electro-music.com   Dedicated to experimental electro-acoustic
and electronic music
 
    Front Page  |  Radio
 |  Media  |  Forum  |  Wiki  |  Links
Forum with support of Syndicator RSS
 FAQFAQ   CalendarCalendar   SearchSearch   MemberlistMemberlist   UsergroupsUsergroups   LinksLinks
 RegisterRegister   ProfileProfile   Log in to check your private messagesLog in to check your private messages   Log inLog in  Chat RoomChat Room 
 Forum index » DIY Hardware and Software » MusicFromOuterSpace.com designs by Ray Wilson
Quantizer
Post new topic   Reply to topic Moderators: dnny, Uncle Krunkus, v-un-v
Page 3 of 4 [87 Posts]
View unread posts
View new posts in the last week
Mark the topic unread :: View previous topic :: View next topic
Goto page: Previous 1, 2, 3, 4 Next
Author Message
lexvortex



Joined: May 14, 2008
Posts: 155
Location: Toronto

PostPosted: Fri Mar 20, 2009 12:50 pm    Post subject: Reply with quote  Mark this post and the followings unread

Hey Funkyfarm and guitarfool,

I too had a really hard time calibrating with a DMM, I found this thread and tried calibrating by ear using a MOTM300 osc and my polyevolver as the references, it worked very well, thanks for the heads up Very Happy

Cheers,
Dave
Back to top
View user's profile Send private message Visit poster's website
funkyfarm



Joined: Jan 21, 2007
Posts: 583
Location: France
Audio files: 3

PostPosted: Sat Mar 21, 2009 1:57 pm    Post subject: Reply with quote  Mark this post and the followings unread

Very Happy

I still have a problem here and i can't achieve calibration, even with my ears.

I play : C2, C3, C4 and get something like C2, A#2 and D3...not very tempered

Check all resistors (all were measured before soldering/hand-matched).
Change U5 TL074.
All Ic seems to be buffered (B, BE...)
10K trimmer seems to be fine.

I guess I have to take it off from panel, desolder all wires.
Inspect pcb patchs, then make all solderings one more time.

i'm vexed, but hopefull thanks to you. Very Happy
Back to top
View user's profile Send private message
numbertalk



Joined: May 05, 2008
Posts: 992
Location: Austin, TX
Audio files: 5

PostPosted: Sun Mar 22, 2009 7:45 pm    Post subject: Reply with quote  Mark this post and the followings unread

Just finished my quantizer tonight and calibrated with a combination of checking the voltages and comparing to another synthesizer and it's really close - might be able to get it a touch better but need sleep now. Starting next step of adding a CGS Diatonic Converter like guitarfool this week. Very exciting.
Back to top
View user's profile Send private message
Useemu



Joined: Jun 06, 2009
Posts: 42
Location: Germany

PostPosted: Sat Jun 06, 2009 3:13 am    Post subject: Reply with quote  Mark this post and the followings unread

funkyfarm wrote:
Hi,
I've build mine with hand-matched 100k and 200k resistors, roughly 0,1% 0,2%, but I have a problem (maybe wrong resistor value somewhere)

If I put 0V and 1V, I get 0,11 or more and hardly 0,786v (for 1V quantisation)
I've changed R13 from 75k (12V psu) to 51k (15V psu), and it's worth...0,560...

http://www.musicfromouterspace.com/analogsynth/VOLTAGEQUANTIZER/VOLTAGEQUANTIZER.php


Hi,

I have had the same problem. 1V in 0,560V out. I tracked it down to R40. Changing R40 from 200k to 51k solved the problem. Take a scope and look at pin 3 of IC U7. Here you must see the output of the counter and the r2r Ladder stepping up to 1V (with 1V input). Count the steps. In my case there were only six steps. That explains the 0,560V at the output. You should see twelve steps here.

You must add an C of 150p in parallel to R40 as well to get the right stepping in halfs. Without C every forth step is whole not half.

Useemu
Back to top
View user's profile Send private message Visit poster's website
funkyfarm



Joined: Jan 21, 2007
Posts: 583
Location: France
Audio files: 3

PostPosted: Sat Jun 06, 2009 1:55 pm    Post subject: Reply with quote  Mark this post and the followings unread

Quote:
Joined: Today
Posts: 1


Welcome ! Very Happy

Quote:
Changing R40 from 200k to 51k solved the problem.

