MUFF WIGGLER Forum Index
 FAQ & Terms of UseFAQ & Terms Of Use   Wiggler RadioMW Radio   Muff Wiggler TwitterTwitter   Support the site @ PatreonPatreon 
 SearchSearch   RegisterSign up   Log inLog in 
WIGGLING 'LITE' IN GUEST MODE

Phase comparator: how many different ways to patch?
MUFF WIGGLER Forum Index -> Modular Synth General Discussion  
Author Phase comparator: how many different ways to patch?
wechard
Hi Everyone -- I'm wondering how many different ways there are to patch up a phase comparator using more basic modules. I know how to do it with an XOR, and also by using a Min/Max and then taking Max-Min. But the Wikipedia article on phase detectors seems to suggest that there are a few other methods, and it's beyond my electronics knowledge to look at that and figure out whether any of them are practical by patching. Can anyone point me to other modular-friendly methods beyond the two I mentioned above? Thanks...
dthorn
The Doepfer PLL module has a phase comparator built into it that can be patched separately from the other components of the module, but I don't know enough about how phase comparators work or what they might be used for outside the PLL to know if that's any use to you.
wechard
Thanks dthorn ... The Doepfer looks like fun, but what I’d like to do is play around with a PLL without setting aside the hp for a dedicated module. And even more than that, I’m interested in the phase comparator idea as a learning/theory experience.
Dcramer
I’m sure I’ve seen posts on here somewhere that describe patching up a PLL from scratch.
Keep searching thumbs up
cptnal
There are indeed....

Raven_Martin, 2016:

https://www.muffwiggler.com/forum/viewtopic.php?t=162994&postdays=0&po storder=asc&start=0
wechard
Thanks again ... Yes, I had already done some searching and found that thread (among others), which is where I learned the two techniques mentioned in my first post. The reason I wanted to ask the question is that those are the only two techniques I was able to find discussions of here on Muffs, but I couldn't tell if the various techniques in the Wikipedia phase detector article (https://en.wikipedia.org/wiki/Phase_detector) are coextensive with those, or if there are other possibilities described in that article that might be realized through patching. I'm pretty sure that none have been described in previous threads, although if someone could point me to one I missed then that would be great...
starthief
This paper has a bit on PLLs:

http://www.erode-sengunthar.ac.in/dept/lm/ECE/LIC/Analog%20Multiplier. pdf

This helped me understand why both the min/max or a four-quadrant multiplier (aka ringmod) will do the job. Using XOR is just sort of a crude-but-good-enough approximation of a four-quadrant multiplier.

It occurs to me I haven't tried patching this since getting Pico A Logic and ModDemix, so there are a couple more ways I can try this. Plus just using Hertz Donut's XOR output was pretty fun too. So that's a whole slew (heheh) of patch experiments I can play with tonight thumbs up
wechard
Hmmm, I'm only getting a single blank page at that link. Do you know if there's another source for the document?
ranix
it's a pdf, maybe choose "save as" when you're on the blank page

I was able to use it to create a phase-locked loop with the following patch:

OSC A sine -> ring mod X input
OSC B sine -> ring mod Y input
ring mod output -> state variable filter (SVF)
SVF low pass output -> OSC A linear frequency input

I could NOT get a lock when using a ladder filter instead of an SVF.
ranix
edit: nevermind, thought I'd done something I hadn't actually done (need another filter)
wechard
Well, that was interesting. Following up on ranix & starthief's comments about four quadrant multipliers, I tried using my u-mod as a phase comparator, with the two sines from a DPO. I found that taking the Ring output and patching it straight into the DPO oscillator 2 1V/Oct input produced absolutely perfect locking of oscillator 2 to oscillator 1 (as long as the base frequency of oscillator 2 was tuned high enough). This was without any filter at all, and the tracking was so good that it was actually boring.

