Author |
Message |
Rykhaard
Joined: Sep 02, 2007 Posts: 1290 Location: Canada
|
Posted: Sat Sep 20, 2008 12:15 pm Post subject:
The 6 position 1V/Oct. rotary switch schematic Subject description: September 2408 - final problem solved |
|
|
I've just completed the successful test of an idea that I had at work last night.
I had originally bought 25 pieces of the LM329 +6.9V voltage reference, to use for making VCOs with. An idea last occurred to me, for building a resistor string off of the LM329, to have 6 divisions providing 1V each, for a 1V/Oct. rotary switch, for my VCOs.
Having only 9 x 100k 1% metal film resistors, I chose the 6 from them that were as close to each other as possible - aiming for the 0.1% difference. (I managed to find 6 of them between 100.2k and 100.5k)
I bread boarded the circuit in the schematic below and adjusted the trimmer to get as close to a 1V/Oct. increment at each step, as possible.
The results??
+6.004
+5.001
+4.000
+3.000
+2.001
+1.003
I am almost ECSTATIC about these results!
I'll be wrapping all of the components for this circuit around my 6 position rotary switch and mounting it into 1 of my 5 VCOs for testing.
If it's as successful as I feel it should be I'll then bread board a new test with 12 0.1%'d resistors to see if I'm able to do the same thing for 12 semitones of increment around a 12 position switch! Last edited by Rykhaard on Wed Sep 24, 2008 9:38 am; edited 3 times in total |
|
Back to top
|
|
|
Rykhaard
Joined: Sep 02, 2007 Posts: 1290 Location: Canada
|
Posted: Sun Sep 21, 2008 9:11 am Post subject:
|
|
|
ERROR - IGNORE THE ABOVE POSTING - ERROR
Everything worked out exactly as hoped for from the bread board, testing it all with a digital multimeter. I was testing the output voltage with a 100k 1% resistor to Ground, as the VCO's 1V / Oct. input would be. (100k 1% to virtual ground.)
But - when I hooked up the bread board output to the 100k 1% resistor input on a 1V / Oct. tuned VCO, the output levels for each step from the bread board, seemed to rise exponentially from step to step.
The error here is beyond my limited theory. The only guess that I have is the requirement of a Constant Current Source for the 1V/Oct. control, which the LM329 voltage source isn't providing. (Though it all measured perfectly to a DMM.)
I'm abandoning the idea and removing it from my web page's Schematics section as well.
|
|
Back to top
|
|
|
RF
Joined: Mar 23, 2007 Posts: 1502 Location: Northern Minnesota, USA
Audio files: 28
|
Posted: Sun Sep 21, 2008 9:37 am Post subject:
|
|
|
Rych,
I hate to see you give up on a project.... and I like this idea.
If it worked OK measuring the output on a DVM (very high impedance - (Megohms, I'll bet)) I wonder what would happen if you buffer it through an op amp voltage follower before sending it to the VCO?
Have not thought it through, but on first glance...maybe?
good luck
bruce |
|
Back to top
|
|
|
Rykhaard
Joined: Sep 02, 2007 Posts: 1290 Location: Canada
|
Posted: Sun Sep 21, 2008 9:52 am Post subject:
|
|
|
RF wrote: | Rych,
I hate to see you give up on a project.... and I like this idea.
If it worked OK measuring the output on a DVM (very high impedance - (Megohms, I'll bet)) I wonder what would happen if you buffer it through an op amp voltage follower before sending it to the VCO?
Have not thought it through, but on first glance...maybe?
good luck
bruce |
Thanks for the suggestion Bruce. I HAD thought of it just before I decided to nix the test. I wasn't sure whether it would work though and frustrated, didn't try it.
Everything is still set up on the bread board for it. Now though, I've sidetracked myself with a new frustration - for VCO #5 that I was just about to tune, it's default voltage setting range for it, has it oscillating up at 120khz! I'm trying to figure out what I might have messed up, in this 2nd conversion of a Bi-N-Tic to VCO only. (My 1st is working fine.)
When I solve this new issue, I'll then try slapping a opamp buffer into the bread board to see how it goes.
I've fairly strong memories though of Harry and Ray correcting me in the past, on previous ideas for a 1V/Oct. sequencer - with Ray cluing me in as to the Constant Current Source, requirement.