You must add an C of 150p in parallel to R40 as well to get the right stepping in halfs. Without C every forth step is whole not half.


how did you find that ?!
i'm amazed.
Why other builders don't have to ? CD40193 IC tolerance/specs ?

Thank you so much.
Back to top
View user's profile Send private message
mono-poly



Joined: Jul 07, 2004
Posts: 937
Location: Rotterdam, Netherlands
Audio files: 2

PostPosted: Sat Jun 06, 2009 10:27 pm    Post subject: Reply with quote  Mark this post and the followings unread

Someone should do a DIY quantizer like this.


http://www.analogic-acs.com/quant.html
Back to top
View user's profile Send private message
Useemu



Joined: Jun 06, 2009
Posts: 42
Location: Germany

PostPosted: Sun Jun 07, 2009 3:48 am    Post subject: Reply with quote  Mark this post and the followings unread

funkyfarm wrote:
Quote:
Joined: Today
Posts: 1


Welcome ! Very Happy

Quote:
Changing R40 from 200k to 51k solved the problem.

You must add an C of 150p in parallel to R40 as well to get the right stepping in halfs. Without C every forth step is whole not half.


how did you find that ?!
i'm amazed.
Why other builders don't have to ? CD40193 IC tolerance/specs ?

Thank you so much.


How did I find this?

Measuring, thinking, calculating, poking around
I have build five items each with the same fault. I triple checked the parts used. Especially the R2R ladders. Nothing wrong with parts value.
1. Looked at pin2 of U7. Measured 1V. That proofed the input to be right.
2. Looked at pin3 of U7 with scope. Saw the steps adding up to approx. 1V before the comparator switched. seemed OK. (came back to here later)
3. Checked the the output pins of the latch (U2, U4). Find MSB_000111_LSB.
This calculates to 7x0.0833=0,5833V That proofed the output stage with R2R and the scaling to be alright.
4. Everything seemed to be alright. Fault still remains. Giving up.
5. paused
6. paused
7. Trained with my ergometer still thinking. Idea!!
8. Back to step 2. Looked again at pin3 of U7 and counted the steps. Only six steps to accumulate to 1V!! Thats it !! There must be something wrong with the R2R from the counter. Checked again the parts value. No faults
9. The output of the R2R ladder from the counter is scaled with R40.
10. Lowering R40 gives the right voltage reading and counts to twelve steps at the scope.
11. Applying an slow triangle to the input. Routing the e CV to an VCO and listening. Buuuh. Every fourth or fifth step was whole not half. An C with 150pF in parallel cured this.

I don't think the problem is in the tolerance or in the specs of the counter. Compare the R2R ladder from the counter to the R2R ladder of the output latch. At the output latch you will find R13 with 51k (for +/-15V) and C18 with 200pF.
Both ladders should be the same!? Shouldn't they? That is just what i did change after some hours of searching. Making both ladders the same.

I am wondering how others get it to work right with this differences in the scaling of the ladders. Maybe they should have a second look.

Did this solution worked for you?
Back to top
View user's profile Send private message Visit poster's website
funkyfarm



Joined: Jan 21, 2007
Posts: 583
Location: France
Audio files: 3

PostPosted: Sun Jun 07, 2009 3:02 pm    Post subject: Reply with quote  Mark this post and the followings unread

That's brilliant ! Very Happy

(do we live on the same planet ?!)

This is a real diy odyssey.

All is very clear.
Just correct the schematic, making counter R40 the same as output latch R13 ; adding missing cap. Basta.

I tell you soon if it works or not...
you've built 5 working items like that ? mine will be the sixth.

Thank you.
Thank you.
Thank you.


Posted Image, might have been reduced in size. Click Image to view fullscreen.
Back to top
View user's profile Send private message
Pehr



Joined: Aug 14, 2005
Posts: 1307
Location: Björkvik, Sweden
Audio files: 2

PostPosted: Mon Jun 08, 2009 1:46 am    Post subject: Reply with quote  Mark this post and the followings unread

Is this a known error in the schematics/PCB?

would be great if Ray is informed.