I'm going to have to think more about why that worked so well -- with both sum and difference frequencies in the u-mod output, the DPO still locked tight to just the difference frequency component even without any filtering. Or was something else going on, and I wasn't hearing a PLL but some other kind of frequency sync?
ranix
Nope, you were hearing a PLL. You can achieve lock without filtering the high frequency component, but it causes harmonic distortion.

If you've got an oscilloscope with FFT or a spectrum analyzer you can see the difference on the oscillator receiving the sync signal. With the high frequency component filtered you'll have less harmonic content than you will with the high frequency component intact.

Depending on what frequencies you're dealing with, you might be able to observe the malformed waveshape on an oscilloscope without using FFT
AndreasD
starthief wrote:
This paper has a bit on PLLs:

http://www.erode-sengunthar.ac.in/dept/lm/ECE/LIC/Analog%20Multiplier. pdf

This helped me understand why both the min/max or a four-quadrant multiplier (aka ringmod) will do the job. Using XOR is just sort of a crude-but-good-enough approximation of a four-quadrant multiplier.

It occurs to me I haven't tried patching this since getting Pico A Logic and ModDemix, so there are a couple more ways I can try this. Plus just using Hertz Donut's XOR output was pretty fun too. So that's a whole slew (heheh) of patch experiments I can play with tonight thumbs up



Thanks for the link!
My light bulb moment was when I learned that the filter was to kill the x+y out of the ringmod and leave the x-y only.
Works brilliant with Befaco's A*B+C and Polaris filter.

But it appears that the filter has an influence in terms where AC only filters could have problems with very low freq x-y signals. And some filters do have a 180deg phase shift, others don't, which does have an influence on locking or not.
wechard
ranix wrote:
If you've got an oscilloscope with FFT or a spectrum analyzer you can see the difference on the oscillator receiving the sync signal.


Yes, I'm looking forward to getting deeper into this when I get my scope (which will be in about two weeks). I'm still not really grasping how a very low frequency x-y signal ends up working on the oscillator like a DC offset (does it come out of the multiplier offset rather than centered on 0V?), but I'm sure that when I take a look at different scenarios on the oscilloscope it will make more sense.
dooj88
wechard wrote:
ranix wrote:
If you've got an oscilloscope with FFT or a spectrum analyzer you can see the difference on the oscillator receiving the sync signal.


Yes, I'm looking forward to getting deeper into this when I get my scope (which will be in about two weeks). I'm still not really grasping how a very low frequency x-y signal ends up working on the oscillator like a DC offset (does it come out of the multiplier offset rather than centered on 0V?), but I'm sure that when I take a look at different scenarios on the oscilloscope it will make more sense.


dig around on the shapeshifter for dummies page he mentioned.

https://rabidelephant.atlassian.net/wiki/spaces/MD/pages/22708228/Inte llijel+Shapeshifter+Guide#IntellijelShapeshifterGuide-

no shame in that either, it's a bit less dry than the manual.
cptnal
dooj88 wrote:
wechard wrote:
ranix wrote:
If you've got an oscilloscope with FFT or a spectrum analyzer you can see the difference on the oscillator receiving the sync signal.


Yes, I'm looking forward to getting deeper into this when I get my scope (which will be in about two weeks). I'm still not really grasping how a very low frequency x-y signal ends up working on the oscillator like a DC offset (does it come out of the multiplier offset rather than centered on 0V?), but I'm sure that when I take a look at different scenarios on the oscilloscope it will make more sense.


dig around on the shapeshifter for dummies page he mentioned.

https://rabidelephant.atlassian.net/wiki/spaces/MD/pages/22708228/Inte llijel+Shapeshifter+Guide#IntellijelShapeshifterGuide-

no shame in that either, it's a bit less dry than the manual.


Nice. Going to have a good look at this over the weekend. This is fun!
huffnPuff
These threads make MW worth it. Thanks!
MUFF WIGGLER Forum Index -> Modular Synth General Discussion  
Page 1 of 1
Powered by phpBB © phpBB Group