducing regenerative feedback between gate & source.

So, if it's just a gain thing, I'd try that.   You don't have any tone controls in there yet, right?

BTW, the drain seems a bit high to me, I might consider upping the drain resistor until I got about 4.5V, but I could be just being picky. If the sound isn't distorted, you're at least close.

[Frank_NH]

After reading through this thread, I'm not sure what JFET you're using or what source, drain resistors you currently have.  Let me know and I'll calculate the gains for you using my speadsheet  .  However, regardless of your setup, a large source bypass cap (10 uF or even 22 uF) should increase the gain noticeably.  Let us know what you find.

[guidoilieff]

After reading through this thread, I'm not sure what JFET you're using or what source, drain resistors you currently have.  Let me know and I'll calculate the gains for you using my speadsheet  .  However, regardless of your setup, a large source bypass cap (10 uF or even 22 uF) should increase the gain noticeably.  Let us know what you find.

I have used trimpots in the S,G and D and found that 1k5 or 2k2 are the right ones for the drain.

I still have no idea why the G resistor is there. I doesn't change anything if I remove it or just put another value, unless i shorted it to ground.


The 10uf capacitor had tremendous effect on the boost, thanks for that! But still have my doubts about that gate resistor.



Now voltages are:
G 0      S 3.9     D 6.3

[guidoilieff]

Put a 10 uF capacitor in parallel with the source resistor - that should bump up the gain quite a bit.

Right...if you look at the AMZ Mosfet boost, that is exactly what is done (with a pot to create a gain control). The source bypass cap will allow more gain by reducing regenerative feedback between gate & source.

So, if it's just a gain thing, I'd try that.   You don't have any tone controls in there yet, right?

BTW, the drain seems a bit high to me, I might consider upping the drain resistor until I got about 4.5V, but I could be just being picky. If the sound isn't distorted, you're at least close.

Its not a gain thing, I just want to know I'm doing it right. I expect the volume to me the same as the input or higher, I don't care by how much.


No, I haven't made any tone/gain controls but I've made a small amp like that with a lm358 http://www.generalguitargadgets.com/pdf/ggg_little_gem_sc.pdf

I'm doing this because there are some projects I want to do that require biasing...
(and because jfets are the closest thing I'll ever be to a valve I guess...)

[Jdansti]

You might consider throwing together a Tillman and see how you like it. You could probably build one on perf in in less than 20 minutes.  I have have one on two of my guitars and while the gain isn't suoer high as noted by others, it's enough for me to use as a booster when switching between lead and rhythm. I just used a 2P2T toggle to bypass it and there's no pop. As mentioned earlier, the stereo jack on the guitar is used to control the power.

[GibsonGM]

I understand, guido...but like I said, and now JDansti says...maybe put together a Tillman AND/OR AMZ mosfet boost.  They are GUARANTEED to work if your fet is ok and you put the parts in the right places, and are solid foundations for future projects (building blocks).  In fact, they really ARE building blocks, based a lot on manufacturer's designs and so on.

It should take you 1/2 hour or less to put one together. 

http://www.muzique.com/schem/mosfet.htm     Note R1, R2 etc at top of schem - these are 'replacing' your gate resistor to properly bias the FET with a voltage reference, standard stuff.   

If you don't use a BS170, you'd use a trim at the drain to get to about 1/2 V+ right at the drain.  I often use 2N7000 with great results this way. 

I use these all the time, never fail and have TONS of gain but fairly transparent sound.   Also on the site are some FET buffers etc. which would go well with this.   And you can always set the gain with a trimmer or resistor (set and forget) so no need for a pot...very versatile.

[Frank_NH]

"I still have no idea why the G resistor is there. I doesn't change anything if I remove it or just put another value, unless i shorted it to ground."

The 1M - 3M gate resistor is there to provide a reference to ground since the input (AC) signal is biased at (or referenced to) 0V.  With no resistor, there may be a possibility that the Vg bias could drift.  The generalized arrangement is shown in the figure below.  If R1 >> R2, then R1 becomes an open circuit and the voltage divider provides a reference of 0V.

BTW - what JFET are you using.  BF245C?  Can't see in the photos.  Thanks.

[PRR]

> Power:8.80v
> Source: 4.46v
> Drain: 6v

No good. Source and drain are only 1.5V apart, which is not enough for many JFETs to come alive.