_________________
YouTube channel
flickr photostream
http://loxodrom.blogspot.com
http://www.garageband.com/artist/loxodrom
http://soundcloud.com/loxodrom
Back to top
View user's profile Send private message
funkyfarm



Joined: Jan 21, 2007
Posts: 583
Location: France
Audio files: 3

PostPosted: Mon Jun 08, 2009 2:04 am    Post subject: Reply with quote  Mark this post and the followings unread

He will be informed when i'll tell him to refund my money Smile , as i've just ordered few days ago another pcb in order to debug my first faulty one...

The scoop is that this second pcb will not work too.
Back to top
View user's profile Send private message
Pehr



Joined: Aug 14, 2005
Posts: 1307
Location: Björkvik, Sweden
Audio files: 2

PostPosted: Mon Jun 08, 2009 2:16 am    Post subject: Reply with quote  Mark this post and the followings unread

great! Smile

or is this a 15V supply quirk? I use 12V...

_________________
YouTube channel
flickr photostream
http://loxodrom.blogspot.com
http://www.garageband.com/artist/loxodrom
http://soundcloud.com/loxodrom
Back to top
View user's profile Send private message
Useemu



Joined: Jun 06, 2009
Posts: 42
Location: Germany

PostPosted: Mon Jun 08, 2009 9:17 am    Post subject: Reply with quote  Mark this post and the followings unread

Pehr wrote:
great! Smile

or is this a 15V supply quirk? I use 12V...


No. You should have the same problem, You have to use 75k for R40.

I digged a little deeper.

The structure here used is called "reverse ladder structure" (wording?) The output is calculated like this:

Uout=Uref*(Rl/(R+Rl))*(Z/(Zmax+1))

Uout=output voltage of the counter ladder
Uref=output voltage of CD40193 approx. 14,8V with 15V vss
R=100k ladder resistors
Rl=R40
Z=bin output from Counter
Zmax max bin output from counter here 63 (bin 111111)

You want to have 12 steps for 1V output. The 13. step should reset the counter and transfer the output to the latch.

Calculating with R40=200k and 12 steps:

Uout=14,8*(200k/(100k+200k))*(12/(63+1))=9,866*0,1875=1,894V
Way to high

Calculate with 200k and 6 steps

Uout=14,8*(200k/(100k+200k))*(6/(63+1))=9,866*0.094=0,923V
Right voltage false stepping That is what i saw on the scope.

Calculating with 200k and 7 steps gives Uour=1,079V the comparator fires, resets the counter and transfer 6 steps to the latch. That gives the wrong output.

Calculate with 51k and 12 steps:

Uout=14,8*(51k/(100k+51k))*(12/(63+1))=4,999*0,1866=0,937V
Just right. The 13. step should exceed 1V and resets the counter:

Uout=14,8*(51k/(100k+51k))*(13/(63+1))=4,999*0,203=1.015V
and bingo, the comparator fires, resets the counter and transfers the 12 steps to the latch.

Now i am sure 200k for R40 are wrong. With 15V Vss you have to use 51k instead. You can do the same calculation for 12V and you will find that you have to use 75k with 12V Vss.

I have Ray send an eMail with the calculation. I'll keep you informed about the answer.

Hope that helps

Bye for today
Back to top
View user's profile Send private message Visit poster's website
Useemu



Joined: Jun 06, 2009
Posts: 42
Location: Germany

PostPosted: Tue Jun 09, 2009 11:20 am    Post subject: Quantizer Reply with quote  Mark this post and the followings unread

Pehr wrote:
Is this a known error in the schematics/PCB?

would be great if Ray is informed.


I am in contact with Ray. He is just moving his lab. If its done he will evaluate my suggestion. I 'll keep you informed
Back to top
View user's profile Send private message Visit poster's website
funkyfarm



Joined: Jan 21, 2007
Posts: 583
Location: France
Audio files: 3

PostPosted: Sat Jun 13, 2009 10:40 am    Post subject: Reply with quote  Mark this post and the followings unread

Quote:
with 1v in, i could only get 0.560...


