Feral Feline
Well-known member
I looked at that Pellucid, ‘twould be grand to have a Boneyard Edition!
This Week on the Breadboard - The Engineer's Middle Finger Compressor
- Thread starter Chuck D. Bones
- Start date
Have you stop struggling with the other You Know What compressor before you start this one?I found another clever compressor design. Hmmmmmm.
https://www.diystompboxes.com/smfforum/index.php?topic=122568.0
Chuck D. Bones
Circuit Wizard
For the moment, yes, but I will come back to the Engineer's Middle Finger.
The problem I was having with the EMF was that at high RATIO settings, the attack was too pronounced. My thinking was that the peak detector was not tracking a rapid rise in the envelope quickly enough. The Engineer's Thumb uses a half-wave rectifier in the peak detector circuit and tracks only the negative peaks in the guitar signal. This can cause a 1/2-cycle delay in the peak detector. Other compressors, such as the Delegate & General Tsao, employ a full-wave rectifier that tracks both peaks. The Pellucid also uses a full-wave rectifier in the peak detector. I wanted to see how much difference that made. The Pellucid's full-wave rectifier does improve the peak tracking and the attack is smoother, however it's my observation that aggressive picking with the bridge pickup creates an attack so abrupt that there is still somewhat of a "pop" on the first cycle of the note. I'm starting to think that I had an unrealistic expectation of how well a compressor can control such an abrupt attack. The other thing I'm looking at is the difference in release between a feed-forward circuit like the ET and a feedback circuit like the Pellucid. The ET measures the signal volume coming into the compressor and uses that to adjust the gain. The Pellucid measures the signal volume coming out of the compressor and uses that to adjust the gain. Both gain control methods have advantages and disadvantages.
The problem I was having with the EMF was that at high RATIO settings, the attack was too pronounced. My thinking was that the peak detector was not tracking a rapid rise in the envelope quickly enough. The Engineer's Thumb uses a half-wave rectifier in the peak detector circuit and tracks only the negative peaks in the guitar signal. This can cause a 1/2-cycle delay in the peak detector. Other compressors, such as the Delegate & General Tsao, employ a full-wave rectifier that tracks both peaks. The Pellucid also uses a full-wave rectifier in the peak detector. I wanted to see how much difference that made. The Pellucid's full-wave rectifier does improve the peak tracking and the attack is smoother, however it's my observation that aggressive picking with the bridge pickup creates an attack so abrupt that there is still somewhat of a "pop" on the first cycle of the note. I'm starting to think that I had an unrealistic expectation of how well a compressor can control such an abrupt attack. The other thing I'm looking at is the difference in release between a feed-forward circuit like the ET and a feedback circuit like the Pellucid. The ET measures the signal volume coming into the compressor and uses that to adjust the gain. The Pellucid measures the signal volume coming out of the compressor and uses that to adjust the gain. Both gain control methods have advantages and disadvantages.
Chuck D. Bones
Circuit Wizard
Good question. Transistor matching is not critical in this circuit.
Chuck D. Bones
Circuit Wizard
Nice, but now it has 4 knobs.
fig
Village Idiot
One is invisible...it's all the rage.
Last edited:
Chuck D. Bones
Circuit Wizard
OK, I've taken this circuit about as far as I can. I'll see if any of the tweaks I did here can be applied to the Engineer's Thumb board. Latest changes:
1. Put one of the LM13700's internal buffers (U2D) to work in the gain control part of the circuit.
2. Added a RELEASE control. In the CS-3, it's called ATTACK, but it's the same control. It governs how quickly the gain rises as the note decays.
3. Put the LM13700's linearizing diodes to work (pin 2). I have no idea why no one else has done this. Since the LM13700's inputs are connected in parallel, one set of linearizing diodes serves both sets of inputs. Adjusted the value of R9 accordingly.
4. Put clipping diodes (D4 & D5) on the output to limit the initial overshoot. These diodes only conduct during the first few milliseconds of the note attack. They don't add any distortion because U1A is hitting the rails during that time.
