Squidward Mod/Help

I have a Squidward PCB that I was about to populate, but I was going to try it using the values of the octave section from the Parentheses. I got some Ge diodes from Small Bear and in the tiny batch I ordered I found one matching pair at 270mV. The only thing I used differently from the Parentheses is a 2N3906 (original Squidward part) rather than the 2N5807 for the signal splitting transistor. I tried the diodes both ways and both sound good - one direction has more bottom end and I think I like that.

Is there a good way to lower the output volume, and possibly increases the sustain? As it is right now there's a big jump in volume and it's very "fuzzy". Since I plan on using this alone at times that huge jump in volume and clipping doesn't suit my needs as much as it would if it were cleaner sounding. For reference here's the relevant part of the schematic.


Screen Shot 2020-04-14 at 4.24.47 PM.png
 
I didn't say I wanted "variable" anything. I'm just trying to see if there's a way to mellow out the signal so it's not so attack heavy. If I wanted an octave fuzz I would have built an octave fuzz. This is the 'modifications' forum, right? Thanks for the help, I guess.
 
Less clipping, maybe change the diodes for something else with higher fv. But then you'll get even more volume, sooo...

You could always stick a pot before the output - lug 3 the end of the circuit, lug 2 to a new output, lug 1 to ground - the most basic volume control you see in 99% of pedals. Not sure if that's what you want though, or if it'd work on this circuit.

Maybe the fuzzy fox would have been better suited since the tentacle is typically paired with something after it. But hey, try it and see. Nothing ventured, nothing gained. I have a hunch changing the clipping will remove the octave effect but have a go.
 
So for typical distortion circuits, higher forward voltage on the diodes means higher clipping threshold, so less clipping (cleaner). This circuit is different. It’s using the diodes as part of a rectifier circuit (turning all the negative parts of your signal positive). This rectification doubles the frequency, but not in a clean way, even if the rectification was perfect. Think of a rectified sine wave. It’s at double the frequency of the sine wave, but doesn’t look like a sine wave any more.

Anyway, my understanding is not super strong on this circuit, but I think lower forward voltage should give closer to perfect rectification, and a bit cleaner sound (though it’ll never be really clean). I’ve heard matching the diodes Vf is important too for this one.

For fun I made an fv-1 patch that just took the absolute value of the input (perfectly rectified). It sounded not too different from my green ringer build. Of course the fv-1 is also capable of a clean polyphonic octave up or down.
 
I don't see how that circuit could even work unless it relies on Ge diode leakage? What am I missing?
 
I did some biasing math thanks to a recent DIY Guitar Pedals video. If you were to take the Ringer section out of the Life Pedal, because of the voltage divider values being used at Q1 you're only getting about 0.82V at the base, then the way that's biased it's only sending a small current of 88uA, but amplifying it to hit the PNP base with 7.2V. On the original Ringer/Tentacle circuit those figures look more like 2.0V at Q1base and a much larger current of 218uA only amplified to 5V at Q2base. I feel like the signal was already clipping a bit coming out of Q2 using the Life Pedal values - then again I'm using a lower HFE PNP (2306) than the specified 5807, so that might have something to do with it.

One of the reasons I was trying to use the Life Pedal values was because I don't have any 160K's on hand (R2 of the voltage divider in the Ringer/Tentacle) so I did some math with what I had on hand. Settled on 470K for R1 and 100K for R2 to get 1.914V at the base of Q1 (realistically it measures a little over 1.5V). I'm also using a 2N3904 for Q1 but I may try a 5088/89 to see if I like that better. Biased that to give me 4.5-5V at the Collector/Q2 Base. Still very dirty sounding. Swapped out the values in the rectifier part of the circuit for the lower value resistors from the Ringer - left in the 100nF caps. Tried a few different matching diode pairs after the low Vf Ge's - 1N5817 was a little better. 1N4148's measuring 0.603 (about twice the Ge's I have) sound best. Definitely getting more sustain and less dirt. It's subtle, but when you're holding a chord you notice. I also have nothing better to do right now!