I've changed R40 from 200K to 51K so i can get the output around 1 volt.
But i had to swap R13 51K (in series with the trimmer) for a 47k resistor, unless i could not go down under 1,100V (approx.)

Now when I play C1, C2, C3, i can get 1,083V, 2,085V, 3.084V.

Which is a great improvment Very Happy
and which is fine to my ears.

But some "1/12 notes" were missing.
So i've tried to add a 220pF cap (260pf measured) in parallel with R40, but i lost my octaves (while pressing C1, C2, C3). (i use a 220pF for 200pF C18)

I've picked up a 150pF (190pF measured), i get my octaves on C keys back again (1.084, 2.084, 3.084) but some strange things occurs...while C octave voltages are ok, the scale is no more in 1/12v...

strange,

i'll see later.

Thank you.
Back to top
View user's profile Send private message
Useemu



Joined: Jun 06, 2009
Posts: 42
Location: Germany

PostPosted: Sun Jun 14, 2009 2:33 am    Post subject: Reply with quote  Mark this post and the followings unread

funkyfarm wrote:
Quote:
with 1v in, i could only get 0.560...


I've changed R40 from 200K to 51K so i can get the output around 1 volt.
But i had to swap R13 51K (in series with the trimmer) for a 47k resistor, unless i could not go down under 1,100V (approx.)

Now when I play C1, C2, C3, i can get 1,083V, 2,085V, 3.084V.



great

Quote:
But some "1/12 notes" were missing.


Same with me until i added 150pF in parallel to R40

Quote:

So i've tried to add a 220pF cap (260pf measured) in parallel with R40, but i lost my octaves (while pressing C1, C2, C3). (i use a 220pF for 200pF C18)


I use 150pF for C18 because i did not have 200pF

Quote:

I've picked up a 150pF (190pF measured), i get my octaves on C keys back again (1.084, 2.084, 3.084) but some strange things occurs...while C octave voltages are ok, the scale is no more in 1/12v...

strange,


indeed.. Scaling is alright for me.

New observation: Stepping from 0,083 to 0,166 did not work in 1/12 scaling. I can see the output oscillating. This only happens at the first step. With initial V set above this point everything works fine. This thing seems to be sensitive.

Some points to look at:

Is the scaling for whole and fourth Ok? If these are ok the problem lies in the two last bits. Maybe noise sensitive.

Try setting initial V to 1V or higher and check the 1/12 scaling again.
Try lower C
Back to top
View user's profile Send private message Visit poster's website
Useemu



Joined: Jun 06, 2009
Posts: 42
Location: Germany

PostPosted: Sun Jun 14, 2009 11:30 am    Post subject: Reply with quote  Mark this post and the followings unread

[quote="funkyfarm"]
Quote:


But some "1/12 notes" were missing.
So i've tried to add a 220pF cap (260pf measured) in parallel with R40, but i lost my octaves (while pressing C1, C2, C3). (i use a 220pF for 200pF C18)

I've picked up a 150pF (190pF measured), i get my octaves on C keys back again (1.084, 2.084, 3.084) but some strange things occurs...while C octave voltages are ok, the scale is no more in 1/12v...

strange,

i'll see later.

Thank you.


Are you sure you screwed up the scaling? If your only test is with the keyboard you might be wrong. Check wit your voltmeter and a pot at the input not with the keyboard. Maybe there is no fault. You use two devices which are not supposed to work together.

I gave it a second thought and some testing this afternoon.

I am aware now that you are using the quantizer with a keyboard. This device is not intended to be used with an stepped input. i.e keyboard.
This device is for making steps out of an continuous input voltage or to make scaling of an analog sequencer easier.
You will run into problems when you must use fixed input voltages from a keyboard. If your input voltage hits the switching point of the comparator the output flickers between two voltages. Or you can miss some steps.

But it should work with a keyboard with some restrictions

I tried using my analog keyboard from the Formant and i can simulate what you described at will. The good news is you can trim it out. Increasing the initial voltage makes the affect go away. I can use the over all tuning of my keyboard as well to get the right scaling.

Why did this happen? The R2R counter ladder is not properly scaled. You can not trim it to the right scaling (no trimer). There is no need to do so until you use an stepped input with right scaling. But if you have the right scaling before, you don't need the device.

