Here's a nice little compressor that should be a fairly easy build. It doesn't have the squish of a DynaComp, but goes for a more subtle compression like the Orange Squeezer. Board mounted pots and an easy fit in a 1590B. Here's the schematic for reference.
Not build yet... but double check with schematic everyday, looks like everything is fine, but dont know the effect about replace scothky diode,.. must to try it
I build this using 2n5458 to replace 2n5457, and a pair 1n5817 to replace 5819. It works, but when I turn up the comp pot, the sound gets louder and bassier. any clue?
Built one, but it sounds weird. The compression is not really subtle in my build, the signal is being 'pumped' way too hard (especially when playing soft on lower strings) on every position of comp pot. Used 5817 diodes and 2n5457 transistor.
I made 2 of these. I didnt have any 2n5457 when i made the first one, so i used a 2sk30 with twisted pins. Other than that, they are identical. And the one with the 2sk30 sounds a lot better. The 2n5457 one distorts easily, and has a trebly kinda noise to it. What purpose does the transistor serve in the circuit? Could it have this big impact, or have i done something wrong on the 2n5457 one?
Speaking of compressors, here are my fixes for the Rothwell Love Squeeze circuit. It uses a JFET instead of an LDR in the feedback loop of an opamp. With no signal applied, the FET’s resistance is about 300 ohms and the stage’s gain is anywhere from unity to 10x, depending on the sustain control setting. When signal hits it, a P-P rectifier hits the JFET’s gate with a negative bias voltage in proportion to the input signal. The JFET’s resistance rises and reduces the gain of the first stage – compression. The advantage of the JFET is that its attack time is incredibly fast, limited only by the rectifier – half a wavelength of whatever fundamental frequency you’re hitting the front end with. The problem is that most people who build this circuit end up with distortion; it behaves like it has very little headroom.
It has plenty of headroom, the problem is the way the sidechain blocks frequencies below 100Hz. It does this to avoid ‘pumping’ or overshoot on the compression envelope. But what you really get is phase shift in the sidechain at low frequencies, combined with lower drive and higher ripple on the P-P rectifier. So at a specific input level – right at the compression threshold – the circuit will snap in and out of compression very quickly, effectively acting like an LFO. It doesn’t do this at higher or lower signal levels, just at the threshold. You hear it as a blatty, sputtering decay on chords. It can be mitigated by increasing the release time (R11), but the only way to cure it is to let the sidechain run full-range. The changes above do that.
(Note: you can get this behavior by simulating the circuit’s transient response in LTSpice, so it is definitely a design flaw and not a layout problem. In theory, certain selected JFETs might have curves that avoid the problem, so it’s not guaranteed to affect all builds… but why not fix it definitively?)
There is a tradeoff, in that you will get some ‘pumping’ on big bass transients, but it is still far less noticeable than on more traditional compressors, because the attack time is so much faster. It seems possible to mitigate the overshoot by using a shelving filter in the sidechain amp instead of moving the highpass corner down. Maybe a -6dB filter below 100Hz. That might be a best-of-both-worlds solution, but I haven’t tried it.
I replace R11 with a 25kB pot in series with a 15k resistor, which yields a much more traditional range of release times (roughly 100-300ms), with fewer intermodulation artifacts from rectifier ripple modulating the compression.
Once this is done, there’s no reason to keep LEDs in the feedback loop, since only the absolute hottest pickups will distort the opamp, and only at max compression and for the first positive half-wave. It’s virtually inaudible.
The other thing is that C12 should be no larger than 4.7nF, which attenuates treble above about 7.5kHz. The stock Rothwell uses 15nF (supposedly), which would explain why it has a reputation for being ‘dark’ sounding.
OMG I was just watching how it is compared to 4knob keeley. This wins it!! AWESOME!
ReplyDeletesurely build it
are you sure that 10uF below the volume oriented correctly? just wondering
ReplyDeleteThat's how it is in the schematic and a verified stripboard layout I referenced. So yes.
DeleteThanks man
DeleteVery good, I was waiting for a layout of this pedal ... Thanks
ReplyDeleteAHA, finally. A good perf Lovesqueeze layout
ReplyDeleteCan 5819 replace 5817?
and is good replace 10u with MLCC?
Don't know for sure, but probably.
DeleteI'm thinking the same, 5819 for 5817. how's the result?
Deletehi,
ReplyDeleteI cant seem to find the transfer image...
It's in the Misc. folder
DeleteNot build yet... but double check with schematic everyday, looks like everything is fine, but dont know the effect about replace scothky diode,.. must to try it
ReplyDeleteInteresting conversation about the diodes of this compressor:
ReplyDeletehttp://www.diystompboxes.com/smfforum/index.php?topic=100976.msg890136#msg890136
Both BAR43S and BAT43 have the same forward voltage (already checked), so you can use the latter. I would use sockets and try BAT43 and 5817
I use BAT 46 and they work well.
