Transistor roulette

[Kipper4]

Tonight i made this simple preamp.

http://www.electroschematics.com/5300/fet-pre-amp/

I didnt have the perscribed transistors but i built it anyway i figured something in my tranny stock box would fit the bill.
I'm astonished after all the things i've built that i only have 1 type of PNP transistor (besides germaniums)
The only ones i have are BC327's i cant remember what project i originally got them for but they've come in handy a few times now.
so thats that i used for T2.
This is where the roulette comes in.
The jfet T1
i tried a 2n5457 it was splatty (let me qauntify Splatty. when i play guitar softly it could bearly the signal bearly passes, when i strum hard its too loud and very dynamic.)
i treid J201 Splatty
MPF102 IIRC same thing
then i discovered another little gem in my box (intended for the bass balls project)
The amazing (amazing because it makes the circuitwork) 2N5485

This thing sounds good enough for a simple 10 parts preamp. that quite frankly i could have made smaller had i planned better.
I'm going to experiment with 9v power and see if it still works.

Anyways I have a question.
How is the gain determined and worked out in this circuit please.
I wish to understand more about transistor circuits.
I'm getting into them a bit more now.
Thanks

Anybody else like to play transistor Roulette?

[pinkjimiphoton]

i do, i do!!!!

i have zero respect, and will try anything, even npn's in pnp circuits. sometimes 180 degrees rotated.

find some cool stuff breaking things.

[Kipper4]

Nice one Jimi.

It works on 9v too. i'm not sure it has as much headroom.
I might try it with a mic tommorow since he describes it as suitable for mic use too.

The reason i'm intrested in the gain is if i change some resistor (not sure which ones affect this) How will it affect the gain of the circuit and what would be a good choice?
what would overload the circuit?
what would give a full distorted saturated sound?
what would give unity gain sound?
That kinda thing.

[CodeMonk]

Not so much anymore.
But when I first started building pedal back in the 80's I didn't know shit.
I would put random parts in random places just to see what would happen.
And when I say random, I mean like put resistors, caps, diodes, etc. from one trace to another without even following where each trace went.
I had a reverb (built from a ratshack\tandy kit) and got it sounding like a damn space ship.
 
And I broke some stuff in the process too.

But pinkijimi is the king of that kind of stuff.

[pinkjimiphoton]

yahsuh,  i F a LOT of $#l7 up!!!! lol

[CodeMonk]

Should have added a smilie to that
(Better late than never).

I actually meant you are the king of doing random stuff.
Although there is your latest experiences with a DOD pedal 

[R.G.]

i tried a 2n5457 it was splatty (let me qauntify Splatty. when i play guitar softly it could bearly the signal bearly passes, when i strum hard its too loud and very dynamic.)
i treid J201 Splatty

This "splatty" is a near-perfect description of misbiasing. The thing is biased into a non-amplifying state, and loud signals push it over into amplifying.

MPF102 IIRC same thing
then i discovered another little gem in my box (intended for the bass balls project)
The amazing (amazing because it makes the circuitwork) 2N5485

Mother Nature is whispering into your ear that JFETs have widely varying VGS and Idss factors. The 2N5485 is my near-favorite because the Vgs clusters around 2-3V. The J201 is about 0.1 to 0.8V, and the MPF102 can be as much as 8-9V.  What you're doing is picking a JFET to make the whole thing bias correctly.

How is the gain determined and worked out in this circuit please.

The JFET amplifies, and feeds directly into the PNP, which also amplifies. There is a lot of gain. This gain is tamed by feedback. R5 and R4 divide the output signal from the PNP down and feed it into the source of the JFET. This is in a phase that opposes the input signal, so it's negative feedback. We can use the same kind of reasoning we use on opamps. If the signal at the JFET source is the same size as the input, the output would be zero, as there is no net voltage across the Vgs. So it's less than that. It will settle down to where the divided-down signal at the junction of R5 and R4 are just almost equal to the signal on the JFET gate (for AC signals, all of this). So the resulting gain will approach (R4+R5)/R4, or about 23. It will be somewhat less than that, because of the non-infinite gain of the JFETxPNP amplifier.

The reason i'm intrested in the gain is if i change some resistor (not sure which ones affect this) How will it affect the gain of the circuit and what would be a good choice?
what would overload the circuit?
what would give a full distorted saturated sound?
what would give unity gain sound?

This is an mid-60s kind of circuit, before opamps were common and cheap. It has some problems, because the same resistors that change the gain also change the DC operating point. It can be tough to separate out the DC conditions to get the biggest swing (i.e. most "headroom" and least "overload", which are near opposites) and at the same time select a gain.

Changing R4 and R5 changes the gain, but it also changes the DC value on the JFET source, and so the JFET bias point at the same time. Just when you get it right with one device, changing to another JFET changes it all again.

I have textbooks that will address these issues, but the subject takes pages of explanation. In brief, for best output swing, you want the collector of the PNP to be sitting at half the supply voltage. That makes the DC value at the top of R5 be (V+/2)*[R5/(R4+R5)]. The value of voltage at the source of the JFET is that DC value on F5, plus some fraction of the JFET's Vgsoff, just enough to get enough current to flow through the PNP to hold its collector at half V+, and to hold the JFET's source at the right point. They all interact.

[Kipper4]

I dont think i could have hoped for a better technical explanation.
Thanks R.G. its appreciated.

[tca]

It is funny you mention this circuit. I have a copy of the ARRL electronics data book (1976) where a similar topology appears: a FET and a PNP. Have to breadboard it and see how it sounds.

Cheers.

[Kipper4]

Your welcome.
It's quite nice. It'd do for a simple onboard booster too.