Think of it like two slightly (miss)matched "windows". From time to time your window won't match and you miss a step.
Back to top
View user's profile Send private message Visit poster's website
funkyfarm



Joined: Jan 21, 2007
Posts: 583
Location: France
Audio files: 3

PostPosted: Mon Jun 15, 2009 3:36 pm    Post subject: Reply with quote  Mark this post and the followings unread

Useemu wrote:

New observation: Stepping from 0,083 to 0,166 did not work in 1/12 scaling. I can see the output oscillating. This only happens at the first step. With initial V set above this point everything works fine. This thing seems to be sensitive.


Yes, that's it ! what i call "C#0" ("considering C0 is 0,0083v...)

but other notes were missing too.
Back to top
View user's profile Send private message
funkyfarm



Joined: Jan 21, 2007
Posts: 583
Location: France
Audio files: 3

PostPosted: Mon Jun 15, 2009 3:41 pm    Post subject: Reply with quote  Mark this post and the followings unread

Useemu wrote:
it should work with a keyboard with some restrictions

I tried using my analog keyboard from the Formant and i can simulate what you described at will. The good news is you can trim it out. Increasing the initial voltage makes the affect go away. I can use the over all tuning of my keyboard as well to get the right scaling.


I'll try.

Useemu wrote:
Why did this happen? The R2R counter ladder is not properly scaled. You can not trim it to the right scaling (no trimer). There is no need to do so until you use an stepped input with right scaling. But if you have the right scaling before, you don't need the device.

Think of it like two slightly (miss)matched "windows". From time to time your window won't match and you miss a step.


stupid question : can't add a trim ?

Thank you so much.
Back to top
View user's profile Send private message
Useemu



Joined: Jun 06, 2009
Posts: 42
Location: Germany

PostPosted: Wed Jun 17, 2009 8:30 am    Post subject: Reply with quote  Mark this post and the followings unread

funkyfarm wrote:

stupid question : can't add a trim ?



Yes
Back to top
View user's profile Send private message Visit poster's website
guitarfool



Joined: Feb 26, 2007
Posts: 160
Location: Maryland
Audio files: 8

PostPosted: Sat Sep 12, 2009 8:54 pm    Post subject: Reply with quote  Mark this post and the followings unread

Useemu wrote:


Changing R40 from 200k to 51k solved the problem.

You must add an C of 150p in parallel to R40 as well to get the right stepping in halfs. Without C every forth step is whole not half.

Useemu


I made this change to mine and it works great now. I guess I hadn't noticed it before because I was mostly using random voltages and LFOs fed through a S&H to feed it Confused Thanks for the fix!
Back to top
View user's profile Send private message Visit poster's website
numbertalk



Joined: May 05, 2008
Posts: 992
Location: Austin, TX
Audio files: 5

PostPosted: Sun Sep 13, 2009 2:32 pm    Post subject: Reply with quote  Mark this post and the followings unread

Interesting, mine seems to work fine without it - I was testing all notes in a full octave. I'll have to go back and check this now.

guitarfool wrote:
Useemu wrote:


Changing R40 from 200k to 51k solved the problem.

You must add an C of 150p in parallel to R40 as well to get the right stepping in halfs. Without C every forth step is whole not half.

Useemu


I made this change to mine and it works great now. I guess I hadn't noticed it before because I was mostly using random voltages and LFOs fed through a S&H to feed it Confused Thanks for the fix!
Back to top
View user's profile Send private message
numbertalk



Joined: May 05, 2008
Posts: 992
Location: Austin, TX
Audio files: 5

PostPosted: Sat Feb 20, 2010 8:28 am    Post subject: Reply with quote  Mark this post and the followings unread