For the blend part of the circuit I notice the FET buffer was cutting some low end. The 1M/100nF R-C filter has a corner freq of only 1.6Hz, so that's not the culprit. I looked up the original Buff n' Blend circuit that uses an MPF102 like I have (when you see it next to the Parentheses schematic you'll see it). That has a 1M from rails biasing the Gate that the Life Pedal octave section doesn't have, so I put that in - got my low end back. It also felt a little quieter than when bypassed. The Life Pedal is biasing the FET's Source at 10k, so I swapped that for the original value of 3.3K. The result, I get my low end back and unison volume. The blend circuit turned all the way clean actually sounds like my clean sound. The octave of the Life Pedal wasn't intended to be used on it's own, so these changes were likely on purpose on EQD's part. Since the Squidward PCB is so small I'll put the Buff N Blend on it's own little vero daughter board, and I should be able to fit it all into a 1590A.

Squidward_Schemo new Values.jpg

Screen Shot 2020-04-16 at 5.18.20 PM.png
 
As far as output goes I'd think that experimenting with various audio taper volume pots would get you where you want to be. You won't necessarily reduce the volume but at least you can get the volume where you want it with the volume pot at noon. Or if you didn't want to add a volume pot just put a trimmer on the board somewhere to act as a volume.
 
Thanks for the discussion you provided about the circuit and your measurements. Interesting stuff.
 
I did some biasing math thanks to a recent DIY Guitar Pedals video. If you were to take the Ringer section out of the Life Pedal, because of the voltage divider values being used at Q1 you're only getting about 0.82V at the base, then the way that's biased it's only sending a small current of 88uA, but amplifying it to hit the PNP base with 7.2V. On the original Ringer/Tentacle circuit those figures look more like 2.0V at Q1base and a much larger current of 218uA only amplified to 5V at Q2base. I feel like the signal was already clipping a bit coming out of Q2 using the Life Pedal values - then again I'm using a lower HFE PNP (2306) than the specified 5807, so that might have something to do with it.

One of the reasons I was trying to use the Life Pedal values was because I don't have any 160K's on hand (R2 of the voltage divider in the Ringer/Tentacle) so I did some math with what I had on hand. Settled on 470K for R1 and 100K for R2 to get 1.914V at the base of Q1 (realistically it measures a little over 1.5V). I'm also using a 2N3904 for Q1 but I may try a 5088/89 to see if I like that better. Biased that to give me 4.5-5V at the Collector/Q2 Base. Still very dirty sounding. Swapped out the values in the rectifier part of the circuit for the lower value resistors from the Ringer - left in the 100nF caps. Tried a few different matching diode pairs after the low Vf Ge's - 1N5817 was a little better. 1N4148's measuring 0.603 (about twice the Ge's I have) sound best. Definitely getting more sustain and less dirt. It's subtle, but when you're holding a chord you notice. I also have nothing better to do right now!

For the blend part of the circuit I notice the FET buffer was cutting some low end. The 1M/100nF R-C filter has a corner freq of only 1.6Hz, so that's not the culprit. I looked up the original Buff n' Blend circuit that uses an MPF102 like I have (when you see it next to the Parentheses schematic you'll see it). That has a 1M from rails biasing the Gate that the Life Pedal octave section doesn't have, so I put that in - got my low end back. It also felt a little quieter than when bypassed. The Life Pedal is biasing the FET's Source at 10k, so I swapped that for the original value of 3.3K. The result, I get my low end back and unison volume. The blend circuit turned all the way clean actually sounds like my clean sound. The octave of the Life Pedal wasn't intended to be used on it's own, so these changes were likely on purpose on EQD's part. Since the Squidward PCB is so small I'll put the Buff N Blend on it's own little vero daughter board, and I should be able to fit it all into a 1590A.

View attachment 4042

View attachment 4043
You may find the 2N4401 Better suited, Less Noise than 2N3904 & Less Gain than 2N5088
 
The Tentacle (Squidward) is biased hot; the transistors are closer to saturation than cutoff. The Life Pedal (Parenthesis) octave is biased cold; the transistors are closer to cutoff than saturation. The Tentacle is a straight-up copy of the Dan Armstrong Green Ringer. It has a gain just under 3x. Unless you have very hot pickups or a booster in front, you won't get it to saturate. The Life octave has a gain of about 10x. Q2 and Q3 can be driven into cutoff without too much effort. If you want sustain, you need gain and limiting. Saturation, cutoff, or both will provide the limiting.