Will post another update later. Thanks for your suggestion. |
|
Back to top
|
|
|
blue hell
Site Admin
Joined: Apr 03, 2004 Posts: 24079 Location: The Netherlands, Enschede
Audio files: 278
G2 patch files: 320
|
Posted: Sun Sep 21, 2008 10:02 am Post subject:
|
|
|
Rykhaard wrote: | ERROR - IGNORE THE ABOVE POSTING - ERROR |
You'll want to have the 100 k resistors way lower to avoid loading them with your CV inputs, more like 1 k or 100 R even. But then the zener/resistor thingie is going to protest so you'll want to have a buffer opamp between the voltage ref and the resistor chain. It could be done using a current source as well, but the trick is to have low impedance outputs. Of course you could also buffer all of the outputs separately, but that's a lot of opamps. _________________ Jan
also .. could someone please turn down the thermostat a bit.
|
|
Back to top
|
|
|
Rykhaard
Joined: Sep 02, 2007 Posts: 1290 Location: Canada
|
Posted: Sun Sep 21, 2008 12:09 pm Post subject:
|
|
|
1stly - Jan? Could you delete that duplicate post of mine, for me please? I've intermittent wireless network connections in the studio and I accidentally posted that message again, thinking it hadn't gone through.
Blue Hell wrote: | Rykhaard wrote: | ERROR - IGNORE THE ABOVE POSTING - ERROR |
You'll want to have the 100 k resistors way lower to avoid loading them with your CV inputs, more like 1 k or 100 R even. But then the zener/resistor thingie is going to protest so you'll want to have a buffer opamp between the voltage ref and the resistor chain. It could be done using a current source as well, but the trick is to have low impedance outputs. Of course you could also buffer all of the outputs separately, but that's a lot of opamps. |
OOOOoooooohhhhh! That is something else that I'd completely forgotten about that I HAVE read about somewhere / some time, over the years!
I had an alternate thought from what you'd stated - I went downstairs and stuck 1/2 of a TL082 as a resistorless input buffer for either of the 6 voltage levels.
I slapped a 1k resistor on the output of the buffer and hooked her all up to a VCO.
The result? Except for a SLEW when I changed the voltage output that I selected, manually - with wire from step chosen, to the opamp + input - it works PERFECTLY!
I had my DMM setup for Frequency and the tracking is well within 1% across the 6 octaves!
Now - can the SLEW be explained? I had RADICAL voltage changes whilst moving wire with my fingers. Moving it with insulated needlenose pliers solved that.
The slew time though, was about 2 seconds from the last selected voltage down to the no input voltage, with the wire not connected.
As soon though, as I connected the wire to a new voltage output, it immediately responded.
I'm going to move all of these parts from the bread board, to a small piece of perf board and attach it to this VCO permanently. (My new 4 x VCO panel has a pot hole drilled for each, for an octave selector rotary switch.)
If this continues to be as successful as it was with the buffer, I will THEN try going for the semi-tone rotary switch with buffer. (On bread board of course.)
Thanks for your knowledge Jan! Greatly appreciated! The last 24 hours has been like the stock market - huge excitement yesterday; huge bummerdom this morning and now huge excitement again! |
|
Back to top
|
|
|
blue hell
Site Admin
Joined: Apr 03, 2004 Posts: 24079 Location: The Netherlands, Enschede
Audio files: 278
G2 patch files: 320
|
Posted: Sun Sep 21, 2008 12:39 pm Post subject:
|
|
|
Rykhaard wrote: | 1stly - Jan? Could you delete that duplicate | I don't see a dup
Quote: | The last 24 hours has been like the stock market |
That's pretty wild _________________ Jan
also .. could someone please turn down the thermostat a bit.
|
|
Back to top
|
|
|
Rykhaard
Joined: Sep 02, 2007 Posts: 1290 Location: Canada
|
Posted: Sun Sep 21, 2008 1:17 pm Post subject:
|
|
|
Blue Hell wrote: | Rykhaard wrote: | 1stly - Jan? Could you delete that duplicate | I don't see a dup
|
It's the message where I've quoted / thanked Bruce (RF). I accidentally posted it, twice. |
|
Back to top
|
|
|
blue hell
Site Admin
Joined: Apr 03, 2004 Posts: 24079 Location: The Netherlands, Enschede
Audio files: 278
G2 patch files: 320
|
Posted: Sun Sep 21, 2008 1:40 pm Post subject:
|
|
|
Rykhaard wrote: | It's the message where I've quoted / thanked Bruce (RF). I accidentally posted it, twice. |
there was another message inbetween ... too complicated apparently for me to see it, anyway, it's gone now. _________________ Jan
also .. could someone please turn down the thermostat a bit.
|
|
Back to top
|
|
|
sneakthief
Joined: Jul 24, 2006 Posts: 569 Location: Berlin
|
|
Back to top
|
|
|
Rykhaard
Joined: Sep 02, 2007 Posts: 1290 Location: Canada
|
Posted: Mon Sep 22, 2008 5:28 am Post subject:
|
|
|
Thanks for posting the link to that one Sneak.
