Phuncgnosis & buffer?

[GibsonGM]

I was wondering if anyone's added an input buffer to the Phuncgnosis, and what it did to improve its performance?  Is it a 'worth doing' mod? 

Mine works nicely, altho it would be nice to have it 'recover' a little faster for more 'waaaw' sound when picking rapidly. That's probably got more to do with my pickups' output than anything, tho (Les Paul).

Thanks

[mac]

I recently posted about some mods I made to the phuncgnosis,

http://www.diystompboxes.com/smfforum/index.php?topic=49246.0

The idea of an input buffer seem interesting. I should try it.



mac

[GibsonGM]

I read about it somewhere once; someone suggested I do it, but I don't know why (rather, how it would improve the effect).  I have mine set up true bypass, so tone sucking isn't an issue. I'm just curious if it improves things, and don't want to tear the board out to put a buffer in if it won't make a nice improvement

[mac]

breadboard a buffer + phuncgnosis, or just a buffer before the one you built. This will give an idea.



mac

[Mark Hammer]

Here's the schem: http://www.geocities.com/tpe123/folkurban/fuzz/phuncgnosis.gif

Think of the Sensitivity control as an input resistor to ground from the .002uf cap.  If that terminating resistor is a high enough value, then it doesn't load down the input too much. 

Look here - http://www.muzique.com/schem/doctor-q.gif - and you'll see that the Dr Q uses a much lower resistance sensitivity pot (100k).  Not only that but when the switch is set to "bass", that sensitivity pot is placed in parallel with a 47k resistor to ground.  Yikes!  Talk about loading.  This is why the FET buffer that Jack Orman stuck on the Dr Quack made such a noteworthy improvement to the sound, because that pot/resistor combo was now being fed by a lower output impedance signal so that sticking 100k and 47k in parallel would not result in much loading of the signal.

In the case of the Phuncgnosis, there IS nothing comparable to the 47k ground resistor on the DrQ, and the sensitivity control acts like a 1M resistor to ground.  In view of that, my sense is that if you feed it with just about anything other than a direct guitar signal (and even there it's a long shot), you probably won't hear that much benefit of a buffer because there isn't all that much loading down of the input to begin with, the way there is in a Dr Q.

So, if you plan on sticking it later in your signal chain, you can skip the buffer.  If you want the flexibility of being able to go direct from guitar to Phuncgnosis, then a buffer MIGHT add a little insurance.

[GibsonGM]

Thanks, Mark...I see said the blind man.  I'm not experiencing much loading (well, when it's on it does load the signal, but that's part of its character).  I run mine before my MXR Dist+, and it works fine as long as you don't use too much gain and if it's tweaked in just right via sensitivity pot.  Doesn't seem like it would be worth mutilating the board to put a buffer in...I might try it some time when I have one breadboarded for something else, just to see what it changes.   I was thinking about 'more sensitivity' (more rapid recovery), but I use a LP, and the pickups are pretty consistent with their output - can't ask for miracles out of such a small and simple filter!   

[Mark Hammer]

Well heck.  If "undersignalitis" is your problem, then just kick up the value of the feedback resistor in U1a from 33k to 39k or even 47k.  That ought to permit some responsiveness even at lower sensitivity settings.

[GibsonGM]

Hmm, thanks Mark, that might be a good one to try! Maybe a 100k pot.  What happens is that when I get going on leads, I end up cranking the phunc to its max treble setting and fragging the front row - so I play softer, but then it never really goes "up and down" unless I slam it & totally stop playing for a second (slow recovery time, you can hear it whooshing back down).   I think my LP pups are so even/compressed in their output that they don't give the ckt much variation...it's not the phunc's fault, lol.  I get very little 'room' to mess around with the 1M sensitivity pot.   

Perhaps playing with that feedback resistor will get me what I want....which is to kick in the full sweep when I play fairly hard, but have it float somewhere juuuust above quiescent when I'm laying back, so I can 'waahwow' some reggae stuff...I really like the ckt's sound, I just need to get more control over it thru tweaking.   

[mac]

I was not satisfy with the sensitivity pot, nor with the recovery time. I found that replacing the only 330k with a 1Meg-log pot improves the dry/wet mix.



mac

[GibsonGM]

Great, more info!    I like those Tim E. circuits - they are simple & give bang for the buck.  Just always looking for ways to improve things.  I'll look into the 1M pot.  I love how the thing sounds on a clean channel...like any env. filter, you have to pay close attention to what you're putting out as far as signal level.  Adding gain changes things - that's what I'm trying to 'dial in' better if I can...how does adding some dry signal back in improve things? 

[Mark Hammer]

Hmm, thanks Mark, that might be a good one to try! Maybe a 100k pot.  What happens is that when I get going on leads, I end up cranking the phunc to its max treble setting and fragging the front row - so I play softer, but then it never really goes "up and down" unless I slam it & totally stop playing for a second (slow recovery time, you can hear it whooshing back down).   I think my LP pups are so even/compressed in their output that they don't give the ckt much variation...it's not the phunc's fault, lol.  I get very little 'room' to mess around with the 1M sensitivity pot.   

