Causality 6 Phaser shifter

[soggybag]

I made up a board for frequency central's Causality 6 phaser. I got it mostly working. But had a few issues.

I put everything into Eagle and made a schematic and then a PCB.

- Regen pot working in reverse, almost has too much range, there is oscillation in the last 20% of travel
- Range pot works but distorts when range is lowest. Buzzing ripping distortions the sweep reaches the peak of the cycle
- Width seems to be working
- Shape seems to be working (note: shape and width have good interaction for interesting sounds)
- Stages 4/6 works the difference is audible but not huge.
- Level pot is reversed. Seems to be about unity at minimum, and way to loud at max. At the max range there's a lot of distortion. The taper is wrong.
- Rate works but all of the range is in the last 15% of travel. 
- Mix pot seems to work but range is reversed. At minimum there is no audio (for both vibe and phase.)
- Vibe switch makes a big pop when switched.

I used a TL074 and 13600 the original asks for TL084 and 13700. I don't think this will make much difference.

[ElectricDruid]

Regen pot working in reverse

Well, it would, wouldn't it? You've got the maximum setting (pin 3) at the bottom. Flip it the other way up.

almost has too much range, there is oscillation in the last 20% of travel

Add an extra resistor above the pot to limit the maximum setting to something reasonable. It could even be a trimmer, if you like. "Last 20%" suggests about 10K/12K as a value.

- Range pot works but distorts when range is lowest. Buzzing ripping distortions the sweep reaches the peak of the cycle

This sounds like a biasing problem. I'd experiment with R28, which limits how low the LFO can go. There's a lot of interactions in this part of the circuit between the Width and the Range, and keeping them all within "reasonable" parameters at all parts of the pot travel while not ruling out other interesting sounds at extremes of the other pot's travel might be quite tricky.
I don't know how the original pedal behaves, so I can't tell you wether this is normal or not. But it could be.

- Width seems to be working

Phew! Isn't it nice when that happens?!?

- Shape seems to be working (note: shape and width have good interaction for interesting sounds)

Excellent! It's a "Skew" control on the slopes of the triangle wave, but it shouldn't affect the rate or the depth/width/amplitude of the LFO, if I'm reading it right.

- Stages 4/6 works the difference is audible but not huge.

Yeah, not surprised. Sometimes this kind of stuff is a bit underwhelming. Yes, it changes the sound, and perhaps one sound is nicer than the other, but there's not so much in it as all that, so at some point it stops being worth putting a switch on it, and you just pick your favourite of the options and simplify a bit.

- Level pot is reversed. Seems to be about unity at minimum, and way to loud at max. At the max range there's a lot of distortion.

Again, you've got pins 1 and 3 of the pot the wrong way around. The pot needs to be flipped. There's a little arrow on the pot symbol which shows you which way "increasing" is. Think about what happens when you push that pot in the direction of the arrow to "Max". Is that what you want?
I note that Rick's schematics don't mention this detail - you're expected to work it out yourself. So no wonder.
This is one of the absolutely easiest things to get wrong in the whole world, and practically everyone here has done it. The people that tell you they haven't either haven't designed stuff themselves or are liars!

The taper is wrong.

Fix the direction and then see how you feel. It's possible/probable a log pot would work best here, if you've got a linear now. But if you've got a log the wrong way around, there's no surprise it feels terrible!

- Rate works but all of the range is in the last 15% of travel.

This strikes me as more serious. I'd be looking for wrong values in other parts of the LFO circuit. Rick knows what he is doing and the values chosen in the original schematic should give a decent range. If that's not the case in your version, I'd be hunting for copy-paste errors where I've forgotten to update the value of resistor after copying it into a new spot. It could also be the LFO capacitor value, C1. Check that too.

- Mix pot seems to work but range is reversed. At minimum there is no audio (for both vibe and phase.)

Again, you've got pin 3 a the bottom, so "max" is actually "min". In this case, that means you go from "fully-phaser" around to "vibrato" as you go clockwise around the pot. Which is fine, if that's what you like. But it doesn't explain why your signal is disappearing. That sounds like a build error, not a schematic error.

The mix pot is used to mix dry signal in. On the original FreqCentral schematic, at maximum you should have a full phaser effect with equal amounts of dry and wet signal. At minimum, you've got only the wet signal, so you should have a vibrato. I don't see anything wrong on your schematic apart from the pot flip (which just puts those two options the other way around), so are you sure you're getting phase-shifted signal at the output of the shifter stages? I could of course have missed something obvious...

I used a TL074 and 13600 the original asks for TL084 and 13700. I don't think this will make much difference.