5. Put in C3 to minimize the DC offset on U1A.
6. Reduced C2 to preserve the brightness at high SUSTAIN settings.
I tried TL072 and NE5532. Both work well. NE5532 might be a little quieter. The only noise I can hear is coming from the guitar. Don't max out the SUSTAIN with single-coil pickups unless you like listening to hum.
With the linearizing diodes in place, the only source of distortion (after the first few ms of the note attack) comes from ripple on the gain control voltage (+ side of C8). A few mV is all it takes. The distortion is mostly 3rd harmonic. The only time you hear it is during the tail end of the note decay when the gain is high and there is low freq content. If you stay off of the low E-string, it's barely noticeable. The Boss CS-3 does the same thing. It's a consequence of using a VCA (voltage-controlled amplifier) for fast response time. Turning RELEASE up above noon makes the distortion go away, but at the expense of a slower response time. Optical compressors have the advantage of the LDR's slower response filtering out the control ripple. It's a trade-off of distortion vs. response time. If you're running a dirt pedal or a cranked amp, then that tiny bit of distortion doesn't really matter.
The control loop is very stable and does not overshoot or ring. It is normal to hear "breathing" when you play chords because the peak detector senses everything and the low freq content is the loudest. Any beating between notes or feedback will cause the volume to go up and down and the gain control will try to track that. Turning up the BRIGHT control helps a little by shifting the emphasis away from the lower frequencies. Turning RELEASE up also helps by slowing down the response time. It's a balancing act.
1. Put one of the LM13700's internal buffers (U2D) to work in the gain control part of the circuit.
2. Added a RELEASE control. In the CS-3, it's called ATTACK, but it's the same control. It governs how quickly the gain rises as the note decays.
3. Put the LM13700's linearizing diodes to work (pin 2). I have no idea why no one else has done this. Since the LM13700's inputs are connected in parallel, one set of linearizing diodes serves both sets of inputs. Adjusted the value of R9 accordingly.
4. Put clipping diodes (D4 & D5) on the output to limit the initial overshoot. These diodes only conduct during the first few milliseconds of the note attack. They don't add any distortion because U1A is hitting the rails during that time.
5. Put in C3 to minimize the DC offset on U1A.
6. Reduced C2 to preserve the brightness at high SUSTAIN settings.
I tried TL072 and NE5532. Both work well. NE5532 might be a little quieter. The only noise I can hear is coming from the guitar. Don't max out the SUSTAIN with single-coil pickups unless you like listening to hum.
With the linearizing diodes in place, the only source of distortion (after the first few ms of the note attack) comes from ripple on the gain control voltage (+ side of C8). A few mV is all it takes. The distortion is mostly 3rd harmonic. The only time you hear it is during the tail end of the note decay when the gain is high and there is low freq content. If you stay off of the low E-string, it's barely noticeable. The Boss CS-3 does the same thing. It's a consequence of using a VCA (voltage-controlled amplifier) for fast response time. Turning RELEASE up above noon makes the distortion go away, but at the expense of a slower response time. Optical compressors have the advantage of the LDR's slower response filtering out the control ripple. It's a trade-off of distortion vs. response time. If you're running a dirt pedal or a cranked amp, then that tiny bit of distortion doesn't really matter.
The control loop is very stable and does not overshoot or ring. It is normal to hear "breathing" when you play chords because the peak detector senses everything and the low freq content is the loudest. Any beating between notes or feedback will cause the volume to go up and down and the gain control will try to track that. Turning up the BRIGHT control helps a little by shifting the emphasis away from the lower frequencies. Turning RELEASE up also helps by slowing down the response time. It's a balancing act.
Last edited:
almondcity
Well-known member
Boneyard Edition when???