Perhaps playing with that feedback resistor will get me what I want....which is to kick in the full sweep when I play fairly hard, but have it float somewhere juuuust above quiescent when I'm laying back, so I can 'waahwow' some reggae stuff...I really like the ckt's sound, I just need to get more control over it thru tweaking. 

I have to confess that I don't understand the workings of diodes as voltage-controlled (or is it current-controlled?) paths as much as I ought to, but the LEDs in the Phuncgnosis are functioning in a manner soooooooort of like the transistors or LDRs to ground found in other similar types of swept bandpass filters.  The different number of LEDs going to V+ for downward sweep, vs the single LED tied to ground for upward sweep only serves to convince me of that.

Now, if the starting point and range of the sweep is influenced by the combined forward voltage of the LEDs, then it stands to reason that it can be modified by the addition (or deletion)  of diodes to the chain in each instance.  So, if my understanding is correct, the downward sweep can be made to start a little higher by replacing one LED with a silicon diode (1N4001, 1N4148, etc.), or made to start a little lower by adding a silicon diode to the existing chain.  Similarly, the starting point of the sweep in the upward setting can by altered by the use of some diode complement different than what is shown (e.g., two 1N4001 in series, which results in about 30% less forward voltage than 1 LED, or a green LED which has a higher forward voltage than red).

The Sensitivity control, in tandem with the gain of that op-amp, serves to determine how much contrast there will be between the start and max-sweep centre-frequencies of the filter.  The filter caps chosen will determine where in the spectrum that amount of frequency-sweep will occur.  The diodes will determine how that amount of intended sweep translates into actual sweep.

What players tend to want is something that reliably triggers on every pluck, so that they don't have to whack at the strings, without sweeping so far that it interferes with the "meat" or nuance of the notes plucked.  In essence they want a sort of "compression" on the sweep so that they get what they want all the time without risking excessive and disruptive sweeps.  That is achieved by having the right combination of sensitivity at the one end of the circuit, and responsiveness at the other end.  If it were an LDR-based swept filter, the quivalent strategy would be to stick a trimpot in parallel with the LDR and another in series with the LDR, so that the minimum and maximum resistance values attainable could be set and effectively "limit" what the sweep would be.

Try monkeying around with the diodes, though.  I think there may be something to it.  Wouldn't it be cute if some combination of different coloured LEDs serendipitously gave you the sweep response you liked best?   You could then mount them on the panel of the pedal with black bezels and watch the colours change as you strum.

[mac]

The sensitivity pot at the input is a kind of dry/wet mix. Adding the second pot let fine tune what the input pot does. I got the idea from this schematic that uses a 470K instead of the only 330K in tim's schem. Why not to put a 1MA pot? Or tweak some caps here and there?

http://matsumin.gozaru.jp/diy/jisaku_fx/Funky_Wah/Funky_Wah_sch_11.BMP

http://matsumin.gozaru.jp/diy/jisaku_fx/Funky_Wah/Funky_Wah_Layout.jpg

As you can see, the input impedance is no longer 1M but less than 25K in this schem, so an input buffer may be necessary, as mark said about dr quack.

Another thing I would change is the fixed 100k resistor with a 100kA pot, ie a vol pot. I guess this may help if it is connected to the drive chn.

Did you try replacing the 33K resistor at the feedback loop with a 50 - 100k pot?

This thing sound good on the clean chn of my laney lc30II (well in fact everything in the clean chn sound good with this amp), but I also like it in the drive chn, nice crunchy autowah.



mac

[mac]

Mark, I replaced the led at the up sweep with one 1n4148, but at the same time I have to change the 330K with a 100kA pot. It seems that the 330K works with 3 leds down and 1 led up at the same time.

Gibson, looking at another Tim's schem I noted that he used different cap values at the feedback loop, ie, photon filter. Instead of the two 681 caps, a 102 and a 471 maybe?



mac

[GibsonGM]

Good stuff, guys! Great breakdown, Mark, thanks for taking the time to get into the inner workings.
I played with caps when I had this boarded, mac...don't recall exactly what sounded best, but they were closer to your suggested ranges rather than the 680pf.  I settled on only using the single big red LED for up sweep, and didn't bother with the down sweep mode.   It's now in a PVC endcap, but I have to remove it to shield it; a good time to add a 1M pot for the 330K resistor and sub in some higher values for the 33k. I might also try the 1N4148 like you did, Mac.   Other colors of LED's didn't do too much in my breadboard tweaking.

I'm assuming the max lows are set by the top cap (680pf) and the 5M feedback resistor at U1b?? Looks like a lowpass op amp filter to me.  And that playing with the bottom cap will regulate how treble the filter can go?

What I'm searching for in playing with this is best explained above by Mark...not having to fight the thing, esp. when switching pickups.  Might not be do-able, but I could maybe get close...