The circuit uses the darlington buffer, which is where the difference between the two chips lies, so it's possible that it's significant. But like you, I *doubt* it makes much difference. I'd expect a bit more LFO feedthrough with the 13600 is all. That would be audible as more "throbbing" at higher LFO rates.
How did you come to use LM13600s, btw? They're not exactly something that we all have lying around these days... Even the DIP version of the LM13700 is discontinued now. I think the SMD version is still going, but I don't know for how much longer. It's already had a *very* long life for a microchip.

HTH

[soggybag]

I used a TL074 and 13600 the original asks for TL084 and 13700. I don't think this will make much difference.

The circuit uses the darlington buffer, which is where the difference between the two chips lies, so it's possible that it's significant. But like you, I *doubt* it makes much difference. I'd expect a bit more LFO feedthrough with the 13600 is all. That would be audible as more "throbbing" at higher LFO rates.
How did you come to use LM13600s, btw? They're not exactly something that we all have lying around these days... Even the DIP version of the LM13700 is discontinued now. I think the SMD version is still going, but I don't know for how much longer. It's already had a *very* long life for a microchip.

HTH

Thanks for the feedback. I have some 13600 in my stash of chips. I have a bunch of 3080 also. I might have some 13700 somewhere but I found these 13600 first so they went in for the first test.

[soggybag]

What's going on with the "vibe" switch? When started I thought this was a switch for vibrato mode. Now it looks like vibrato mode is covered by the mix knob?

What's happening with this switch? Looking at the original schematic closely, the switch was labeled: "Phaser/Filter". I had thought it was a vibrato mode and misinterpreted and mislabeled it in my version.

[ElectricDruid]

What's going on with the "vibe" switch? When started I thought this was a switch for vibrato mode. Now it looks like vibrato mode is covered by the mix knob?

What's happening with this switch? Looking at the original schematic closely, the switch was labeled: "Phaser/Filter". I had thought it was a vibrato mode and misinterpreted and mislabeled it in my version.

It switches the caps of two of the stages and turns them into lowpass filters instead of allpass filters. That gives you two fixed allpass stages, two swept allpass stages, and two swept lowpass stages. Hence the "phaser / filter" designation - it does literally switch between those options for those two stages.

[duck_arse]

both your range pot and mix pot are going to fight with ground all the time, as they will have V/2 at one end, coming from the opamps. you need some blocking caps. also, I'm not so sure you can string that many stages together without blocking caps somewhere, the darlingtons will have Vbe drops, but then you jamb them up against V/2 again. fights all round.

unless OTA's are a different country. you need an opamp-ish expert to tell, but.

also - to Druids' point about pots - the little arrow indicates clockwise rotation, more, up, in respect of the control knob on the front panel, not how the pot connections affect your circuit. you need to think about what makes the circuit go faster or louder or distort more - is it more resistance is faster, or the other way round? - and then work out which end of the pot you want as up. which determines your logs from your anti-logs.

[ElectricDruid]

both your range pot and mix pot are going to fight with ground all the time, as they will have V/2 at one end, coming from the opamps.

That was how Rick did it, but I agree it's a bit weird. Changing the range scales the whole waveform down closer to ground, rather than shifting the whole thing up or down.

you need some blocking caps. also, I'm not so sure you can string that many stages together without blocking caps somewhere, the darlingtons will have Vbe drops, but then you jamb them up against V/2 again. fights all round.

unless OTA's are a different country. you need an opamp-ish expert to tell, but.

I'm not that expert, but Rick's original gets by without the DC-blocking caps. But he used the 13700, and the buffers on that are organised to avoid the DC feedthrough you get with the 13600. So you may well be right in this case. I'd certainly want to get it working with 13700s in it first, and *then* see if it works with 13600's rather than trying to debug it with an IC choice of unknown validity.

[soggybag]

Thanks for all of the input! I learn something new every time I post here!

So an issue with the Range and Mix pot is that the output of the op-amp wants to stay at 4.5V but the pot to ground is always pulling it to 0v DC. A potential fix might be to place a cap between GND and the lower leg of the pot?

The Mix pot be better something like this:

[ElectricDruid]

So an issue with the Range and Mix pot is that the output of the op-amp wants to stay at 4.5V but the pot to ground is always pulling it to 0v DC.

Yes. In the case of the Range pot, I think it's deliberate. For the mix pot, it's an error, since it'll make the pot scratchy (it's got 4.5V DC across it).

A potential fix might be to place a cap between GND and the lower leg of the pot?

That would work, yes. Or take the lower end of the pot to Vref instead of Gnd.

The Mix pot be better something like this:

Yes, good idea. That gives you more flexibility without having to add another pot (goes from 0/100% to 100%/0%, instead of from 100%/0 to 50/50%). If you'Re doing that, you could turn the output mixer into an active circuit instead of the passive mix followed by make-up gain that it has now. Here's an example, doing the same Wet/dry mix with a PT2399 delay:


I suppose the downside of the active mixer is the inversion, if you care. In a phaser, the phase is all screwed up anyway, so perhaps it doesn't really matter!