Fall, Rise and Staircase up/down waveform generator?

[anotherjim]

It may just not be possible without doing it in digital, that is if the sudden step change is absolutely the effect you want. It could still make a useful LFO, but with much slower changes between steps and set the pots so they change level by degrees in imitation of traditional LFO waves.
From what I can see of the original LFO, it's biased low down toward 2.5v, but I might be working that out wrong. Your 100k Iabc control resistor may just be keeping the sweep in an area it doesn't like. BTW, if you experiment with that resistor value, make sure it can't exceed the Iabc 2mA safe current limit. 10k isn't a bad minimum from 9v, 4k7 is an absolute minimum.

[Rob Strand]

Rob: The 47k was a the non-inverting input of IC3.
Also, I meant that I took out the diodes to ground and instead connected the lugs to ground directly, which should have removed the offset, no?

The idea is fine but the 47k value isn't correct.

With the diodes the voltage at the bottom of the pot is,
2xdiode    = 1.2V to 1.4V, say 1.3V
3xdiodes  = 1.8V to 2.1V, say 2V

With the diode the offset is independent of the supply but with the grounded resistor

Consider one 4017 output driving a 10k pot.  There is only one 4017 output on at a time. 
With the pot on minimum the 10k forms a voltage divider with the grounded resistor. 
We want the voltage at the top of the added resistor to be 1.3V to be equivalent to two diodes
or 2V to be equivalent to 3 diodes.   The output at the resistor now depends on supply voltage
going to the 4017.

Suppose the supply is 9V.
If we want 1.3V output, then there's 9-1.3=7.7V across the 10k pot,
that means a current of 7.7V / 10k = 770uA flows through the 10k pot.
If we want 1.3V across the ground resistor that means that the grounded
resistor needs to be  1.3V / 770uA = 1.68k

If we want 2V we end up with the ground 2.86k

So the resistor needs to be somewhere in the 1k7 to 2k9 zone.

A lot smaller than 47k

[anotherjim]

My one-pot breadboard rig up (the 4017 doesn't actually exist, the +supply replaces it).

I get the best range with D2 just another 1N4148 and D3 a green LED.
The maximum step voltage is +9v-(D2+D1). I get about +7.7v which the opamp tracks OK.
The minimum step voltage is D3-D1. If I replace D3 with a diode string, the closest I could get to 1.4v is 1.6v unless I futz about selecting in x3 diodes with lower Vf (and there are really going to be x8 D1's). With the LED I get 1.45v.

I thought about the transient response of the buffer opamp. It could overshoot so added in a pre-filter R3/C1.

I thought about the LM358 crossover distortion might cause a bump, so added R2.

With the step voltage range now limited, the opamp can be a faster one with better transient response.

[strungout]

Jim: Tried your setup. It's good. Couldn't hear ticking. I left out the glitch filter, it blurred the steps too much. I reduced the 120k feeding the phaser to 47k. More or less ohmage blurs the steps.

I got some PLC662, still useful?

[anotherjim]

I forgot about pot resistance when calculating the glitch filter. This means there is no easy way to implement it there. It was meant to slow the step edge by about 1us but that increases as the pot add resistance as it's turned down. It might do something good if the cap went down from 1nF to 100pF but I'm not convinced now.

You could try the LMC662 (PLC662 is a Honda bike). Some reduction in power consumption. It doesn't really have any benefits in the circuit now apart from its pulse response of about 2us to swing which in itself is a glitch filter I suppose. I'm not very good at finding these details in datasheets though and the LM358 might be doing something similar anyway.

The 662 chip can be another choice for a low power LFO. If I read the data right, it idles on 400uA per amplifier.

[strungout]

I finished building it. It's not very noisy. No discernable ticking. Awesome. There's really not much to say... Everything went smoothly. The only problem I had was I wired the pots in reverse.

It plays really nice. It's fun to be able to create rythms with the sequence, though I haven't played it a lot yet. I made some (bad) soundclips. The first is the phaser mode 4/8 stages, the second is on Vibe. Just strummed some chords...

https://soundcloud.com/user-165425177/phasersequencer02/s-wj4NGBfB4wS?si=04f05bb93b53487c9419c383d2e5ab54&utm_source=clipboard&utm_medium=text&utm_campaign=social_sharing

https://soundcloud.com/user-165425177/phasersequencer01/s-g5JpZhupIxa?si=45f7ff43efe94f5bb1e82991e28ecfcf&utm_source=clipboard&utm_medium=text&utm_campaign=social_sharing




It's a mess...


The verified schematic. I used 3mm red LEDs for D1-9.



Thanks again for all the input guys.

[Rob Strand]

I finished building it. It's not very noisy. No discernable ticking. Awesome. There's really not much to say... Everything went smoothly. The only problem I had was I wired the pots in reverse.

It plays really nice. It's fun to be able to create rythms with the sequence, though I haven't played it a lot yet. I made some (bad) soundclips. The first is the phaser mode 4/8 stages, the second is on Vibe. Just strummed some chords...

All the hard work has paid off.   The steps sound smooth to me, no glitching, ticks, or noise.  Well done.   

Some interesting and some unexpected sounds in there.   I'm guessing the middle part of the first clip is 8-stages.

[ElectricDruid]

All the hard work has paid off.   The steps sound smooth to me, no glitching, ticks, or noise.  Well done.   

Some interesting and some unexpected sounds in there.   I'm guessing the middle part of the first clip is 8-stages.

+1 agree with Rob. You've definitely succeeded in getting all the wrinkles out of that and it sounds really good. Nice work.

[strungout]

Couldn't have done this well without you guys!

Rob: yeah, it's 8 stages when it gets wonky. Good wonky. I'm glad my build doesn't howl. I read some have trouble with that, because of the regen that is then going through 4 extra stages before getting back to the 2nd.

Anyway, I'm very pleased with how it turn out.

[Rob Strand]

Rob: yeah, it's 8 stages when it gets wonky. Good wonky. I'm glad my build doesn't howl. I read some have trouble with that, because of the regen that is then going through 4 extra stages before getting back to the 2nd.

I didn't think it was wonky, it just had a bit more of a flanger vibe to it.  You do get more pitch-shifting with more stages.