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 
go to the radio page Live at electro-music.com radio 1 Please visit the chat
poster
 Forum index » DIY Hardware and Software » Jürgen Haible designs
Bringing the Vocoder to life
Post new topic   Reply to topic Moderators: fonik, Scott Stites
Page 1 of 7 [167 Posts]
View unread posts
View new posts in the last week
Mark the topic unread :: View previous topic :: View next topic
Goto page: 1, 2, 3, 4, 5, 6, 7 Next
Author Message
jhaible



Joined: May 25, 2007
Posts: 2014
Location: Germany
Audio files: 24

PostPosted: Wed May 11, 2011 1:16 pm    Post subject: Bringing the Vocoder to life
Subject description: this is the debugging / building thread for the adventurous 10+
Reply with quote  Mark this post and the followings unread

This will be your thread for exchanging information about building / optimizing / debugging.

I rely on the group to self-organize, and won't take any leading role here personally. Smile

JH.

_________________
"I tell you the truth, if anyone says to this mountain, 'Go, throw yourself into the sea,' and does not doubt in his heart but believes that what he says will happen, it will be done for him. Therefore I tell you, whatever you ask for in prayer, believe that you have received it, and it will be yours." (Mk 11,23f)
Back to top
View user's profile Send private message Send e-mail Visit poster's website
The Bad Producer



Joined: Mar 08, 2009
Posts: 282
Location: The Manhole

PostPosted: Wed May 11, 2011 3:18 pm    Post subject: Reply with quote  Mark this post and the followings unread

If I may be so bold as to ask the first question!

How hard/pointless do you guys think it would be to hand-match all the caps, as opposed to getting 1%? Would this even be viable in this situation?

(I've got a fair amount of time on my hands at the moment!)

Charlie

_________________
http://loudestwarning.tumblr.com/
Back to top
View user's profile Send private message Visit poster's website
/mr



Joined: Aug 05, 2007
Posts: 223
Location: Elektron City, Sweden
Audio files: 1

PostPosted: Wed May 11, 2011 4:25 pm    Post subject: Reply with quote  Mark this post and the followings unread

The Bad Producer wrote:
How hard/pointless do you guys think it would be to hand-match all the caps, as opposed to getting 1%? Would this even be viable in this situation?

Good question (I hope) - I was about to ask the same thing. Smile I think it would be doable. Some discussions about this from last year in this post and Jürgen's answer that follows: http://electro-music.com/forum/post-301322.html#301322

What kept me from asking is that I didn't find any really good alternatives... i.e., considerably cheaper polypropylene 2%-5%-10% capacitors with both the right lead spacing (5 mm) and small enough size (max 8½ mm). Does anybody else have more luck?

(Btw, it would be bad to choose caps with more aging problems - so I asked an R&D representative at Vishay Capacitors about different tolerances vs long-term precision, and he says that higher-tolerance and lower-tolerance caps have the same long-term behaviour since their process is the same. The lower-tolerance ones are just closer to the nominal value.)
Back to top
View user's profile Send private message
diablojoy



Joined: Sep 07, 2008
Posts: 809
Location: melbourne australia
Audio files: 11

PostPosted: Wed May 11, 2011 4:31 pm    Post subject: Reply with quote  Mark this post and the followings unread

150 each of 10n + 47n to within 1% you would need to buy thousands
of each at 5% values with no guarantee of complete success i guess it depends on what value you rate your time at but cost wise probably wouldn't work out that much cheaper
you could check out rush on line capacitors presently at 0.75 GBP each for both values or 225.00 GBP + postage the price only recently went up from 0.47 Sad doh! they also have the 10n value in polystyrene in a 'd' tolerance 0.5% at 0.75 GBP though not certain they will fit the board at 7mm x 7mm x 13 mm

cheers
Back to top
View user's profile Send private message
/mr



Joined: Aug 05, 2007
Posts: 223
Location: Elektron City, Sweden
Audio files: 1

PostPosted: Wed May 11, 2011 4:48 pm    Post subject: Reply with quote  Mark this post and the followings unread

diablojoy wrote:
150 each of 10n + 47n to within 1% you would need to buy thousands of each at 5% values

1% precision is not necessary - what is important is matching the 8 caps in each filter, and not changing cap value too much between adjacent filters. See my link above for a description of how matching & selection can be done. For the 144 caps you could probably buy 200 and follow the routine there. Smile
Back to top
View user's profile Send private message
diablojoy



Joined: Sep 07, 2008
Posts: 809
Location: melbourne australia
Audio files: 11

PostPosted: Wed May 11, 2011 8:07 pm    Post subject: Reply with quote  Mark this post and the followings unread

Quote:
1% precision is not necessary - what is important is matching the 8 caps in each filter, and not changing cap value too much between adjacent filters.




ah yes i see my bad Embarassed , i should have realised about matching the cap's firstly in groups of 8. I had completely forgotten those posts must go back over all of them again
Back to top
View user's profile Send private message
decaying.sine



Joined: Aug 31, 2009
Posts: 92
Location: New Haven, CT, USA

PostPosted: Thu May 12, 2011 10:30 am    Post subject: Reply with quote  Mark this post and the followings unread

Would there be any appreciable benefit of sorting and matching specific 1% resistors into identical ohm groups and using those in certain areas of, say, the 8 pole filters (or other places)? In other words would this improve performance?
Back to top
View user's profile Send private message
StephenGiles



Joined: Apr 17, 2006
Posts: 507
Location: England

PostPosted: Thu May 12, 2011 11:09 am    Post subject: Reply with quote  Mark this post and the followings unread

decaying.sine wrote:
Would there be any appreciable benefit of sorting and matching specific 1% resistors into identical ohm groups and using those in certain areas of, say, the 8 pole filters (or other places)? In other words would this improve performance?


