A BSIAB without JFETs?

[DSV]

Hi all!

I am currently developing Infinity², an updated version of my guitar preamp. As you may know, the distortion section is based on the BSIAB, and uses JFETs.

The idea I have is to replace these with another "technology", be it opamps or other transistors. My goal is:
- to avoid JFET inconsistency;
- to have a circuit whose distortion is independent from the PSU voltage.
Of course, it has got to sound as good as the BSIAB gain stages.

I have started with an opamp approach, by:
- measuring the FETs' Vth and Idss;
- simulating everything on LTSpice (and hoping the results are accurate) with an FFT analyisis
- reproducing the gain and the filtering of each stage
- implementing it with opamps.

The problem is that, even by trying all sort of clipping (LEDs, Si and Ge diodes, transistors) in soft- and hard- configurations, the sound and the dynamic response do not satisfy me at all. Furthermore, the circuit is much more prone to oscillation.

Are there other ways I should explore in order to reach my goal, or should I drop this idea off and just implement the circuit with selected JFETs and a voltage regulator for the PSU?

Thanks,

Dan

[Thomeeque]

Try tubes 

[oddist]

Hi,
Did you try MOSFETs as clipping devices?  How about opamp configuration - inverting or non-inverting?  With cascaded stages
where the signal gets large, inverting is better, IMO.  I think with non-inverting, you get a component of the input coming through
at unity gain, and it seems to do weird things when input is large. 
Also, keep impedances low and multiple stages with lower gain in each.
Best wishes!

[dschwartz]

try cmos inverters..good sound, cheap, six inverters, few parts..a little bit of noise, but not that much.

[DSV]

Try tubes 

Hi,
Did you try MOSFETs as clipping devices?  How about opamp configuration - inverting or non-inverting?  With cascaded stages
where the signal gets large, inverting is better, IMO.  I think with non-inverting, you get a component of the input coming through
at unity gain, and it seems to do weird things when input is large. 
Also, keep impedances low and multiple stages with lower gain in each.
Best wishes!

I tried all sort of configurations. The best performance in terms of noise and oscillation stability was indeed obtained by cascaded inverted stages. IMHO the problem really lies in the clipping method (I tried MOSFETs too, in most of the configurations suggested here http://www.muzique.com/lab/zenmos.htm, also in the feedback loop, but to no avail). Impedances were pretty low, too. I guess opamps are not the way to go for me.

try cmos inverters..good sound, cheap, six inverters, few parts..a little bit of noise, but not that much.

I should try them. Howerever I do not know much about their implementation. Is there a tutorial explaining how they work, and in particular:
- how the gain is defined;
- how the inter-stage filtering is performed (unless it is only passive filtering)
- things like biasing, PSU etc. etc. etc.

[oddist]

One last thing with opamps - are you filtering in the feedback loop (LPF)?  Helps reduce oscillation and tames fizziness.
I've had pretty good luck with cascaded opamps both with LED and MOSFET clippers, but using three or four stages with
lower gains.
Good luck in your search!

[DSV]

Yes, I used caps from 22pF to 5nF. About the clipping, I liked the sharp attack and sound of the LEDs, but not the transition between distortion and non clipping, too brutal. Si diodes, on the other hand, had a better "decay", but lacked sparkle and didn't interact well with the other controls. Ge diodes were too mushy and undefined. MOSFET clippers were better, but still not satisfying enough.

I think opamps + clipper diodes are great for overdrive and distortion stompboxes, as  the amp plays an important role when pushed into overdrive or by "smoothing" and shaping the signal. However for preamps their distortion is too raw IMHO. Andrew Barta from Tech21 has managed to get extremely good results only with opamps, but they do not have clippers (so PSU voltage is important, as well as the opamp type ... and I am too lazy to go and "audit" opamps  )

I checked a few CMOS inverter circuits, and, if I understand well, they are implemented like inverting opamps, both for gains and filtering. On the other hand their maximum gain is roughly 100 for each stage.

Am I right?

[dschwartz]

check the double D from runoffgroove, red lama, hot tubes, and hot harmonics, and analyze them. cmos inverters sound great for guitar, very "ampy" sound

[WGTP]

http://www.diystompboxes.com/analogalchemy/sch/vulcan.html

This is Joe Davisson's Vulcan which is different than anything else I have seen around here.  It's a great distortion/OD and 1 or 2 stages should work.  It uses 2N5089 transistors and rocks.  Do a search for more info...   

This is the Blackfire.  Doug Hammond used a stage for the output of the Highway 89.   

http://www.diystompboxes.com/analogalchemy/sch/blackfire.gif