Three More Shaka Questions

[StickMan]

First off, things are going well with this, if a little slow.

I'm putting together a mishmash of ideas from the Shaka I, II, III, IV, and V schematics.  For instance, I'm putting in the switchable feedback clipping diodes, from one version, but dropping the buffer.  I've got the MOSFET shunt to ground diodes in there, but I'm also making it switchable with a pair of LEDs.  I'm also toying with replacing the trim pot in front of the booster with an off board pot to allow the drive to be pushed up high enough to get the LEDs or MOSFETs clipping, but then not drive the booster -- or to turn off all of the diode clipping and push the booster hard (this would be like a BSIAB2, with an opamp for the first booster).

Finally, I'm going to add a more exciting tone control to the circuit.  Here's question 1:

I've been looking for a schematic for an active Baxandall tone stack with a single supply.  I can't find any.  I figure I should just connect the initial input to that section to Vref using a resistor -- but how big?  Also, I assume that I should just connect the +ve of the opamp directly to vref.

Question 2:

I've got the shunt to ground diodes switchable.  Nothing, MOSFETs (as in Shaka V) or LEDs.  For some reason, the LED setting is MUCH louder than the other two settings, even with no shunting.  How can this be?  Surely it can only drop the volume?

Question 3:

My feedback loop looks like the Shaka V with a 1N914 diode pair switchable in.  In other words, I haven't added the second capacitor or change the component values to match the schemo that had the switchable diodes circuits in it.  When I switch in the diodes, the gain out of the opamp drops considerably.  I can here it, but I can also see that the LEDs shunting to ground are glowing more dimly, so I'm losing some of that clipping.  How can I stop that happening?  I tried adding a resistor to that circuit loop, but I didn't see (or hear much difference).  I've pondered adding a connection to ground through a resistor, but that would mean changing the switch to DPDT, which I'd rather not do.  Is there a better way?

Here's the schematic with the switchable feedback diodes:



And while I'm at it, why the hell does the feedback loop connect back to ground through the .47uF cap instead of Vref?


thanks,

dave

[Doug_H]

Wow, I wish Aron would take this one down...

This is a very old and bad mod idea on my part. The .001u feedback cap actually attenuates the signal in the passband... Not good... I never did anything with this.

If you want to do the Shaka, I suggest starting with the Shaka 3 and modding from there.

Doug

[StickMan]

Doug,

Don't worry, I'm really just using the idea of putting the switchable "hot", "warm", "cool" diode feedback circuits in there.  I like the idea of being able to mix and match the various diode clipping circuits.

The starting place for me was actually the Shaka V schematic.

But you didn't answer any of my questions!!!!!


dave.

[StickMan]

Sorry, but I'm bumping this since I'd really like to hear some suggestions.

dave.

[WGTP]

http://aronnelson.com/gallery/WGTP/MosfetDiode?full=1

This is an alternate way to wire the Mosfets.  Aron's method uses the diodes built into the Mosfets and has a similiar clipping threshold to SI diodes, although it sounds slightly different.  The alternate method uses the actual Mosfets and has a much higher clipping threshold similiar to LED's.  It might be interesting to make it switchable between the 2 methods of using a Mosfet.

The higher clipping threshold allows more signal thru to the next stage, resulting in more volume which is why your LED's are louder.  They also distort less.   

[StickMan]

That's the only way I've seen the MOSFET's used.

The LED shunt clipping is actually louder than no shunt clipping at all!  That's what I don't understand.  I can understand it being louder than the MOSFETs.

Here's the schematic for the Shaka 5, which is the base I'm working from:

[WGTP]

Look closely, there are 2 different ways to wire the mosfets.

Sounds like some wiring problems if the LED's are louder than nothing.   

[StickMan]

I notice two things.  First, the MOSFETs on your diagram are P-channel, which doesn't match the IRF510 & equiv transistors.  Secondly, the diodes appear to be reverse biased with regards to the MOSFETs.

Am I missing something here??


dave.

[WGTP]

I used the same Mosfets Aron used, but notice that I redrew the picture from his schematic.  The Ge diodes are reversed which forces the signal thru the mosfet, rather than the built in diode.  I have this arrangment on a breadboard right now in the feedback loop and it has a nice thich smooth sounding distortion to it.  Here is the link to the thread about it if your interested.   

http://www.elixant.com/~stompbox/smfforum/index.php?topic=38581.0

[aron]

I have this arrangment on a breadboard right now in the feedback loop and it has a nice thich smooth sounding distortion to it.

Interesting. With what pedal? I remember really liking it much more the way I had the diodes oriented.

Hmmmm

Aron

[StickMan]

When I first looked at the circuit with the MOSFETs, I figured that the coupling of the Gate meant that you'd have a resonably large voltage on the Gate (at least some of the time).  My thinking was that this would essentially boil down to using the MOSFET as a variable resistor, shunted to ground -- a bit like the simple compressor (using an LM386) that some posted here a few weeks back.

I couldn't reconcile how that would work with the other MOSFET, however, with the gate essentially at ground.

My understanding now is that the MOSFETs are used backwards to the ordinary current flow, which basically makes the Gate voltage moot, and then the diode aspect kicks in.  Is this correct?

I looked through the other thread linked from here, and from what I could see, there were two ways to use the MOSFETs referred to under the following terms:

1.  Normal body diode mode.
2.  Soft knee mode.

I'm having trouble seeing how the two are actually put together, though.  Is the only difference the direction of the mated germanium diode?

thanks
dave

[WGTP]

As you can see from the previous post there was a lot of discussion and some of us had difficulty understanding it.  Thanks to R.G. and John Green for their explainations.  That's all I know about it.

I wanted to minimize the op-amp distortion when trying the Mosfet's (which have a high clipping threshold), so I put the Mosfets in the feedback loop of an inverting stage with a 100K-470K resistor and drove it with a non-inverting op amp stage with the gain adjusted to only slightly clip on the initial attack.  The op-amp is rail-to-rail for max clean signal.  2K resistor between stages.

Aron, your correct the body diode does have a nice crisp sounding distortion, with a much lower threshold, around .7v IIFC.  That's why I said a switch to use both methods would be cool.   

Here it is in a slightly different Muff Fuzz type circuit.