What does my Gated Comparator/Klee do? Funky Drummer?

From circuitbending to homebrew stompboxes & synths, keep the DIY spirit alive!

Moderators: Kent, luketeaford, Joe.

User avatar
lessavyfav
Ultra Wiggler
Posts: 991
Joined: Wed Jun 02, 2010 11:00 pm
Location: Brooklyn, NY

What does my Gated Comparator/Klee do? Funky Drummer?

Post by lessavyfav » Wed Oct 17, 2012 5:23 pm

I've FINALLY, after a 2 month absence, finished wiring up my Funky Drummer Panel. Now I am faced with the embarrassing challenge of figuring out what the hell this Gated Comparator thing is doing?

I feel like a dumb ass but I just assumed it would be clear by the time I built it. :oops: It is doing something cool but I have no idea if it is intended. I think I get this much:
-there is a comparator that compares CLOCK against IN based on a threshold set by the sensitivity knob.
-this triggers a gate and also puts a gate signal into the line of 8 stages.
-The gates in the line get pushed along down the line at the rate of clock.
-You can get the gate state along any of the line of 8 stages
-There is a mix and a mix INV out which outputs a mix of the knob level of any of the 8 stages that are currently high.

-There is a VC sens input to manipulate the sens value. It has an attenuator.

That all seems to be true based on my experiemnts. Now here is where it gets foggy....

-There is a loop in CV as well as a loop enable CV and switch which don't seem to be working. Not sure what loop CV does? Assume Loop Enable CV takes a gate to enable the loop.

-There is some man load and man loop switch that don't appear to be working. The names seemed clear but I don't know how they are supposed to work I realize now :sadbanana:

-There is an rnd and an rnd inv output which don't seem to be working either. I assume they randomly output gate signals.

:help:

Image

User avatar
pre55ure
Super Deluxe Wiggler
Posts: 1455
Joined: Tue May 01, 2012 7:15 pm
Location: SoCal
Contact:

Post by pre55ure » Wed Oct 17, 2012 7:11 pm

I will be anxiously awaiting others responses on this as well. I just ordered the panel from clarke68 and also just sort of assumed that what it does would be clear by the time I finish building it. :despair:

moogah
Wiggling with Experience
Posts: 452
Joined: Mon Nov 22, 2010 10:27 am

Post by moogah » Wed Oct 17, 2012 7:26 pm

Each of the 8 steps can be either on or off, exactly an 8-step gate sequencer.

However, unlike a gate sequencer there isn't a single output which shows weather the selected step is on or off. Here, the status of all 8 stages are rotated (to the right, it think) so that step 8 becomes step 1 and step 1 becomes step 2 etc. This happens each time a clock pulse arrives, again like a gate seq.

Now, the output, instead of being the status of the selected stage is the _summed value_ of each stage that is ON. Because the on/off value is rotated and each stage can have it's own value this produces strange sorta tangental sequences of voltages that can be changed in odd ways by changing the value of one or more stages.

the value of each step is the output of a comparator. Each time a clock pulse arrives the comparator looks at the input signal and loads it's result (on/off) into step 1 of the sequence and rotates what was step 1 to step 2.


When noise is used at the input you can get nice repeating random sequences. I'm told the noise ring is like this and that random looping sequencer is also like this. I'd try using a couple of lfo's one as the clock and the other as the input.. might be neat, should be a ton of ways to create long evolving patters with this.

The first thing I'd try is to set it up like the random sequencer and then experiment with what the loop stuff does...

Take this all with a grain of salt because I don't have a working one to check this with :)



edit: holy crap I can't grammar
Last edited by moogah on Wed Oct 17, 2012 7:43 pm, edited 1 time in total.

User avatar
DGTom
Super Deluxe Wiggler
Posts: 2615
Joined: Tue Aug 19, 2008 2:42 am
Location: Pt.Adelaide, Sth.Aust.

Post by DGTom » Wed Oct 17, 2012 7:33 pm

'tis a tough nut to crack, quite abit of info in various threads, but still, one of those ones you need to spend some quality time with to gel with.


viewtopic.php?t=68935

viewtopic.php?t=51937

viewtopic.php?t=13834

viewtopic.php?t=23471


nothing quite like it once you tame it tho.

