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 » YuSynth
VCO tracking problem...
Post new topic   Reply to topic Moderators: Scott Stites
Page 1 of 1 [7 Posts]
View unread posts
View new posts in the last week
Mark the topic unread :: View previous topic :: View next topic
Author Message
minisystem



Joined: Nov 16, 2012
Posts: 35
Location: Toronto

PostPosted: Tue Jan 22, 2013 9:31 pm    Post subject: VCO tracking problem...
Subject description: realistic expectations for tracking accuracy?
Reply with quote  Mark this post and the followings unread

My surface-mount version of the Yusynth VCO is done and the waveforms looks great. Stability seems very good too. I now have a few questions about 1V/octave tracking. I've browsed around this forum and found previous threads describing problems and misunderstandings about how to scale the VCO so that it tracks at 1V/octave over 6+ octaves.

I'm starting with a CV of 0V, setting frequency to 55 Hz and then going to CV of 1V and adjusting T2 and going back to 0V and adjusting frequency until I get perfect tracking between 55 Hz and 110 Hz. Then up to A2 (220 Hz), which leaves A1 a little off, but not by much. By the time I get to the point where tracking between 0 and 3 V (55 Hz and 440 Hz) looks good, the octaves in between (110 Hz and 220 Hz) are off by 15-25 cents (I'm using a guitar tuner). If I keep going up to A7 so that 0V is 55 Hz and 6V is 3520 Hz, the octaves in between are off by anywhere from 10-25 cents

I get the counter-intuitive adjustment of T2, so I don't think that's the issue. Anyway, then I tune using T1 so that 0V is bottom C and A3 gives 220 Hz. At this point, I get pretty decent tracking over 2 octaves, but beyond that, everything is out by 10-30 cents.

My VCO is using this BCM847 matched pair with a maximum delta Vbe of < 2mV: http://www.nxp.com/documents/data_sheet/BCM847BV_BS_DS.pdf
The matched pair is thermally coupled to a 3300ppm/ÂșC 1K 0805 chip resistor.

All other semi-conductors are just smt versions of what Yves specifies on his VCO page. This is the OPA2137 I'm using: http://www.digikey.ca/product-detail/en/OPA2137UA/OPA2137UA-ND/275834

Should I expect better tracking performance? If so, any idea what to troubleshoot? [/url]
Back to top
View user's profile Send private message
minisystem



Joined: Nov 16, 2012
Posts: 35
Location: Toronto

PostPosted: Tue Jan 22, 2013 9:33 pm    Post subject: Reply with quote  Mark this post and the followings unread

I just noticed the OPA2137 I'm using has a typical input offset voltage of 2.5mV (max 10 mV), whereas another, slightly more expensive version, has a typical input offset voltage of 1.5 mV (max 3 mV). Perhaps I should have used the OPA2137 with the tighter specification?
Back to top
View user's profile Send private message
minisystem



Joined: Nov 16, 2012
Posts: 35
Location: Toronto

PostPosted: Wed Jan 23, 2013 8:11 pm    Post subject: Reply with quote  Mark this post and the followings unread

OK, so I've compared my VCO's tracking to a Modcan VCO. The Modcan VCO is 7 years old and tracks perfectly over at least 7 octaves to within 1-3 cents. So, it's safe to assume that I should expect tight tracking over 5-6 octaves. Any ideas on how to fix this?
Back to top
View user's profile Send private message
brock



Joined: May 26, 2011
Posts: 112
Location: Canada

PostPosted: Thu Jan 24, 2013 12:02 am    Post subject: Reply with quote  Mark this post and the followings unread

Just curious, but how can you assume any correlation between the performance of the Modcan VCO and the VCO you are building?

Regarding the problem you have, assuming your input voltage source is accurate, which seems a safe assumption since you can measure the Modcan accurately over 7 octaves, the only thing that would likely cause the tracking to be so bad is that the log conformance of your transistor pair sucks. Don't suppose you have something else to put in there? Less than 3 octaves seems unusually poor. I suspect you could get 5 octaves from hand matched 2N3904s.
Back to top
View user's profile Send private message
minisystem



Joined: Nov 16, 2012
Posts: 35
Location: Toronto

PostPosted: Thu Jan 24, 2013 8:09 am    Post subject: Reply with quote  Mark this post and the followings unread

brock wrote:
Just curious, but how can you assume any correlation between the performance of the Modcan VCO and the VCO you are building?


