trying to find reading material on jfet preamp design, cold clipping stages..

[Doctor Algernop]

I'm trying to learn how to understand the different stages used in a typical FET preamp and different ways of implementing them in ways that make sense. I'm starting with zero understanding of how to construct a JFET gain stage.   I've been modifying the do boogey and none more black as well as many others but I'd like to try and understand  as much as I can about it's

[antonis]

I'm starting with zero understanding of how to construct a JFET gain stage.

I'm afraid of you have to first understand the "classic" BJT analog, then pass to Ebers-Moll equation model (all those nasty  things about emitter intrinsic resistance, small signal variations, ΔV_be e.t.c.) to get an, more or less, in depth undestanding of the "notorious" transconductance g_m (notorious for FETs because of their wide spread and remarkably lower values compared with BJTs..) and earn some headache with all those I_DSS, V_pinch-off, I_D/V_GS curves, e.t.c.

Or just consider FETs as voltage controled current devices and search for anything relative in AMZ & GEOFEX.. 

[Doctor Algernop]

thank you very much for the advice

I've been building just under a year now and I'm finally starting to make sense of a lot of things that seemed like reading mandarin to me at first. I've been reading thru a lot of stuff about MOSFETs and JFETs but your right, I really need to start with the basics such a BJT's as there a lot easier to grasp. I find all the pedals I really love of fet based tho so I tried to RUSH(great band) ahead for sure

I find opamp distortions are great if you want something that sounds nice and tight without a boost or tube screamer. a FET based amp sim and a tube screamer are a match made in heaven

[tubegeek]

thank you very much for the advice

I've been building just under a year now and I'm finally starting to make sense of a lot of things that seemed like reading mandarin to me at first. I've been reading thru a lot of stuff about MOSFETs and JFETs but your right, I really need to start with the basics such a BJT's as there a lot easier to grasp.

Great books, cheap used:

Malvino, especially "transistor circuit approximations," and later books which move into ICs etc.

http://www.diystompboxes.com/smfforum/index.php?topic=110832.msg1018065#msg1018065

[merlinb]

You can get great tube-like distortion from JFETs by using them as source followers and overdriving them with an opamp or two. That way you get the predictable control of gain that an opamp affords, and the relatively trouble-free biasing of source followers. Just a thought.

[Frank_NH]

Hi Merlin,

Could you expand on your idea?  I'm assuming you would put a JFET buffer like the one below at the output of a tube-screamer style gain stage with soft clipping diodes?  Or would the input limits of the JFET do the clipping?  I may have to try this out in LTSpice...  

[antonis]

I think that Merlin told you about a source follower driven by an already distorted signal...

You'll take advantage of more predictable gain of op-amp and easy follower biasing.. 


But, IMHO, a source follower should be used in cases of relatively high signal source impedance (Z>50k or so) because of FETs low tranconductance (compared to typical BJT values..) which results in relatively high Zout in a follower configuration..
(it just a matter of comparison between expressions for Id vs Vgs & Ic vs Vbe..)

FET follower Zout is practically independable of signal source impedance and it's just 1/gm where BJT follower Z_out is Z_source/h_fe + 1/gm

For typical resistor & current values and for a Z_source of, say 5k, a BJT follower has a Zout of 10 times less than a FET one - I don't like to make calculations here so take my word for that.. 

Op-amp ouput is considered very low so I can't find any benefit of using FET follower other than bias easiness..

BUT, breadboard behaviour may fight book analysis... 

[merlinb]

Could you expand on your idea?  I'm assuming you would put a JFET buffer like the one below at the output of a tube-screamer style gain stage with soft clipping diodes?  Or would the input limits of the JFET do the clipping?

Yes a buffer like that, but no clipping diodes. I would use the opamps for clean gain and EQ, and let the JFETs do the clipping by overdriving them (add a resistor in series with the gate to protect against gate overload). They clip quite softly, very much like a valve, and you can control the symmetry by adjusting the bias or swapping in different FETs.

[antonis]

 

Sorry Merlin, I understood totally different arangement ...