The 4.46V seems awful high, if source resistor is few-K as you say.

Make drain resistor 200 ohms. Now diddle the source resistor to get the source to about half of that 4.4V, say 2.0V to 2.5V. Now increase drain resistor to get 5V-6V at drain. Is that different? Better?

Your voltages suggest you DO have a working pinout.

> still have my doubts about that gate resistor.

Just do it and forget it. 1Meg is plenty high to have "no" effect on the audio (you have shown this). *If* you come off a guitar or some stompboxes, it is not strictly essential, there's a DC path through the guitar or a bleed resistor. But some stomps lack any bleed resistor. Then, without a ~~1Meg gate to ground, the gate bias will "float" to any darn DC value it wants to. This will almost certainly be well away from the best amplification bias. So you have a booster which boosts in many situations but not in others, maybe different from moment to moment. This resistor (and other details) is simple defensive interfacing.

[GibsonGM]

^ Do that.  Best advice you'll get all day, guido...   

[guidoilieff]

> Power:8.80v
> Source: 4.46v
> Drain: 6v

No good. Source and drain are only 1.5V apart, which is not enough for many JFETs to come alive.

The 4.46V seems awful high, if source resistor is few-K as you say.

Make drain resistor 200 ohms. Now diddle the source resistor to get the source to about half of that 4.4V, say 2.0V to 2.5V. Now increase drain resistor to get 5V-6V at drain. Is that different? Better?

Your voltages suggest you DO have a working pinout.

> still have my doubts about that gate resistor.

Just do it and forget it. 1Meg is plenty high to have "no" effect on the audio (you have shown this). *If* you come off a guitar or some stompboxes, it is not strictly essential, there's a DC path through the guitar or a bleed resistor. But some stomps lack any bleed resistor. Then, without a ~~1Meg gate to ground, the gate bias will "float" to any darn DC value it wants to. This will almost certainly be well away from the best amplification bias. So you have a booster which boosts in many situations but not in others, maybe different from moment to moment. This resistor (and other details) is simple defensive interfacing.

I made the Drain resistor 7v and the Source resistor 4v (less than that the gain drops) and now is louder than connecting the bass directly to the pc speakers, but I used a 13v power source because I left my 9v transformer at a friends house.

[Jdansti]

You can test it with a 9v battery. Actually, you'll probably use a 9v battery in the guitar and it should last a very long time.

[ggedamed]

I think that the main problem is that you're using a BF245C (!!!) in a 9V circuit. That's why your source voltage is approaching 5V and that's why your circuit is attenuating. Kreuzer's circuit is made for BF245A which has a pinch voltage Vp of less than -2V (mine have it less than -1.5V). BF245C has a much larger Vp and it's not easy to bias it in this circuit, as your measurements show.
Use this opportunity to build the JFET measurement jig from the Fetzer Valve page. You'll need it plenty anyway if you intend to build JFET circuits.

Here are two common-source JFET circuit calculators I found on the interwebz: http://www.guitarscience.net/calcs/cs.htm and http://www.diydave.be/tools/fetcalc/index_eng.html. Maybe they are of some use to someone.

[amptramp]

Go to:

www.geofex.com

and look under projects at the JFET matching fixture.  It is an op amp circuit that sets the non-inverting input of an op amp at half the supply voltage.  The inverting input is driven by the source of a JFET with the source driving a 10K resistor to ground, the drain at supply voltage and the gate connected to the op amp output.  The gate voltage is compared to the half supply voltage by an external meter.

[Gus]

you can build something with a NPN transistor
you can design for the gain you want and get a good high input resistance

[guidoilieff]

I think that the main problem is that you're using a BF245C (!!!) in a 9V circuit. That's why your source voltage is approaching 5V and that's why your circuit is attenuating. Kreuzer's circuit is made for BF245A which has a pinch voltage Vp of less than -2V (mine have it less than -1.5V). BF245C has a much larger Vp and it's not easy to bias it in this circuit, as your measurements show.
Use this opportunity to build the JFET measurement jig from the Fetzer Valve page. You'll need it plenty anyway if you intend to build JFET circuits.

Here are two common-source JFET circuit calculators I found on the interwebz: http://www.guitarscience.net/calcs/cs.htm and http://www.diydave.be/tools/fetcalc/index_eng.html. Maybe they are of some use to someone.

I think thats the problem, thanks!