Well, I expected the Modcan VCO to track very well, but I had never actually measured it. I wanted to see what tracking well over 7-8 octaves really meant - did it mean close-to-perfect tacking within 1-2 cents over that range or more like within 10-15 cents? For the Modcan, it was the former rather than the latter. I guess my question is should I expect such precise tracking over 5-6 octaves for the Yusynth design?

brock wrote:
Regarding the problem you have, assuming your input voltage source is accurate, which seems a safe assumption since you can measure the Modcan accurately over 7 octaves, the only thing that would likely cause the tracking to be so bad is that the log conformance of your transistor pair sucks. Don't suppose you have something else to put in there? Less than 3 octaves seems unusually poor. I suspect you could get 5 octaves from hand matched 2N3904s.


Yes, input voltage is a Kenton Pro-Solo MIDI-CV converter and I measured the output, so CV source is good. I could swap out the existing matched pair for another of the same just in case this specific pair is out of spec? The BCM847 has been used by at least one other builder in the Yusynth VCO. Mentioned here and here.

I did fiddle with it a little more and if I start the scaling procedure at higher frequencies and work down I'm getting better tracking - I now get pretty good tracking (within 5-10 cents for most octaves) from C7 down to C2, with the caveat that C0 and C1 are out by more than a quarter tone.
Back to top
View user's profile Send private message
brock



Joined: May 26, 2011
Posts: 112
Location: Canada

PostPosted: Thu Jan 24, 2013 11:51 am    Post subject: Reply with quote  Mark this post and the followings unread

I would think you can expect decent tracking over a 5-6 octave range because many builders have achieved this. The question is will the BCM847 be suitable, and from your results so far the answer seems to be no. I suspect swapping the part out for another from the same batch will yield the same results as I've noticed that discrete transistors from the same batch are getting closer in tolerance matching every year, probably due to improved manufacturing processes. However, the only harm in trying it will be wear and tear on the board. The BCM847 doesn't have a spec for log conformance, most transistors don't, even matched pairs, so it's hard to say it's out of spec. You might want to try a device that has known good log conformance such as the SSM2212 that Ian Fritz suggested or a THAT300 (even though it doesn't have a conformance spec) or the Linear Integrated Systems LS318 (not so easy to get).

One other thought. Have you tried changing R15? I think this will change the operating point on the voltage curve. This doesn't seem like it is the problem you're having but it might be worth checking out.
Back to top
View user's profile Send private message
minisystem



Joined: Nov 16, 2012
Posts: 35
Location: Toronto

PostPosted: Thu Jan 24, 2013 8:39 pm    Post subject: Reply with quote  Mark this post and the followings unread

brock wrote:
I would think you can expect decent tracking over a 5-6 octave range because many builders have achieved this. The question is will the BCM847 be suitable, and from your results so far the answer seems to be no. I suspect swapping the part out for another from the same batch will yield the same results as I've noticed that discrete transistors from the same batch are getting closer in tolerance matching every year, probably due to improved manufacturing processes. However, the only harm in trying it will be wear and tear on the board. The BCM847 doesn't have a spec for log conformance, most transistors don't, even matched pairs, so it's hard to say it's out of spec. You might want to try a device that has known good log conformance such as the SSM2212 that Ian Fritz suggested or a THAT300 (even though it doesn't have a conformance spec) or the Linear Integrated Systems LS318 (not so easy to get).


Well, starting with C1 at 0V I now have pretty good tracking from 1V to 7V with C2 to C8. C1 is off by about 50 cents but C2 is only 10-12 cents off, C8 is dead on and in between it varies from spot on to no more than 5-6 cents. Perhaps this is the best I can achieve with the BCM847.

brock wrote:
One other thought. Have you tried changing R15? I think this will change the operating point on the voltage curve. This doesn't seem like it is the problem you're having but it might be worth checking out.


Might be worth a try. Thanks for your advice. Smile
Back to top
View user's profile Send private message
Display posts from previous:   
Post new topic   Reply to topic Moderators: Scott Stites
Page 1 of 1 [7 Posts]
View unread posts
View new posts in the last week
Mark the topic unread :: View previous topic :: View next topic
 Forum index » DIY Hardware and Software » YuSynth
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