DeleteYou can mark this one as verified. I just finished the build - it sounds great!
ReplyDeleteDid u get pop sound while playing muted notes?
DeleteI build this using 2n5458 to replace 2n5457, and a pair 1n5817 to replace 5819. It works, but when I turn up the comp pot, the sound gets louder and bassier. any clue?
ReplyDeleteBuilt one, but it sounds weird. The compression is not really subtle in my build, the signal is being 'pumped' way too hard (especially when playing soft on lower strings) on every position of comp pot. Used 5817 diodes and 2n5457 transistor.
ReplyDeleteCiao, ho Sentito il video e suona benissimo, COMPLIMENTI!! VOLEVO CHIEDERTI SE c'è il disegno elettrico. Devo farlo subito perché è bellissimo!!
ReplyDeleteQ bc5457,,,,OR BS170,,,or,,bc549 is good? Provisional. Grazie, sei bravissimo. Saluti dall'Italia.
ReplyDeleteIt needs to be a JFET like a 5457. The BS170 is a MOSFET, and the BC549 is a BJT.
DeleteHello!
ReplyDeleteI made 2 of these. I didnt have any 2n5457 when i made the first one, so i used a 2sk30 with twisted pins. Other than that, they are identical. And the one with the 2sk30 sounds a lot better. The 2n5457 one distorts easily, and has a trebly kinda noise to it. What purpose does the transistor serve in the circuit? Could it have this big impact, or have i done something wrong on the 2n5457 one?
Dont put leds.
ReplyDeleteHi Dedalo! Why ? Can u explain please !! Thanks
Deletehttps://atomiumamps.tumblr.com/post/158975466951/speaking-of-compressors-here-are-my-fixes-for-the#notes
ReplyDeleteSpeaking of compressors, here are my fixes for the Rothwell Love Squeeze circuit. It uses a JFET instead of an LDR in the feedback loop of an opamp. With no signal applied, the FET’s resistance is about 300 ohms and the stage’s gain is anywhere from unity to 10x, depending on the sustain control setting. When signal hits it, a P-P rectifier hits the JFET’s gate with a negative bias voltage in proportion to the input signal. The JFET’s resistance rises and reduces the gain of the first stage – compression. The advantage of the JFET is that its attack time is incredibly fast, limited only by the rectifier – half a wavelength of whatever fundamental frequency you’re hitting the front end with. The problem is that most people who build this circuit end up with distortion; it behaves like it has very little headroom.
ReplyDeleteIt has plenty of headroom, the problem is the way the sidechain blocks frequencies below 100Hz. It does this to avoid ‘pumping’ or overshoot on the compression envelope. But what you really get is phase shift in the sidechain at low frequencies, combined with lower drive and higher ripple on the P-P rectifier. So at a specific input level – right at the compression threshold – the circuit will snap in and out of compression very quickly, effectively acting like an LFO. It doesn’t do this at higher or lower signal levels, just at the threshold. You hear it as a blatty, sputtering decay on chords. It can be mitigated by increasing the release time (R11), but the only way to cure it is to let the sidechain run full-range. The changes above do that.
(Note: you can get this behavior by simulating the circuit’s transient response in LTSpice, so it is definitely a design flaw and not a layout problem. In theory, certain selected JFETs might have curves that avoid the problem, so it’s not guaranteed to affect all builds… but why not fix it definitively?)
There is a tradeoff, in that you will get some ‘pumping’ on big bass transients, but it is still far less noticeable than on more traditional compressors, because the attack time is so much faster. It seems possible to mitigate the overshoot by using a shelving filter in the sidechain amp instead of moving the highpass corner down. Maybe a -6dB filter below 100Hz. That might be a best-of-both-worlds solution, but I haven’t tried it.
I replace R11 with a 25kB pot in series with a 15k resistor, which yields a much more traditional range of release times (roughly 100-300ms), with fewer intermodulation artifacts from rectifier ripple modulating the compression.
Once this is done, there’s no reason to keep LEDs in the feedback loop, since only the absolute hottest pickups will distort the opamp, and only at max compression and for the first positive half-wave. It’s virtually inaudible.
The other thing is that C12 should be no larger than 4.7nF, which attenuates treble above about 7.5kHz. The stock Rothwell uses 15nF (supposedly), which would explain why it has a reputation for being ‘dark’ sounding.
This comment has been removed by the author.
ReplyDeletehttps://atomiumamps.tumblr.com/post/158975466951/speaking-of-compressors-here-are-my-fixes-for-the#notes
ReplyDeleteIT REALLY HELPED A LOT. NOW IT SOUNDS RIGHT
So remove led and replace 15nf with 4.7nf?
Deletedon't work for me, just volume pot. can anybody help me?
ReplyDeleteI made this with few subsitutions
ReplyDeleteI subsituted 22ohms with 18 ohms
1n5817 with 1n5819
I hear lot of poping while.im.playing idk how to decrease it.. any suggestions?