The size of it with surface mount components is wonderful! I'm still not a huge fan of soldering surface mount chips though. (I'd need a 1/64" tip, as my 1/32" is a little wide - distributing too much of my iron's 700º heat to surface mount items.) |
|
Back to top
|
|
|
Rykhaard
Joined: Sep 02, 2007 Posts: 1290 Location: Canada
|
Posted: Mon Sep 22, 2008 7:07 am Post subject:
New but related question ...... Subject description: ...... new but related question |
|
|
K. With the success of using Jan's mention of the opamp buffer, I've worked the rest out in my mind, hopefully logically and would like to ask all of the other viewers of this message whether I'm correct in my thoughts:
I wish to do the same idea of thing for semitones, around a rotary switch. That is 1/12th of 1 octave, for each step or 1 volt divided by 12. (0.083333)
If I am sticking with my LM329 +6.9V voltage reference then I need a string of voltage divider resistors such as the following:
LM 329 --> R1 (= 0.9V voltage drop - using a trimmer); R2 ( = 5 x 1V voltage drop) and R3 to R14 (= 12 x (1V / 12 resistors voltage drop)
So - if I were to use 10k resistors, I would be looking at (from top (LM329) to bottom):
LM329
200k trimmer (0.9V drop plus head room for tuning the overall string)
600k ((1V x (12 x 10k resistor value) x (5 octaves)))
10k = C
10k = C#
10k = D
10k = D#
10k = E
10k = F
10k = F#
10k = G
10k = G#
10k = A
10k = A#
10k = B
to opamp buffer for output.
Correct?
Another realization last night - if I wanted to have the ability to shift the tuning by 1 octave, I could have the resistor value for 1 octave (120k), switchable in either BEFORE or AFTER the rotary switch resistor string.
The overall resistance of the entire string would be the same. The actual voltage coming out would depend on whether the rotary switch were before or after the Octave Resistor.
Correct?
I'll try it out on the bread board to see where I'm at. I remembered also that I had brought home somewhere between 50,000 and 100,000 resistors (including 1%'s) from the trash pile at work. I'll hunt out a 1% value that starts with 1 in it's value, so that I can get an extra digit in the display of my digital multi meter, for 0.1%'ing them.
EDIT:
Update: I just finished wiring up the perfboarded version with an opamp buffer and fine tuned it over 6 octaves. Working beautifully!
Tomorrow - I'll try the semi-tone version. Logically, it should work out the same way. Here's to hoping! (I'll adjust it's schematic on my web page tomorrow, as well.) |
|
Back to top
|
|
|
Uncle Krunkus
Moderator
Joined: Jul 11, 2005 Posts: 4761 Location: Sydney, Australia
Audio files: 52
G2 patch files: 1
|
Posted: Mon Sep 22, 2008 3:06 pm Post subject:
|
|
|
Hey Rykhaard,
Make sure you check out this thread -
http://electro-music.com/forum/topic-25065.html&highlight=octave+switcher
I had similar problems sorting out putting buffers on my octave switchers. They have to be buffered. That board from Elby which uses SMDs is the one I based my stripboard layout on. It works perfect now. _________________ What makes a space ours, is what we put there, and what we do there. |
|
Back to top
|
|
|
Rykhaard
Joined: Sep 02, 2007 Posts: 1290 Location: Canada
|
Posted: Mon Sep 22, 2008 9:13 pm Post subject:
|
|
|
Thankee Uncle Krunkus! I read the whole thread. Mine with the Jan suggested buffer, is working perfectly. It's actually over 5 octaves that it's almost perfectly tuned. With the lowest switch selection the VCO is humming away at 250hz. At the highest, it were measured at 8024hz. Not too bad! Only out by 0.25% up there!
At work I layed out the plan of action for my semitone rotary. Searching through my boxes of a bazillion resistors (which have all turned out to be 1%'s! Woohoo!) I found 3k01's.
I worked out all of the values for that, as well as the switching logic for moving the 'octave resistor' before or after the rotary switch.
