Introducing the "G3" High Gain Fet Distortion Circuit ...

[MartyMart]

http://aronnelson.com/gallery/Martys-layouts-and-photos/G3_Fet_Distortion?full=1

Schem - rough !!  :
http://aronnelson.com/gallery/Martys-layouts-and-photos/G3_schematic

This circuit is based on the "G5" valve amp project, using an EF86 pentode for the first stage
and 12AX7 ( EC83 ) for stages two and three.

The tone is full with quite "thick" mids, think "TS-9 mid hump" and you're not far off.
It sounds nice and bright too, with very good range from totally clean, with boost using
the volume pot, to almost "Metal" grind with gain at max, and many softer "blues" sounds
inbetween.
I recommend sticking to the parts list, however, for a slightly brighter and thinner tone, you
can change the three 22n caps to 10n.
For even more "sparkle" simply remove the 15k/2n2 network at the back of the circuit and go
directly from the final 10n cap to volume pot lug 3 .
This may work well for people with "darker" sounding amps, but not recommended for bright
Fender amps.

I hope someone will build and enjoy this, it took a lot of "thinking" and "tweaking" to get here
and I'm very happy with the amount of gain from only three stages.
Keep leads short and neat to avoid oscillations, mine were all around 6cm and produced no
noise whatsoever.

Cheers and "TURN THAT UP" !
Marty.

[brett]

Hi Marty
Very interesting design.
Lots of time spent on it, judging by the fine-tuned parts.

You mention taking out the heavy low-pass filter at the ouput.  (15k and 2n2, fc=5kHz)
Another option is to lessen the effect:  10k and 2n2 gives fc=7.5 kHz, and this often works for me (taming harsh fuzzes etc)

keep up the good work

[Torchy]

Nice one Marty, have to try this out ...

[John Lyons]

Marty
Can you post a schematic, or did I miss it? I'd like to check it out.

Thanks

John

[MartyMart]

Thanks Brett, my "maths" was off ... I was aiming for around 8k rolloff !
15k/1n should get close

An interesting addition ( which I've not seen used on these fet circuits before ) is to use an
spdt switch, just before the roll off network to add a pair of back to back red 3mm LED's to ground.

They "just" light up, so are no where near saturated, but it adds a bit of extra "grit" at full bore and
doesn't seem to effect the lower gain range.

I may also try Brian Marshalls idea of positive feedback, from the gain lug 3 node through perhaps
a 2M2/6n8 cap to output ... unless it may cause some motorboating ?

MM.

[KerryF]

Looks nice.  Maybe I will try that soon.

[MartyMart]

Marty
Can you post a schematic, or did I miss it? I'd like to check it out.

Thanks

John

John, I dont have any software for that ( mac ) and tried to "hack" one together with photoshop, but kept
running into problems cutting/pasting !!
I'll hand draw one and post up a photo ... OK ?

MM.

[petemoore]

  Vero is cool...I haven't gotten any yet.
  For ''looking inside the 'clock' to try to tell how/why/what makes it 'tick," Schematics make for the best reading.
 
 

[Eric H]

Marty, where is the G5 amp project located?

-Eric

[MartyMart]

Schematic is at top post, very rough but readable !

G5 is on a french site, posted recently ... lost the link ! sorry

MM

[fxdfxd]

Hi

Projet G5 is hosted here : http://projetg5.com/

But this is not the original "projet G5" schematics (see mine here : http://f.defaut.free.fr/drupal/), which own only one 12ax7.
The schematics provided comes from the 'projet G1', that is hosted on the same website

-fx

[Xavier]

Hey Marty, thanks for the new design !!!

Mid hump, high gain, almost metal.......your sound description sounds quite close to the BSIABII to my ears. Is it very different? Does it have the chug-chug also ?

very interesting

[petemoore]

  Readable enough..Thanks MM @!!
 Does the 'Gullivered' [Gate ties to V+/Gnd.] Mosfet transistor bias right up with those values or is some fiddling with biasing necessary to optimize the mosfets function?

[MartyMart]

Hey Marty, thanks for the new design !!!
Mid hump, high gain, almost metal.......your sound description sounds quite close to the BSIABII to my ears. Is it very different? Does it have the chug-chug also ?
very interesting

No, I wouldn't say that it's "BSIAB - like" at all, has a "fuller" tone ( not "Fulltone" !! )
.... more like an LP through a big amp/cab at half power, with 15inch cones

 Readable enough..Thanks MM @!!
 Does the 'Gullivered' [Gate ties to V+/Gnd.] Mosfet transistor bias right up with those values or is some fiddling with biasing necessary to optimize the mosfets function?

No problems, the pair of 1m5's 18k drain/2k2 seemed to keep my 20 BS170's happy , they all biased in around
4.5 to 4.9 v ... mine reads exactly :
D  4.72v
G  2.04v
S  0.44v

Battery 8.82v

MM.

[petemoore]

  Wire-Right 'n Fireup's always nice, not that I wouldn't do the biasing measures if necessary, it's just nice to know what your'e getting into so plans may better accomodate the biasing arrangements.

[John Lyons]

Thanks for poting the schematic Marty

I think I can make it all out in you schematic.
John

[Marcos - Munky]

Soundclips?

[slacker]

If anyone's interested I drew up a schematic from Marty's sketch

[MartyMart]

Thanks Slacker, that's great !
Anyone know a "mac" platform program to easily create schems ??

Edit : Follow Brett's advice about the high end roll off network, use either a 10k/2n2 (2,200pf)
or 15k/1n (1000pf) for best results.

MM.

[Gus]

I see a few things.

One the input Z looks to be on the low side for a fet input, you might want this I don't know what the design was for.  You have the mosfet set for max gain with the source resistor bypassed with the cap.  Now the 1.5meg from drain to gate Reduces the input Z then you have the 1meg input R in parrallel.

Two why 560K?  Fets have low gate leakage so it should not add to much noise.

Three check GEOFEX maybe use a lin pot with a gate to ground R to get the curve you want with a fet input and ......

J201s have fairly tight specs for a FET IMO. Now look at where you have the trim pot(take it out and measure it) and note the drain voltage.  This gives you I drain.  the source R is part of the input headroom measure the voltage S to ground.  Then measure the IDSS of the j201. gate connected to source, source to ground postive of a 9V battery to a DMM set to ma to the drain the current read is your IDSS.

Now if the J201s are from the same batch and you measure the IDSS and resistors for what sounds good for different fets you should be able to reporduce the circuit with the bias points you like from this data by knowing the IDSS of each fet

I picked IDSS because it is easy to measure a 9V battery a meter and some jumpers

Now lets take a well known microphones the neumann u87 KM84 etc the fet in the older ones had the drain R fixed and a target voltage for the drain.  The source R was adjusted picked to get this Drain voltage.  So maybe lets guess the trim is set to 50K(just a guess) make the drain a 47K or 51K and then use a say a 5 K trim pot for the source then pick a fixed resistor close to that value.

Just giving you some feedback about things you might have fun adjusting

I very good site is Vishay semiconductors section on FETs look for the app notes good stuff.

Fets are fun but harder to design with than say a BJT.