I am so frustrated. After the posts here I went back and more closely checked my quantizer and it is totally not quantizing well (not sure how I thought it ever was?!?). I tried the solution suggested here (swapping R40 for a 51K + 150pf cap in parallel) but still could not get it calibrated. So I switched that back to the 200K resistor and still no luck. I had given up and posted something about wanting to sell my module (interfaced with a CGS diatonic converter) which otherwise seems to work fine, but of course no one is interested in a module that cannot be calibrated. Plus I would really love this module to work and to keep it! Could anyone help? I've tried calibrating it using the "offset" pot on the panel. I'll measure the voltage with the pot at say 10 o'clock (not 0, since I've read and experienced the module can behave strangely around that setting), then I will turn the pot slowly until the next step is reached, measure that voltage and calibrate for .0833V change. I can get it ok for the first step but then it'll all over the place from there, and if I come back to my original starting point that voltage is totally off. I tried this with a tuner as well and it's atrocious - it will start at C, go to C# pretty close, then to D way flat, then D# way sharp (?!?) then drop to B! I also tried feeding it 0 volts from my CV keyboard, followed by 1V, to try to get a 1V change from it, checking with both a DMM and a tuner, and still, no luck. What am I doing wrong?? I was so excited about this module and am so bummed it's just sitting on top of my cabinet now, unused and unloved.
Back to top
View user's profile Send private message
kkissinger



Joined: Mar 28, 2006
Posts: 1354
Location: Kansas City, Mo USA
Audio files: 42

PostPosted: Sat Feb 20, 2010 9:20 am    Post subject: Reply with quote  Mark this post and the followings unread

numbertalk wrote:
After the posts here I went back and more closely checked my quantizer and it is totally not quantizing well


Well, if you want to part with these modules please PM me.

I am in the process of building two of these modules -- however, I will admit -- I got a little anxious and only after soldering everything together did I read that I should have matched the 1% resistors. -- well, I'll just go with what I have and see how it works **sigh**

The first question is: are your results consistent? That is, for a given input voltage do you get the same output voltage? If "yes", then you could have an open or a short in one of your data lines -- that is, pins 4, 7, 13, and 14 of U2. You can work backwards from there... to U3, etc.

Look for a pattern -- if every other note is wrong, then the low-order bit connection is messed up (for example).

I have yet to bring my modules to life so I'll get to experience the "trial by fire" soon. Smile

_________________
-- Kevin
http://kevinkissinger.com
Back to top
View user's profile Send private message Visit poster's website
numbertalk



Joined: May 05, 2008
Posts: 992
Location: Austin, TX
Audio files: 5

PostPosted: Sat Feb 20, 2010 10:50 am    Post subject: Reply with quote  Mark this post and the followings unread

Thanks for the reply kkissinger.

Checked for shorts, couldn't find any. Also seem to have all the correct components in the right places and right orientations where required. I'm using .1% resistors throughout.

Just took some measurements and this is what I got - any ideas?

I started with the Initial V pot at around setting 3. It measured .998V (it fluctuates a little, which I'm assuming is normal, but it was roughly .998)
I turned the pot clockwise and the next step I got was not +.083, so I trimmed to get 1.081 here (which would be an increment of .083)
Then I turned the post counter-clockwise to re-check the previous step and I got .983.
Turning the pot clockwise again I still get 1.081 at the next step
Turning the pot further clockwise the next step is .963!!! Not good!
Turning the post another step clockwise the next value is 1.149
Back to top
View user's profile Send private message
numbertalk



Joined: May 05, 2008
Posts: 992
Location: Austin, TX
Audio files: 5

PostPosted: Sun Feb 21, 2010 2:21 pm    Post subject: Reply with quote  Mark this post and the followings unread

By the way been working with Ray on this via email - don't want to jinx it but we *might* be getting somewhere - my clock rate was way too high. Changed R1 and it's acting much, much better now. Still getting weird readings on some of the steps (while incrementing one of the steps will actual put out a lower voltage before jumping back up) but better. Will post any updates here.
Back to top
View user's profile Send private message
Display posts from previous:   
Post new topic   Reply to topic Moderators: dnny, Uncle Krunkus, v-un-v
Page 3 of 4 [87 Posts]
View unread posts
View new posts in the last week
Goto page: Previous 1, 2, 3, 4 Next
Mark the topic unread :: View previous topic :: View next topic
 Forum index » DIY Hardware and Software » MusicFromOuterSpace.com designs by Ray Wilson
Jump to:  

You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
You cannot attach files in this forum
You can download files in this forum


Forum with support of Syndicator RSS
Powered by phpBB © 2001, 2005 phpBB Group
Copyright © 2003 through 2009 by electro-music.com - Conditions Of Use