Both octave circuits are insensitive to HFE, provided that the diodes are installed with the cathodes pointing toward the base of the last transistor. Anything with an HFE >100 will work fine. You don't need high HFE, but it does no harm. The resistors in these octave circuits set the gain and biasing, not the transistor HFE. Unless the HFE is <20 in which case all bets are off.

Some JFETs source followers (Q1 in the Life & Buff N Blend) will bias ok with the gate at ground potential, provided they have a high enough Vp. Vp for PF5102 is -0.7V to -1.6V. That means the source voltage will be between 0.7V and 1.6V at idle and the most drain current we can expect is 160uA. Biasing the gate up would actually help in the Life Pedal. Vp for MPF102 can be as high as -8.0V. One that hot would not be much use in a pedal that runs on 9V. Biasing the gate voltage up (as in the BNB) will increase drain current. However, some MPF102s will not work in that circuit because if their Vp is >4.5V they will saturate at idle and no signal will get thru. Both circuits require JFETs that are hand-picked because some of the ones that are in-spec will work poorly, or not at all.
 
The Tentacle (Squidward) is biased hot; the transistors are closer to saturation than cutoff. The Life Pedal (Parenthesis) octave is biased cold; the transistors are closer to cutoff than saturation. The Tentacle is a straight-up copy of the Dan Armstrong Green Ringer. It has a gain just under 3x. Unless you have very hot pickups or a booster in front, you won't get it to saturate. The Life octave has a gain of about 10x. Q2 and Q3 can be driven into cutoff without too much effort. If you want sustain, you need gain and limiting. Saturation, cutoff, or both will provide the limiting.

Both octave circuits are insensitive to HFE, provided that the diodes are installed with the cathodes pointing toward the base of the last transistor. Anything with an HFE >100 will work fine. You don't need high HFE, but it does no harm. The resistors in these octave circuits set the gain and biasing, not the transistor HFE. Unless the HFE is <20 in which case all bets are off.

Some JFETs source followers (Q1 in the Life & Buff N Blend) will bias ok with the gate at ground potential, provided they have a high enough Vp. Vp for PF5102 is -0.7V to -1.6V. That means the source voltage will be between 0.7V and 1.6V at idle and the most drain current we can expect is 160uA. Biasing the gate up would actually help in the Life Pedal. Vp for MPF102 can be as high as -8.0V. One that hot would not be much use in a pedal that runs on 9V. Biasing the gate voltage up (as in the BNB) will increase drain current. However, some MPF102s will not work in that circuit because if their Vp is >4.5V they will saturate at idle and no signal will get thru. Both circuits require JFETs that are hand-picked because some of the ones that are in-spec will work poorly, or not at all.
Thanks Chuck. Sounds like you know what you’re talking about after all. I’m only starting to read about FETs because of this little project. What you’re saying about high output pickups is relevant to this build for me since I’m barely driving the circuit on the bridge pickup of my Jazzmaster. That’s what I’m trying to perfect before boxing anything up. I might go back to the Life Pedal values for the reasons you pointed out. I’m starting to think the big boost in volume is not a bad thing an that’s where the blend control comes in.
 
I have a Squidward PCB that I was about to populate, but I was going to try it using the values of the octave section from the Parentheses. I got some Ge diodes from Small Bear and in the tiny batch I ordered I found one matching pair at 270mV. The only thing I used differently from the Parentheses is a 2N3906 (original Squidward part) rather than the 2N5807 for the signal splitting transistor. I tried the diodes both ways and both sound good - one direction has more bottom end and I think I like that.

Is there a good way to lower the output volume, and possibly increases the sustain? As it is right now there's a big jump in volume and it's very "fuzzy". Since I plan on using this alone at times that huge jump in volume and clipping doesn't suit my needs as much as it would if it were cleaner sounding. For reference here's the relevant part of the schematic.


View attachment 4025
I built a “electro distortion” style boost (w/+w/o clipping) after the squid board in the same enclosure and it mellows and exaggerates the squid pretty well. Using correct input and output caps for the boost is pretty key.
 
Well I'm glad DigitalZombie's thread was raised from the dead.

I learned a few things and will be experimenting with my own Green Ringer > Boost project (or was it Boost > GR?).
 
Back
Top