User avatar
lessavyfav
Ultra Wiggler
Posts: 991
Joined: Wed Jun 02, 2010 11:00 pm
Location: Brooklyn, NY

Post by lessavyfav » Wed Oct 17, 2012 7:44 pm

Thanks DGTom, I'll check these out. How the loop parts are supposed to work will hopefully be cleared up in the links. I can't tell if they are not working correctly or if I don't know what they should do. :-)
I'm Tim and I'm excited for modular synthesis! :nana:

User avatar
ringstone
antipodean experimentalist
Posts: 2789
Joined: Wed Mar 17, 2010 4:31 am
Location: Trudging through the Capital Wasteland

Post by ringstone » Thu Oct 18, 2012 7:37 am

How the Funky Drummer got his name

Last week I pulled my Funky Drummer back apart and tweaked him a little bit, I had 4 different color leds for the comparator stages but wasn't happy with the orange, as they were pretty hard to tell apart from the red - so I stuck in blue instead.

Also I modded the comparator input to compare against a positive voltage range only, rather than the default bipolar range. It gave me a lot better range for the typical material I use for the input, although I guess that makes it slightly less useful for general comparator duties.

Previously I modded the Loop Enable switch to a DPDT to switch out the In jack when the external loop was engaged (otherwise both streams are interspersed). The Loop In is used for the output from another Gated Comparator, or one of the individual stage outputs. BTW this is especially handy when used in conjunction with the Man. Loop switch, as it enables the current comparator buffer to cycle as is, "freezing" it.

The Man. Loop switch in "On" position sends the output of stage 8 directly back into the comparator - in other words it "carries" the bit that is normally lost out of the end of the data stream back into the first bit again. It makes more sense to "mute" the data coming in the "In" jack in this case, otherwise it ends up being mixed in with the cycled data, eventually causing the buffer to fill completely. The "Man Load" position (momentary) forces the comparator input data high, so a bit will always be written while it is engaged - useful for "programming" in your own sequence.

Anyway, when I powered him back up, he was in a really good mood and started laying down some funky rhythms straight away. I was caught on the hop a bit, but did manage to record this short segment:

http://soundcloud.com/mindlesseye/how-t ... er-got-his

This is the Gated Comparator and Pulse Divider driving my home-made 808 clone box, FutureRetro XS for bass through both channels of a Thomas White LPG (fast/slow). The 808 box turns gates into triggers so it runs well off the outputs from the individual stages, I processed the gate from another couple of stage through the "OR" function and then a Slope Detector to get a shorter gate for pinging the LPG.

Cheers
Blair

richard
bananaphile
Posts: 7164
Joined: Sat Mar 14, 2009 2:17 am

Post by richard » Thu Oct 18, 2012 8:14 am

I modded the Loop Enable switch to a DPDT to switch out the In jack when the external loop was engaged (otherwise both streams are interspersed
everyone should do this. It should be in the schematic IMO as the manual loop function is close to useless without it.

Finding the right levels for the inputs is the most important part to actually get it doing anything. Also, trying every kind of input to everywhere, greek orgy style especially - patented Scott technique coming up - trying inputting audio to the CV input and CV to the main input, often gives better results than the conventional way

I have had one for two years and I still can't say that I really understand it fully!

do check out the other threads.
Bastard Science Vol.1 and 2 (Hordik, Buchla, Serge, EMS, Oberheim) https://soundanatomy.bandcamp.com/album ... ence-vol-1
https://soundanatomy.bandcamp.com/album ... ence-vol-2

Tales from the Voodoo Box (EMS Synthi A solos) https://soundanatomy.bandcamp.com/album ... ynthesiser

Richard Scott - Several Circles https://cuspeditions.bandcamp.com/album ... al-circles

User avatar
tIB
Stainer!
Posts: 9618
Joined: Mon Jun 08, 2009 5:28 am
Location: UK

Post by tIB » Thu Oct 18, 2012 8:27 am

How does the loop enable thing work? Mine doesnt have it (early metalbox one), I might look at modding mine though maybe it an older PCB.

User avatar
ringstone
antipodean experimentalist
Posts: 2789
Joined: Wed Mar 17, 2010 4:31 am
Location: Trudging through the Capital Wasteland

Post by ringstone » Thu Oct 18, 2012 8:54 am

tIB wrote:How does the loop enable thing work? Mine doesnt have it (early metalbox one), I might look at modding mine though maybe it an older PCB.
Loop Enable switches the input coming in from the Loop In jack into the first stage of the shift register. Actually I thought it went through the comparator, but it bypasses this - so you really need to use a "pre-processed" stream, with clearly defined on/off voltages. I tried using other sources such as LFOs, etc. but noticed this caused issues such as the output being unstable. It wasn't until I went over the schematics that I realised this was down to the design.

Another trick is to use different types of LFO waveforms to clock the gated comparator, rather than a square wave. You can get quite pleasing "off-kilter" rhythms this way.

Cheers
Blair

User avatar
rico loverde
Super Deluxe Wiggler
Posts: 4036
Joined: Fri Aug 06, 2010 10:40 am
Location: Portland, Oregon

Post by rico loverde » Thu Oct 18, 2012 9:00 am

great thread, this module def did my head in.
works for Darkplace Manufacturing

richard
bananaphile
Posts: 7164
Joined: Sat Mar 14, 2009 2:17 am

Post by richard » Thu Oct 18, 2012 12:09 pm

ringstone wrote:
Another trick is to use different types of LFO waveforms to clock the gated comparator, rather than a square wave.
Nice! I had discovered and then forgotten that, it fits my "put all the wrong signals into all the wrong places" thesis quite nicely! This module is like an unifinished essay, it needs some help and imagination from the user to be great!
Bastard Science Vol.1 and 2 (Hordik, Buchla, Serge, EMS, Oberheim) https://soundanatomy.bandcamp.com/album ... ence-vol-1
https://soundanatomy.bandcamp.com/album ... ence-vol-2

Tales from the Voodoo Box (EMS Synthi A solos) https://soundanatomy.bandcamp.com/album ... ynthesiser

Richard Scott - Several Circles https://cuspeditions.bandcamp.com/album ... al-circles

User avatar
tIB
Stainer!
Posts: 9618
Joined: Mon Jun 08, 2009 5:28 am
Location: UK

Post by tIB » Thu Oct 18, 2012 12:25 pm

ringstone wrote:
tIB wrote:How does the loop enable thing work? Mine doesnt have it (early metalbox one), I might look at modding mine though maybe it an older PCB.
Loop Enable switches the input coming in from the Loop In jack into the first stage of the shift register. Actually I thought it went through the comparator, but it bypasses this - so you really need to use a "pre-processed" stream, with clearly defined on/off voltages. I tried using other sources such as LFOs, etc. but noticed this caused issues such as the output being unstable. It wasn't until I went over the schematics that I realised this was down to the design.

Another trick is to use different types of LFO waveforms to clock the gated comparator, rather than a square wave. You can get quite pleasing "off-kilter" rhythms this way.

Cheers
Blair
Thanks for that, though I wasnt clear: mine is the older version without the loop switch, input or bit switches. Like this:

Image

I wonder if there are different revisions of the PCB or if the loop functionality is a bend of sorts?

EDIT: found it, its a different revision board.

richard
bananaphile
Posts: 7164
Joined: Sat Mar 14, 2009 2:17 am

Post by richard » Thu Oct 18, 2012 12:38 pm

can you create a loop by patching a stage back to the input Tib?
Bastard Science Vol.1 and 2 (Hordik, Buchla, Serge, EMS, Oberheim) https://soundanatomy.bandcamp.com/album ... ence-vol-1
https://soundanatomy.bandcamp.com/album ... ence-vol-2

Tales from the Voodoo Box (EMS Synthi A solos) https://soundanatomy.bandcamp.com/album ... ynthesiser

Richard Scott - Several Circles https://cuspeditions.bandcamp.com/album ... al-circles

User avatar
tIB
Stainer!
Posts: 9618
Joined: Mon Jun 08, 2009 5:28 am
Location: UK

Post by tIB » Thu Oct 18, 2012 12:44 pm

^ possibly, if I modulate the threshold with the random melody. Ill have a look later...

User avatar
ringstone
antipodean experimentalist
Posts: 2789
Joined: Wed Mar 17, 2010 4:31 am
Location: Trudging through the Capital Wasteland

Post by ringstone » Thu Oct 18, 2012 9:25 pm

tIB wrote:
ringstone wrote:
tIB wrote:How does the loop enable thing work? Mine doesnt have it (early metalbox one), I might look at modding mine though maybe it an older PCB.
Loop Enable switches the input coming in from the Loop In jack into the first stage of the shift register. Actually I thought it went through the comparator, but it bypasses this - so you really need to use a "pre-processed" stream, with clearly defined on/off voltages. I tried using other sources such as LFOs, etc. but noticed this caused issues such as the output being unstable. It wasn't until I went over the schematics that I realised this was down to the design.

Another trick is to use different types of LFO waveforms to clock the gated comparator, rather than a square wave. You can get quite pleasing "off-kilter" rhythms this way.

Cheers
Blair
Thanks for that, though I wasnt clear: mine is the older version without the loop switch, input or bit switches.
Sorry tIB, we've discussed your gated comparator before but I had assumed it was much the same except without the individual stage pots. It's probably a lot more like Tombola's random sequencer but with gate outputs.

Cheers
Blair

User avatar
lessavyfav
Ultra Wiggler
Posts: 991
Joined: Wed Jun 02, 2010 11:00 pm
Location: Brooklyn, NY

Post by lessavyfav » Thu Oct 18, 2012 9:54 pm

Thanks everyone! I got some more time today with it. Got the manual load and loop working and making sense. Thanks Ringstone for the loop in tip. I was definitely getting unstable results with an LFO in there- I'll mess with it more tomorrow. Also found that the sens knob takes massaging. "sens" is kind of misleading, right? It seems more like a bipolar reference voltage that the comparator compares against.

I tried patching the comp out to the loop in to make a loop differently but didn't really feel like it was working right. I'm still trying to figure out if my loop enable switch is on when flipped left or right ;-) I had thought that loop enable getting switched on would just lock in whatever was currently in the stream into a loop but so far haven't gotten that kind of a result. A little more in the AM.

The mix out as implemented is really weird and neat! Only Now I want 16 stages rather then 8!
I'm Tim and I'm excited for modular synthesis! :nana:

User avatar
ringstone
antipodean experimentalist
Posts: 2789
Joined: Wed Mar 17, 2010 4:31 am
Location: Trudging through the Capital Wasteland

Post by ringstone » Thu Oct 18, 2012 9:58 pm

lessavyfav wrote:The mix out as implemented is really weird and neat! Only Now I want 16 stages rather then 8!
This is where your loop in comes in handy, if you build 2, you can take the output from stage 8 on the first and run it to the loop in on the second, thereby extending it from 8 to 16 stages...

Cheers
Blair

User avatar
lessavyfav
Ultra Wiggler
Posts: 991
Joined: Wed Jun 02, 2010 11:00 pm
Location: Brooklyn, NY

Post by lessavyfav » Thu Oct 18, 2012 10:04 pm

I just understood the rand and inv rand outputs I think- the 8 stages make a bit(or byte or whatever) 0100110, say, and a DtoA converts that to a CV value, yes? Mine is broken I think but if that is the right Idea I can start debugging.
I'm Tim and I'm excited for modular synthesis! :nana:

User avatar
ringstone
antipodean experimentalist
Posts: 2789
Joined: Wed Mar 17, 2010 4:31 am
Location: Trudging through the Capital Wasteland

Post by ringstone » Thu Oct 18, 2012 10:23 pm

lessavyfav wrote:I just understood the rand and inv rand outputs I think- the 8 stages make a bit(or byte or whatever) 0100110, say, and a DtoA converts that to a CV value, yes? Mine is broken I think but if that is the right Idea I can start debugging.
Yes, that's exactly it! You therefore have two completely different CV outputs that are derived from the same circuit.

Check there's no problems with your TL072, that IC buffers and inverts the random output.

Cheers
Blair

User avatar
lessavyfav
Ultra Wiggler
Posts: 991
Joined: Wed Jun 02, 2010 11:00 pm
Location: Brooklyn, NY

Post by lessavyfav » Thu Oct 18, 2012 10:41 pm

Thanks, will do- also wondering if I need to put links in at the A, B, C,D etc. the PCB has lines but thinner then normal link indicators. Also in the notes I just saw that I should. Link A to B and C-D, and some others to get all the possible values... Ok, me, go to bed don't get dressed and Gogo the studio on a midnight munches run of SDIY madness...
I'm Tim and I'm excited for modular synthesis! :nana:

User avatar
ringstone
antipodean experimentalist
Posts: 2789
Joined: Wed Mar 17, 2010 4:31 am
Location: Trudging through the Capital Wasteland

Post by ringstone » Thu Oct 18, 2012 11:10 pm

lessavyfav wrote:Thanks, will do- also wondering if I need to put links in at the A, B, C,D etc. the PCB has lines but thinner then normal link indicators. Also in the notes I just saw that I should. Link A to B and C-D, and some others to get all the possible values... Ok, me, go to bed don't get dressed and Gogo the studio on a midnight munches run of SDIY madness...
Well, I left those off but put an 8 pin DIL header across ABCDEFGH. That way I can install jumpers if I want to add these in. Or you could have a separate breakout panel with switches if you liked:
Pad pairs A-B, C-D, E-F and G-H are to allow for optional switches to add the 5th through 8th bits to the random melody output. If these are left open, the melody will be limited to 16 notes. Optionally a header can be installed, allowing removable links to be used to set the desired number of bits semi-permanently.

Pad pairs J-K, L-M, O-P, Q-R are so the same can be done to the first four bits, but unlike the ones above, these are pre-linked by PCB tracks. These tracks will need to be cut if adding switches for these bits. The tracks that are to be cut form a "V" between each pair of pads.


Cheers
Blair

User avatar
bkbirge
Super Deluxe Wiggler
Posts: 1942
Joined: Sun Nov 28, 2010 8:48 pm
Location: Houston, TX
Contact:

Post by bkbirge » Sat May 17, 2014 1:40 pm

Just finished wiring mine up and put it in the rack, does anyone have a patch they can clue me in on to test out the comparator section? The other stuff works but I'm not getting any LED action on the 8 stages and not really sure if I've patched it in correctly.

Modular & Dork Rock:
https://bkbirge.bandcamp.com/
Racks:
https://www.modulargrid.net/e/racks/view/88730
https://www.modulargrid.net/e/racks/view/89177
https://www.modulargrid.net/e/racks/view/379579


"The person who says it cannot be done should not interrupt the person doing it." ~ Puck, 1903

User avatar
lessavyfav
Ultra Wiggler
Posts: 991
Joined: Wed Jun 02, 2010 11:00 pm
Location: Brooklyn, NY

Post by lessavyfav » Sat May 17, 2014 2:07 pm

This old nugget... Fond memories. The patch you'll do is - noise in to the "in" clock in to clock then mess with the sensitivity till you get "all" the lights lighting up, after that dial it back till you get a more random distribution.

Or- skip the noise in and just use the clock and mess with the manual switch- click up to light a slot and it should move down the line at the clocks rate. Flick that switch down to get the pattern to loop.
I'm Tim and I'm excited for modular synthesis! :nana:

User avatar
bkbirge
Super Deluxe Wiggler
Posts: 1942
Joined: Sun Nov 28, 2010 8:48 pm
Location: Houston, TX
Contact:

Post by bkbirge » Sat May 17, 2014 2:55 pm

lessavyfav wrote:This old nugget... Fond memories. The patch you'll do is - noise in to the "in" clock in to clock then mess with the sensitivity till you get "all" the lights lighting up, after that dial it back till you get a more random distribution.

Or- skip the noise in and just use the clock and mess with the manual switch- click up to light a slot and it should move down the line at the clocks rate. Flick that switch down to get the pattern to loop.
Thanks! Still get nada so time to troubleshoot.

Modular & Dork Rock:
https://bkbirge.bandcamp.com/
Racks:
https://www.modulargrid.net/e/racks/view/88730
https://www.modulargrid.net/e/racks/view/89177
https://www.modulargrid.net/e/racks/view/379579


"The person who says it cannot be done should not interrupt the person doing it." ~ Puck, 1903

User avatar
bkbirge
Super Deluxe Wiggler
Posts: 1942
Joined: Sun Nov 28, 2010 8:48 pm
Location: Houston, TX
Contact:

Post by bkbirge » Thu Jun 05, 2014 5:01 pm

Got some more time to troubleshoot, one of the beads looked connected but it wasn't so I had no positive power. Now it all has pretty lights and makes random melodies. Question though, the attenuators on the outputs of the individual registers, how much should they modify the signal? Because they aren't much at all.

Modular & Dork Rock:
https://bkbirge.bandcamp.com/
Racks:
https://www.modulargrid.net/e/racks/view/88730
https://www.modulargrid.net/e/racks/view/89177
https://www.modulargrid.net/e/racks/view/379579


"The person who says it cannot be done should not interrupt the person doing it." ~ Puck, 1903

Post Reply

Return to “Music Tech DIY”