Twin-T notched JFET preamp with bass in mind

[onboard]

This started out with the front end of a ’57 Fender Bassman in mind. I couldn’t get the cathode follower stage down, ditched the tone stack, and ended up just tinkering around with biasing cascaded jfet gain stages. It ended up more of it’s own circuit than an emulation.

You could use any number of filtering stages to get fixed/variable tone shaping. I just happened to be playing with Twin-T type notches and stuck it in there. If my math is right, the notch is centered around 340Hz  and the bass roll-off at the end of the circuit is adjustable from 32 â€" 160Hz. The low end coming through this circuit is deep, even with 0.1uf input/output caps.

Bypassing my bass amp’s preamp and running this straight to my power amp return sounded fantastic to me. No distortion, although you could get clipping in a number of ways if that’s what you want. The notch sounds right and doesn’t seem to be too wide. Overall, I like the voicing. There seems to be some compression, too â€" though that could be my imagination.

If you wanted to switch the notch in and out of the circuit without huge volume differences, a 1M resistor in it’s place is fairly equivalent to the filter’s attenuation. Inefficient, but it works. Active filtering would overcome the insertion loss, but I think you would quilckly end up with an entirely different circuit â€" which wouldn’t necessarily be a bad thing.

[B Tremblay]

Very nice looking circuit!

You could convert the twin-T to a bridged-T, which would let you use two less components.

[vortex]

Hey cool. Nice work and thanks for sharing.
I too am working on a FET based Bass preamp. Is there a reason you bailed on the tone stack?
I am working on the premise that if you are copying a tube amp tone stack design with FETs that the cap values ( resistors too?)should be reduced to a tenth of their tube amp equivalents due to lower voltage. I read that somewhere but am not sure that it is correct. Any thoughts?

[onboard]

Thanks B! Props to the RunoffGroove JFET emus that got my wheels turning. For some reason the twin-t seemed to respond better than a bridged, but I might not have been working the numbers out right.

A notch filter seemed more elegant than the typical tone stacks for just a mid scoop. If you're going for the different response curves of an original, definately go with the stock tone stack. Get the Duncan Tone Stack Calculator if you don't already have it!

Got any links vortex? The Prof. Tweed page at ROG mentions Doug Hammond's technique, but I didn't turn up any specific info. My go at it was simply twiddling pots for biasing the drains, keeping track of the response and voltages.  

The '57 tweed Bassman's first cathode resistor is 820 ohms, the 1.5k are pulled from the Prof. Tweed. (that's what I had in the drawer...)

The input 12AX7's cathode resistor bypass cap is 220uf, so I think you're right on with there with the 1/10th value, 22uf. For tone stacks, I'ld go with the stock values.

[B Tremblay]

For some reason the twin-t seemed to respond better than a bridged, but I might not have been working the numbers out right.

That may be caused by the lower Q of the notch produced by your twin-T.

For example, you used a 47k and 10nF for half of the twin-T, which results in a 338Hz corner frequency.  The other filter uses a 27k and 22n, producing a 268Hz corner.  The closer that these two frequencies are, the higher the Q.

Swapping the 27k for a 22k would move that corner to 328Hz, making the notch narrower.  It should sound similar to the bridged-T you tried.

[moosapotamus]

Nice looking circuit, onboard.

Swapping the 27k for a 22k would move that corner to 328Hz, making the notch narrower.

Cool. Alternatively, could you maybe replace that resistor (R6) with something like an 18K in series with a 10K pot and have a variable Q control?

~ Charlie

[B Tremblay]

Alternatively, could you maybe replace that resistor (R6) with something like an 18K in series with a 10K pot and have a variable Q control?

You sure can!  However, it may seem subtle when playing clean.  That's the sort of adjustment that is much more apparent when using a fuzz.

[moosapotamus]

Ain't nothing wrong with a little fuzz!
The notch frequency might have some impact on how noticeable the variable Q would be too, yes/no?
So, maybe also try some different values for C3, or a dual pot for R4/R5?  

Thanks
~ Charlie

[B Tremblay]

Sebastian Tepper (stm) wrote the definitive article for applying bridged-T filters to DIY-FX.  I've found it to be absolutely indispensable:

http://diystompboxes.com/sboxforum/viewtopic.php?t=25788

[moosapotamus]

   

Thanks!
~ Charlie

[Dragonfly]

Sebastian Tepper (stm) wrote the definitive article for applying bridged-T filters to DIY-FX.  I've found it to be absolutely indispensable:

http://diystompboxes.com/sboxforum/viewtopic.php?t=25788

WOW !!!!  you arent kidding....absolutely FANTASTIC article/thread !!!!!

andy

[Ben N]

Where's the pic?

[onboard]

Sebastian Tepper (stm) wrote the definitive article for applying bridged-T filters to DIY-FX.  I've found it to be absolutely indispensable

That's *the* thread alright. It's archived in my pc...

The values I used are slightly off from the R, 1/2R, C, 2C, twin-t ratio- namely the resistors. I'll play with how narrow I can get the notch and still have that "scooped" curve sound - if it's gets too narrow the response changes more to a, well, notch  

So is it just my imagination, or would this circuit cause some compression? Either way, there's a certain something about the note attack and bloom that feels good. In light of the tube/jfet similarities only going so far, I'll still take a chance and say it's tubey.