SUB 37 LFO 1 & 2 Phase

[oli@sub37]

Yesterday I was working on a sequenced sound, using the MOD 2 LFO (Saw Down) to modulate the filter in time with the sequence. Since at some point down the line I wasn't using the MOD 1 bus anymore but wanted to control the LFO modulated filter intensity using the Mod Wheel, I decided to copy over the settings from MOD 2 to MOD 1 just to have the preset using "standard" settings.

Interesting enough, the MOD 1 LFO (Saw Down) didn't produce the same result! It sounded "rounder" with less "attack", i.e the filter was opened and closed slightly earlier than with the MOD 2 LFO (Saw Down). That made me investigate what's going on, and I did the following:

- Only NOISE On
- NOISE Level to 0
- MOD 1 & 2 LFO (Saw Down)
- MOD 1 & 2 Destination Noise Level
- MOD 1 Amount +5 / -5
- MOD 2 Amount -5 / +5

If the two LFO were in phase then they should completely cancel out each other, producing no noise at any moment. But that wasn't the case. The two settings (MOD 1 AMT +5 & MOD 2 AMT -5 / MOD 1 AMT -5 & MOD 2 AMT +5) produced two different rhythmic noise patterns. The first setting made a very short noise with a strong attack, indicating that MOD 1 opened earlier and was then cut down by MOD 2. The second setting made a "swish" type of noise with a slower attack and slower release, indicating again that MOD 1 LFO was cutting before MOD 2 LFO started opening.

When I tried the same today, both LFOs were perfectly aligned, and the cancellation was working as expected!

Is it normal that the phase of MOD 1 & 2 LFOs can be slightly offset? And if so, is it on purpose or just in the nature of an analogue synth? If not, how can they be re-aligned?


And going further: It would be cool if there was a way to change the phase of the LFOs (in the firmware)!

[DrJustice]

You do not mention how your RATE, SYNC and KB RESET are set.

If I do as you describe, then switch on SYNC and KB RESET, set the RATE equal for MOD 1 and 2, I do get a steady noise level as expected. KB RESET resets the LFO phases when you hit a key, and SYNC makes them follow the ARP/MIDI clock rate with a divisor set by RATE.

[mikeh11]

The good Dr. says it all. Here's an exposition that I wrote up off-line while playing with the synth:

Is it normal that the phase of MOD 1 & 2 LFOs can be slightly offset?

Probably. You're thinking that if they are synced to the ARP clock that they must be phase-locked. It sounds like these LFOs don't work that way.

...is it on purpose or just in the nature of an analogue synth?

Probably on purpose. It might not be analogue, however.

...how can they be re-aligned?

KB Reset.

Repeating your Noise Level test, with KB Reset on for both LFOs, I get silence. (Well, a tiny pop at the beginning.)

My experiment, with square waves and the LEDs on the LFOs

- ARP rate at 60
- set both LFOs identical: rates just below 12, square waves, Sync on, KB Reset on
- hit a note, LFOs are now synchronized, the two LEDs blink in unison
- move LFO1 rate way up, then back to 6
- now the LFO lights are out of sync

That makes me think that when we changed the "charge" rate of the LFO, we changed the phase relation between the two. I think I read somewhere that the LFOs and envelopes are digitally generated. If so, it seems like these LFOs are good digital simulations of analog LFOs. Even when "Sync" is on (synched to the ARP clock) you can still adjust things, and the wave forms can shift around a little. To sync them back up, leave on "KB Reset" and hit a new note.

- play a new note, the LEDs blink together again
- turn off "KB Reset" on both LFOs
- hit a new note, still unison, since they were already synchronized
- move the LFO1 rate again
- now a new stable LED pattern

(I originally started out with pitch mod: set LFO1 rate to 3, with a pitch mod of +2. Set LFO2 rate to 6, with pitch mod +1. Get descending steps. Change the phase, get a different pattern. Reduce LFO1 to 1.4, different pattern.)

For anyone interested, here are some specs on the LFO waveshapes. The phase relation is indicated by the front panel markings, and only the triangle and SAH waves are bipolar. (Thanks to "KB Reset" for making this easy.)

Wave / Range / KB Reset point
triangle / bipolar / bottom
square / unipolar / high
saw up / unipolar / zero
ramp down / unipolar / high
SAH / bipolar / -

[oli@sub37]

DrJustice, you're right, I didn't mention the SYNC (on) and RATE (160) settings and didn't think the KB RESET (off) would matter.

Mike, thank you very much for the in depth explanation. I was able to reproduce the behaviour reliably. What threw me off is that I did expect switching presets would reset any SYNC'd LFO, but it doesn't. Now I know. better! But I have to admit I wish loading a preset would reset the Phase of SYNC'd LFOs.

Did a sound today that uses the S&H LFO synced to the Sequencer to offset the filter for every note. However if I set KB RESET to ON, then the filter is always the same because the first S&H value is always the same. Either I need a different way to synchonize the S&H LFO or I need a different source of random values on each Note ON.

[DrJustice]

...
For anyone interested, here are some specs on the LFO waveshapes. The phase relation is indicated by the front panel markings, and only the triangle and SAH waves are bipolar. (Thanks to "KB Reset" for making this easy.)

Wave / Range / KB Reset point
triangle / bipolar / bottom
square / unipolar / high
saw up / unipolar / zero
ramp down / unipolar / high
SAH / bipolar / -

That's good info - thanks, Mike!