Thinking about it afterward, on 2 destinations:
1) For the VCOs, I already have (and will have) the Octaves rotary switch - I wont need an octave switch on the semitone rotary
2) For my step sequencer's number X rebuild, I'll be using the semitone rotary switches for each step BUT - if I added 3 position octave rotaries for each step as well, then I wouldn't need the single octave switch for the semi. I like that idea much better though, it'll double the # of rotary switches for each step from 1 to 2. (Or in my case, from $5 to $10. )
I still like that idea better.
Will be testing my semitone layout on bread board in the morning. If successful as presumed, I'll mount it into the VCO that now has it's rotary switch (as well as lower it's initial voltage level, from 125hz, down to 20hz or something. (Hmmmmm - 40hz to 1280hz with a 6 position switch. Makes me think I should go to an 8 or 12 position switch for the VCOs (Ray) will also be used as LFOs. )
K. I'm rambling ...... |
|
Back to top
|
|
|
Rykhaard
Joined: Sep 02, 2007 Posts: 1290 Location: Canada
|
Posted: Tue Sep 23, 2008 11:16 am Post subject:
|
|
|
Well kids - here's an odd one.
I did all of the testing the for the semitone switch and got everything tuned up properly on the breadboard, going to the summing node of the buffer opamp through a 100k 1% (as the virtual ground summing node on the VCO has everything coming into, each.)
Tuning was absolute PERFECT for 1 octave with 0.0833 increments on every step. Each of the 12 steps were perfectly matched 30k9 1% metal film resistors, to 2 decimal points.
Hooking it all up to the VCO that has the properly tuned octave switch, the tuning in the Octave switch changed by 1 or 2 % and the semitone switch itself (originally hooked up through a 100k 1% metal film to the buffer which then outputs through a 1k to the summing node's 100k 1% and then changed so that the switch output went directly into the buffering opamp - no difference) and the semi-tone switch is changing on each of it's 12 steps, by a WHOLE tone.
I'm stumped. For further reference, the VCO data:
- formerly a Ken Stone Bi-N-Tic, converted to +/-5V (10vpp) Sawtooth waveform only VCO
- 3 x CV inputs (2 with 100k attenuators) into each of their own 100k 1% metal film resistors
- 100k Tune pot into a 1 meg 5%
- Octave Switcher 6 position octave switch into it's buffer, then directly (no 1K) into it's own 100k 1% metal film.
- Semitone Switcher 12 position semitone switch it's it's buffer, through a 1k, into it's own 100k 1% metal film.
Any ideas, anyone?? Thanks for any help. I'm off to work. |
|
Back to top
|
|
|
blue hell
Site Admin
Joined: Apr 03, 2004 Posts: 24079 Location: The Netherlands, Enschede
Audio files: 278
G2 patch files: 320
|
Posted: Tue Sep 23, 2008 12:09 pm Post subject:
|
|
|
Rych, please post a sketch of what you did exactly, it's kind of hard to guess now ... _________________ Jan
also .. could someone please turn down the thermostat a bit.
|
|
Back to top
|
|
|
Rykhaard
Joined: Sep 02, 2007 Posts: 1290 Location: Canada
|
Posted: Tue Sep 23, 2008 9:42 pm Post subject:
|
|
|
Blue Hell wrote: | Rych, please post a sketch of what you did exactly, it's kind of hard to guess now ... |
I'll see if I can go one better. I'll take photos of both sides of the perf board for each of the Octave Switch; the Semi-tone switch as well as the CV Input section to the VCO and notate everything in the photos, to try and have everything as clear to understand as possible.
I'll also hand draw exact schematics for each of them - possibly in combination with the section for the VCO as to where they're going.
I was trying to puzzle out the trouble all night at work but I can't think of anything that I could have connected incorrectly. I'll have to go down and look at everything again with a magnifier to see what I've done. There's quite possibly a non-connection or mis-connection SOMEwhere - as the semitone switch IS tracking properly. Just at 0.166666666V instead of 0.0833333333V.
I'm trying to figure out exactly what, could cause something of that sort to happen. (Learn Kirchhoff's Laws - learn Kirchhoff's Laws, I keep telling myself. )
One thing that I remembered at work - I had to add an extra 100k 1% to the virtual ground summing section of the VCO as it had only 5 input resistors. (Coarse; Fine; CV1; CV2 and 1V/Oct.)
I can't logically see THAT causing any trouble at all. I made sure to solder the extra resistor TO the summing point.
The point of this paragraph though, that I just remembered - the wire connection that was originally soldered onto the perfboard of the Octave Switcher, was desoldered from it and soldered to the output from the Semitone Switcher. (The Octave Switcher's output was wired to the new resistor that were added. It's still working as it were before, with a minor trimming adjustment to make.)