Chuck D. Bones
Circuit Wizard
OK, now it's done. I'm still going to take another crack at the original ET, but this version which uses the output signal to control the gain is finished. I changed how the SUSTAIN control works. Like the CS-3, the SUSTAIN control limits how high the automatic gain control can set the gain. The lower the voltage on C8, the higher the gain. Turning SUSTAIN down raises C8's minimum voltage. I put the other internal buffer in the LM13700 to work as a voltage clamp. To ensure that the SUSTAIN setting doesn't drift, Vref is now regulated at +5V. If you don't have a B5K for the SUSTAIN control, then use B10K and raise R21 to 10K. I used tantalum for C8, although aluminum will work. I rearranged how the BRIGHT control works. Instead of raising the treble gain, it lowers the bass gain. Has the same effect on the tone, but this way it's less likely to run out of headroom. With BRIGHT at zero, the freq response is flat. A good initial setting is all controls at noon. I changed D4 & D5 to red LEDs. The only time they conduct is during the initial overshoot, which lasts a few milliseconds. The max output with LEVEL dimed is about 1.5V p-p. Unity is around 10:00. If you want it to drive your amp harder, stick a booster after it.
Last edited:
Chuck D. Bones
Circuit Wizard
U2B serves two purposes:
1. It buffers the DC voltage coming from the SUSTAIN pot.
2. It acts like a diode and only sources current into C8. If Q1 & Q2 want to raise the voltage on C8, U2B does not interfere.
When there is no guitar signal, C8 discharges thru R18 and VR4 until U2B starts conducting. U2B sets the minimum voltage on C8. The voltage on C8 feeds U2D (another buffer) which feeds current to U2A & U2C via R19 & R20. The current in R19 & R20 set the gain of U2A & U2C. U2A & U2C control the feedback around U1A. The overall effect is when the voltage on C8 rises, the gain of U1A goes down, and vice-versa. By setting the minimum voltage on C8, the SUSTAIN control is setting the maximum gain. I stole the SUSTAIN control idea from the Boss CS-3. The CS-3 uses an opamp & a diode between the SUSTAIN pot and capacitor; I used a Darlington emitter-follower (U2B).
1. It buffers the DC voltage coming from the SUSTAIN pot.
2. It acts like a diode and only sources current into C8. If Q1 & Q2 want to raise the voltage on C8, U2B does not interfere.
When there is no guitar signal, C8 discharges thru R18 and VR4 until U2B starts conducting. U2B sets the minimum voltage on C8. The voltage on C8 feeds U2D (another buffer) which feeds current to U2A & U2C via R19 & R20. The current in R19 & R20 set the gain of U2A & U2C. U2A & U2C control the feedback around U1A. The overall effect is when the voltage on C8 rises, the gain of U1A goes down, and vice-versa. By setting the minimum voltage on C8, the SUSTAIN control is setting the maximum gain. I stole the SUSTAIN control idea from the Boss CS-3. The CS-3 uses an opamp & a diode between the SUSTAIN pot and capacitor; I used a Darlington emitter-follower (U2B).
Chuck D. Bones
Circuit Wizard
I'm also cooking up an optical compressor based on the Hollis Flat-Line.
Chuck D. Bones
Circuit Wizard
Final update:
I changed C7 (tant), C8 (film), R18, R21, VR2, VR3, added R23. 78L05s can be a little noisy and R23 kills the noise. C4 can be 22nF or 33nF. 33nF adds a little more midrange when BRIGHT is turned up. With compressors, it's a trade between speed and distortion. Fast ones can make distortion and low distortion ones have a slower attack. Setting RELEASE below noon will produce some distortion on the lower strings, but the initial pop when SUSTAIN is dimed has been minimized. Not a bad compressor, but not as good as the Delegate IMO. Next up: one more look at the Thumbsucker.
I changed C7 (tant), C8 (film), R18, R21, VR2, VR3, added R23. 78L05s can be a little noisy and R23 kills the noise. C4 can be 22nF or 33nF. 33nF adds a little more midrange when BRIGHT is turned up. With compressors, it's a trade between speed and distortion. Fast ones can make distortion and low distortion ones have a slower attack. Setting RELEASE below noon will produce some distortion on the lower strings, but the initial pop when SUSTAIN is dimed has been minimized. Not a bad compressor, but not as good as the Delegate IMO. Next up: one more look at the Thumbsucker.
Last edited:
