Author |
Message |
kokoon
Joined: Jan 09, 2006 Posts: 158 Location: slovenia
|
Posted: Mon Jul 02, 2007 3:26 am Post subject:
|
|
|
HOW BEAUTIFUL THAT SOUNDS!!!
so do i understand this right:
the circuit takes an AC signal (actually technically as a control voltage) and outputs 5 streams of pulse waves, with duty cycles and phase positions determined by the level of the input at the given moment (windowed x5 vertically - each of the 5 streams corresponds to one window).
then the output streams of your choice (switch) are effectively summed together.
that's an awesome idea!
some questions:
- why are you talking about the "digital part" of the ciruit?
- you wrote that, if you set the divider CV so that the entire input falls just into the first window, you get a single SQUARE wave... wouldn't that be nearly impossible, wouldn't you get a PULSE unless the divider is set EXACTLY to the center of the (symmetrical) ramp wave?
- i know it's not exactly "summing" you have there (why not?) but have you thought about mixing the 5 streams with different levels? since they're pulse shape maybe even LPF them separately?
i think, if summing the separate streams is possible then there are even more interesting options - manipulating the separate streams before summing.
could you maybe post some more samples? maybe with a sine wave? or even with a random audio input?
again, great sounds!!! |
|
Back to top
|
|
|
kokoon
Joined: Jan 09, 2006 Posts: 158 Location: slovenia
|
Posted: Mon Jul 02, 2007 3:42 am Post subject:
|
|
|
hmm.. thinking now... i guess there's a comparator there instead of a summer so you get constant level of the final output... but on the other hand, isn't it true that none of the separate streams' high states overlap? couldn't you just sum what you're sending to the U3 comparator?
thinking about modular use... you could just make all 5 pulse streams available. |
|
Back to top
|
|
|
frijitz
Joined: May 04, 2007 Posts: 1734 Location: NM USA
Audio files: 54
|
Posted: Mon Jul 02, 2007 8:07 am Post subject:
|
|
|
kokoon wrote: | so do i understand this right:
the circuit takes an AC signal (actually technically as a control voltage) and outputs 5 streams of pulse waves, with duty cycles and phase positions determined by the level of the input at the given moment (windowed x5 vertically - each of the 5 streams corresponds to one window). |
That's roughly correct. Don't confuse the audio signal in (AC coupled saw in the schematic) with the control signal ("shape" and "shape mod" in the schematic).
kokoon wrote: |
- why are you talking about the "digital part" of the ciruit? |
I call it that simply because it's all 1's and 0's (not analog).
kokoon wrote: |
- you wrote that, if you set the divider CV so that the entire input falls just into the first window, you get a single SQUARE wave... wouldn't that be nearly impossible, wouldn't you get a PULSE unless the divider is set EXACTLY to the center of the (symmetrical) ramp wave? |
By that argument you would never have a square wave in any synth system. "EXACTLY"??? What do you want? D3 sets the maximum voltage on the reference chain, so the signal in will produce a square wave if its amplitude has the same value. If the saw in is more than 5 V amplitude then you don't quite get to the square. If it is less you can dial it in by ear with the "shape control". So no, this isn't a precision laboratory instrument. It's a simple circuit producing a huge range of timbres.
kokoon wrote: |
- i know it's not exactly "summing" you have there (why not?) |
Three reasons. (1) The pulses seem to have the best symmetry when picked off at a level near their high end. (2) There is a bit of cross talk between stages at the 3914 outputs. (3) The circuit was designed to have a +/- 5 V output independent of the ps voltage. Please let me know if you have a better way to do this.
kokoon wrote: |
but have you thought about mixing the 5 streams with different levels? since they're pulse shape maybe even LPF them separately? |
Yes, of course I have. That's not the purpose of this circuit.
kokoon wrote: |
i think, if summing the separate streams is possible then there are even more interesting options - manipulating the separate streams before summing. |
Depends what you are interested in. Most people want simple inexpensive circuit boards, and this one gives you a big bang for the buck.
Quote: |
"thinking about modular use... you could just make all 5 pulse streams available." |
They are available. See the white wires on the board photo? Those are the leads to and from the switches. |
|
Back to top
|
|
|
vtl5c3
Joined: Sep 08, 2006 Posts: 425 Location: PDX
Audio files: 13
|
Posted: Mon Jul 02, 2007 9:07 am Post subject:
|
|
|
Nice sample, Ian. I'd like one of those boards, if you decide to do a run.
Romeo |
|
Back to top
|
|
|
Funky40
Joined: Sep 24, 2005 Posts: 875 Location: Swiss
Audio files: 1
G2 patch files: 5
|
Posted: Mon Jul 02, 2007 9:18 am Post subject:
|
|
|
@Ian
sounds very nice.
i take 2
OT:
Scott Stites wrote: | Especially the sound that appears at the 40 second mark. Again - wow! |
then you'll like the polivoksfilter clone from Marc Bareille
http://m.bareille.free.fr/modular1/vcf_polivoks/vcf_polivoks.htm
edit: look my audiofile ( under my nick. But synth olli had posted a much better one ) |
|
Back to top
|
|
|
kokoon
Joined: Jan 09, 2006 Posts: 158 Location: slovenia
|
Posted: Mon Jul 02, 2007 2:59 pm Post subject:
|
|
|
frijitz wrote: |
That's roughly correct. Don't confuse the audio signal in (AC coupled saw in the schematic) with the control signal ("shape" and "shape mod" in the schematic).
...
They are available. See the white wires on the board photo? Those are the leads to and from the switches. |
i see. i hope you didn't get me wrong, i didn't mean to be rude or disrespectful, i think this circuit is just fantastic and i thank you for posting it. it's just that i was so overwhelmed when i actually understood how it works ... the ideas just kept popping
but i really understand, the module works what it's meant to work and if someone wants something else they (we) can work it out ourselves. not being around in the scene for that long, i didn't know project's cost and simplicity were so important.
by the way, i really don't have the skills necessary to make any "improvements" (like you suggested, with the summing), i just thought you didn't think of it (silly me) and hoped someone'd get as excited as i have about the idea of processing the separate streams to make it work.
again, i really like how this sounds and when i'll build a modular, a couple of these will definitely be a part of it! |
|
Back to top
|
|
|
frijitz
Joined: May 04, 2007 Posts: 1734 Location: NM USA
Audio files: 54
|
Posted: Mon Jul 02, 2007 3:34 pm Post subject:
|
|
|
kokoon wrote: |
again, i really like how this sounds and when i'll build a modular, a couple of these will definitely be a part of it! |
Great! I'm glad you are interested in it. Another idea is to add a second set of switches feeding a second comparator and to dynamically mix the two resulting signals, possibly with an optional inversion. Another idea I worked on was to feed pulses 1, 3 and 5 into one comparator and pulses 2 and 4 into another and then varying the sign and magnitude of the second set with a separate CV input (using a VCA). The first set of pulses swung 0 to -5 V and the second set from 0 to x, with -5<x<5.
All great fun!
Ian |
|
Back to top
|
|
|
kokoon
Joined: Jan 09, 2006 Posts: 158 Location: slovenia
|
Posted: Tue Jul 03, 2007 1:02 am Post subject:
|
|
|
good idea. i like the second one better, i like tools to be prearranged a bit so they are somewhat limited (but still highly functional) as opposed to offering tons of options to distract your attention.
if only the signals overlapped somehow. i know it's impossible with just LM3914, you could get a fixed voltage(just DC bias) out of it at best if using all the outputs.
would it be possible to delay some signals? since you can operate between two comparators, could you maybe use just a simple capacitor? the second comparator wouldn't produce the same width of pulses but maybe that'd make it sound even more interesting.
i'm sorry i'm thinking aloud again, i'm just asking to know what you think, not to say you should do that
on the other hand, the beauty of this circuit is that the output always has fixed level, and that you can be sure it's constructed entirely by pulse waves. |
|
Back to top
|
|
|
richardc64
Joined: Jun 01, 2006 Posts: 679 Location: NYC
Audio files: 26
|
Posted: Tue Jul 03, 2007 5:08 am Post subject:
|
|
|
frijitz wrote: | Another idea I worked on was to feed pulses 1, 3 and 5 into one comparator and pulses 2 and 4 into another and then varying the sign and magnitude of the second set with a separate CV input (using a VCA). The first set of pulses swung 0 to -5 V and the second set from 0 to x, with -5<x<5.
All great fun!
Ian |
Heh...or something like your Double Pulser, 2 VCAs with CVs going in opposite directions. There's no reason to stop at 5 pulses, is there? One VCA could get any combination of odd numbered pulses and the other the even numbered, inverted.
Great fun indeed, even to just think about. |
|
Back to top
|
|
|
psylux
Joined: May 15, 2007 Posts: 44 Location: Dallas, TX
|
Posted: Wed Jul 04, 2007 8:26 pm Post subject:
Re: That's not a waveshaper ... |
|
|
Can't wait to build one. What program did you use to generate that schematic gif?
frijitz wrote: |
Photo, schematic and sound clip are attached. In the sound clip, the drone uses a continuously varying waveshape and the other parts use an ADSR to sweep the timbre with different depths and initial waveforms.
Ian |
|
|
Back to top
|
|
|
frijitz
Joined: May 04, 2007 Posts: 1734 Location: NM USA
Audio files: 54
|
Posted: Mon Jul 09, 2007 7:41 pm Post subject:
|
|
|
richardc64 wrote: | Heh...or something like your Double Pulser, 2 VCAs with CVs going in opposite directions. There's no reason to stop at 5 pulses, is there? One VCA could get any combination of odd numbered pulses and the other the even numbered, inverted.
Great fun indeed, even to just think about. |
Lots of possibilities -- so many it's hard to decide where to stop.
The limit is five pulses with the current approach, since only alternate outputs are used, but if the individual amplitudes are controlled then you can use all ten. |
|
Back to top
|
|
|
frijitz
Joined: May 04, 2007 Posts: 1734 Location: NM USA
Audio files: 54
|
Posted: Mon Jul 09, 2007 7:50 pm Post subject:
Re: That's not a waveshaper ... |
|
|
psylux wrote: | Can't wait to build one. What program did you use to generate that schematic gif? |
Boards will be available this week
I do my schematics with a drawing program, since I don't like how the CAD programs look for presentation. Just some shareware thing that I made my own symbols for. |
|
Back to top
|
|
|
frijitz
Joined: May 04, 2007 Posts: 1734 Location: NM USA
Audio files: 54
|
Posted: Mon Jul 09, 2007 8:00 pm Post subject:
Boards |
|
|
The first set of boards for this project worked fine and a larger batch is on order. They should be here by Thursday and unless there are unexpected problems I can start shipping them next week. I've PM'ed the ordering info to people who sounded interested. (I've changed my pricing structure a bit to encourage multiple boards per order and to adjust for postal changes.) Anyone else who didn't get a PM and wants a board please PM me. |
|
Back to top
|
|
|
krisp14u
Joined: Nov 11, 2006 Posts: 206 Location: uk
|
Posted: Thu Nov 29, 2007 2:31 am Post subject:
|
|
|
Hi All
I missed out on these boards and would like one
Is there going to be a reissue of the PCB ?
if not dose anyone have a 5pulser PCB to sell or trade
I would do a straight swap with one of fonik's ps3100 PCB's
cheers
Paul _________________ Cheers
Paul Darlow
www.krisp1.com |
|
Back to top
|
|
|
prophei
Joined: Jan 27, 2007 Posts: 234 Location: san francisco, ca
Audio files: 4
|
Posted: Thu Nov 29, 2007 7:37 am Post subject:
|
|
|
i completely missed this one too, and am interested! |
|
Back to top
|
|
|
Luka
Joined: Jun 29, 2007 Posts: 1003 Location: Melb.
|
Posted: Thu Nov 29, 2007 7:44 am Post subject:
|
|
|
was the design of the pcb just what was in the schematic on the first page, or has it been developed further?
if it is something bigger i would love a pcb, otherwise i think i can breadboard that baby. _________________ problemchild
melbourne australia
http://cycleofproblems.blogspot.com/
http://www.last.fm/user/prblmchild |
|
Back to top
|
|
|
frijitz
Joined: May 04, 2007 Posts: 1734 Location: NM USA
Audio files: 54
|
Posted: Thu Nov 29, 2007 8:25 am Post subject:
|
|
|
Luka wrote: | was the design of the pcb just what was in the schematic on the first page... |
Yes, exactly that. There is a stripboard layout of a previous version here:
http://home.comcast.net/~ijfritz/sy_cir8.htm
This earlier version uses a 0-5V saw input (for dedicating to a VCO core), but you could easily mod it for +/-5V input wave.
If you do the page 1 version, I would suggest adding a switch for the first pulse. (This is always on in this version, but I've recently decided it's better to have a switch for it.)
BTW, there are more demos using the 5Pulser here:
http://electro-music.com/forum/topic-19504.html
Ian |
|
Back to top
|
|
|
Dave Kendall
Joined: May 26, 2007 Posts: 421 Location: England
Audio files: 3
|
Posted: Thu Nov 29, 2007 8:27 am Post subject:
|
|
|
If there's another run, I'd be up for 2, maybe more if funds allow.....
cheers,
Dave |
|
Back to top
|
|
|
Rykhaard
Joined: Sep 02, 2007 Posts: 1290 Location: Canada
|
Posted: Thu Nov 29, 2007 8:30 am Post subject:
|
|
|
If there're any left, I'd be interested in 1 or 2 as well. I'd bought 4 LM3914's earlier this year for a different project, that still hasn't occurred. I'd much rather use them in THIS one!
Speaking OF PCBs ...... Ian - when you have a moment, could you PM me with what you may still have available PCB wise, for sale?? I'd much rather work from yours, rather than laying out my own versions of them. |
|
Back to top
|
|
|
JovianPyx
Joined: Nov 20, 2007 Posts: 1988 Location: West Red Spot, Jupiter
Audio files: 224
|
Posted: Thu Nov 29, 2007 8:35 am Post subject:
|
|
|
I've been thinking about using waveshaping in the digital domain with an FPGA. I may try to model this circuit or something like it. In the digital domain, I may be able to play some different/additional games with numbers. |
|
Back to top
|
|
|
frijitz
Joined: May 04, 2007 Posts: 1734 Location: NM USA
Audio files: 54
|
Posted: Thu Nov 29, 2007 9:09 am Post subject:
|
|
|
ScottG wrote: | I've been thinking about using waveshaping in the digital domain with an FPGA. I may try to model this circuit or something like it. In the digital domain, I may be able to play some different/additional games with numbers. |
I'd love to see that. There's a lot that can be done with pulses. Even fixed-width ones:
http://www.variophon.de/index_e.htm
Ian |
|
Back to top
|
|
|
Peter Grenader
Joined: Jan 05, 2007 Posts: 14 Location: USA
|
Posted: Thu Nov 29, 2007 12:22 pm Post subject:
3914 sequencer |
|
|
My very first Milton prototype did just this. I just pulled out the board, tested it, still works. It's pretty spiffy way of sequencing, but it's part intensive. The way I ended up doing the VC on the analog Milton was a lot cleaner, easier, cheaper (using a A to D and the preset inputs of a 4516 counter).
hope this helps,
- P |
|
Back to top
|
|
|
JovianPyx
Joined: Nov 20, 2007 Posts: 1988 Location: West Red Spot, Jupiter
Audio files: 224
|
Posted: Thu Nov 29, 2007 1:40 pm Post subject:
|
|
|
Ian,
I believe that I understand how your circuit works, very clever. Given that I will have to invent a digital implementation of the LM3914, I should be able to increase the number of bars. The whole scheme should be fairly simple to implement digitally, just a bunch of comparators and a big OR gate.
In the sound clip, you were just modulating via the shape input? |
|
Back to top
|
|
|
Peter Grenader
Joined: Jan 05, 2007 Posts: 14 Location: USA
|
Posted: Thu Nov 29, 2007 1:51 pm Post subject:
|
|
|
We're using a microproccessor, we don't have to go into the details of replicating comparator gates like that. We simply read the data coming from an eight input serial ADC and assign commands to a series of hex numbers coming from the converter. This way we can change the input sensitivity to anything we wish, we're not limited to a comparator ladder's scaling. The micro responds by executing functions when numbers are present at the VC Input, like go to stage number X at hex number Y, and so on. Much easier. |
|
Back to top
|
|
|
Tim Servo
Joined: Jul 16, 2006 Posts: 924 Location: Silicon Valley
Audio files: 11
|
Posted: Thu Nov 29, 2007 1:58 pm Post subject:
That's not a waveshaper ... Subject description: one possible mod |
|
|
Hey Ian,
How about adding another 3914 (run in Bar mode) and a set of analog switches to replace the mechanical switches and allow voltage control of switch 1 to 5 closure?
Tim (just thinking out loud) Servo |
|
Back to top
|
|
|
|