Right. I haven't gotten to any greater degree of clarity by babbling on and on so I'll post again once I have taken pics and properly notated them along with schematics drawn. |
|
Back to top
|
|
|
Rykhaard
Joined: Sep 02, 2007 Posts: 1290 Location: Canada
|
Posted: Wed Sep 24, 2008 9:41 am Post subject:
|
|
|
Thanks for the offers of assistance Jan.
In examining the new perfboard for the semitone rotary switch, I found that the Ground connection for the LM329 +6.9 volt reference, wasn't even connected!
I soldered her up; hooked the VCO back up to the modular; did some slight tuning and now both the Octave and Semitone switchers are working perfectly!
Off now to do a PCB for 4 of the semitone rotary switches for my sequencer as well as 3 more Octave and Semitone switches for my other VCOs. |
|
Back to top
|
|
|
blue hell
Site Admin
Joined: Apr 03, 2004 Posts: 24079 Location: The Netherlands, Enschede
Audio files: 278
G2 patch files: 320
|
Posted: Wed Sep 24, 2008 11:49 am Post subject:
|
|
|
Rykhaard wrote: | [...]now both the Octave and Semitone switchers are working perfectly! |
_________________ Jan
also .. could someone please turn down the thermostat a bit.
|
|
Back to top
|
|
|
diablojoy
Joined: Sep 07, 2008 Posts: 809 Location: melbourne australia
Audio files: 11
|
Posted: Wed Oct 01, 2008 3:56 am Post subject:
|
|
|
sounds fantastic great work would really like to see your final cct's i have been thinking of trying something like this also but not advanced as far
as this yet. having only just recently ordered some MFOS 1V/OCT VCO's boards and wanted to use octave and semitone switching on them in place of the pots. I didn"t really want to go down the smt path. Your solution is far better |
|
Back to top
|
|
|
Rykhaard
Joined: Sep 02, 2007 Posts: 1290 Location: Canada
|
Posted: Wed Oct 01, 2008 8:06 am Post subject:
|
|
|
diablojoy wrote: | sounds fantastic great work would really like to see your final cct's i have been thinking of trying something like this also but not advanced as far
as this yet. having only just recently ordered some MFOS 1V/OCT VCO's boards and wanted to use octave and semitone switching on them in place of the pots. I didn"t really want to go down the smt path. Your solution is far better |
Thankee muchlee!
The only difference in the final, from the schematic at the top, is to add an opamp buffer at the switches selector. Hook the entire mess up to the 100k 1% CV input of the VCO and THEN adjust the trimmer to tune the switch.
I'll have to do a new schematic for the semitone one. I've gotten both of them done for 1 of my VCOs and they're both working perfectly in Octave and Semitone selection.
(I'm currently developing a step sequencer with the same controls, per step. ) |
|
Back to top
|
|
|
yusynth
Joined: Nov 24, 2005 Posts: 1314 Location: France
|
|
Back to top
|
|
|
Rykhaard
Joined: Sep 02, 2007 Posts: 1290 Location: Canada
|
Posted: Wed Oct 01, 2008 9:03 am Post subject:
|
|
|
Thanks for letting me know of that, Yves! I had seen it before and was quite interested IN it.
Reflecting on it, I wanted to make an individual one for a single VCO. That's why I went with my resign above (with a buffer, as well, of course ). I was able to build it fairly quickly - having MANY 1% resistors that I could draw 0.1%'d resistors from, with my digital multimeter. |
|
Back to top
|
|
|
Uncle Krunkus
Moderator
Joined: Jul 11, 2005 Posts: 4761 Location: Sydney, Australia
Audio files: 52
G2 patch files: 1
|
Posted: Wed Oct 01, 2008 2:58 pm Post subject:
|
|
|
That's a very nice circuit Yves. Exact and thorough. _________________ What makes a space ours, is what we put there, and what we do there. |
|
Back to top
|
|
|
diablojoy
Joined: Sep 07, 2008 Posts: 809 Location: melbourne australia
Audio files: 11
|
Posted: Wed Oct 01, 2008 8:58 pm Post subject:
|
|
|
Rykhaard wrote: | (I'm currently developing a step sequencer with the same controls, per step. ) | Very will be doing the same with
MFOS seq's but not until next year i think, i already have a lot of other modules to build first. my osc boards arrived today yay!+ the vcf boards i got last week |
|
Back to top
|
|
|
|