I'm afraid I can see some divorce cases arising out of all this Crying or Very sad Crying or Very sad Crying or Very sad
Back to top
View user's profile Send private message
/mr



Joined: Aug 05, 2007
Posts: 223
Location: Elektron City, Sweden
Audio files: 1

PostPosted: Fri May 13, 2011 12:47 am    Post subject: Reply with quote  Mark this post and the followings unread

decaying.sine wrote:
Would there be any appreciable benefit of sorting and matching specific 1% resistors into identical ohm groups and using those in certain areas of, say, the 8 pole filters (or other places)? In other words would this improve performance?

Matching capacitors or resistors is beneficial to the limit of the precision of the other components. I.e., if the capacitors and the construction's value selection already add up to more than 1% error compared to the theoretically perfect filter, it's probably no use matching the 1% resistors.

Matching capacitors and resistors in absurdum might be beneficial only if the R & C values in the construction are perfectly chosen. I'd expect them to be a bit off from perfect though, but I guess only Jürgen can answer that.

Personally I don't think that it is worthwhile trying to achieve much more than 1% precision in the components of these filters. After all, the neighbouring filter channels are 20% apart, so a total deviation of, say, 3% is not that much. (Unless the phase angles in the transitions between channels are critical... they might or might not be.) Smile

Speaking of the channel filters - they seem to be built from
2-pole LPF + 2-pole HPF + 2-pole LPF + 2-pole HPF.
The HPF construction looks standard, but I can't seem to quickly find this LFP construction elsewhere. Is it common? Where does it come from? Smile

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



Joined: May 25, 2007
Posts: 2014
Location: Germany
Audio files: 24

PostPosted: Fri May 13, 2011 2:56 am    Post subject: Reply with quote  Mark this post and the followings unread

The design is relatively insensitive to component tolerances - that's why 1% is good enough.
That was EMS's method. Sennheiser achieved similarly steep filters with only 6 poles, but required much better component tolerances.
Synton used a similar approach - higher filter order, lower Q of individual stages.
The LPF is quite unique ineed - I've never seen it usesd othe rthan from EMS.
BTW, a single LPF stage is a resonant, 2pole, 6dB/Octave filter.

JH.

_________________
"I tell you the truth, if anyone says to this mountain, 'Go, throw yourself into the sea,' and does not doubt in his heart but believes that what he says will happen, it will be done for him. Therefore I tell you, whatever you ask for in prayer, believe that you have received it, and it will be yours." (Mk 11,23f)
Back to top
View user's profile Send private message Send e-mail Visit poster's website
/mr



Joined: Aug 05, 2007
Posts: 223
Location: Elektron City, Sweden
Audio files: 1

PostPosted: Fri May 13, 2011 11:07 am    Post subject: Reply with quote  Mark this post and the followings unread

jhaible wrote:
BTW, a single LPF stage is a resonant, 2pole, 6dB/Octave filter.

Only 6 dB/oct, not the ordinary 12?
Increased steepness does make less difference in HPF than in LPF... but anyway... why? Laughing

PS: Can the "preliminary BOM" on http://www.jhaible.de/vocoder/living_vocoder.html be considered "official BOM" for us pre-series builders? I see that a small error has been corrected since my first download of the document, so I guess some kind of revision check has been made. It's important to know if it's alright to trust it for the component orders coming up, or if it needs to be checked for more errors before we order. Smile
Back to top
View user's profile Send private message
jhaible



Joined: May 25, 2007
Posts: 2014
Location: Germany
Audio files: 24

PostPosted: Fri May 13, 2011 11:19 am    Post subject: Reply with quote  Mark this post and the followings unread

/mr wrote:
jhaible wrote:
BTW, a single LPF stage is a resonant, 2pole, 6dB/Octave filter.

Only 6 dB/oct, not the ordinary 12?
Increased steepness does make less difference in HPF than in LPF... but anyway... why? Laughing


The resonant peak is nore important than the slopes here, as long as there are _any_ slopes. That's why they gat away whith it.
But why did they *do* it?
I thought it was obvious from the first glance on the circuit. And I think it's probably the most brilliant choice EMS made. Still no idea? Smile
(I'll tell you, of course, but don't want to spoil the grin on your face when you find out for yourself. So?)

/mr wrote:

PS: Can the "preliminary BOM" on http://www.jhaible.de/vocoder/living_vocoder.html be considered "official BOM" for us pre-series builders? I see that a small error has been corrected since my first download of the document, so I guess some kind of revision check has been made. It's important to know if it's alright to trust it for the component orders coming up, or if it needs to be checked for more errors before we order. Smile


It's from a forum member who is also among th efirst 10. I haven't personally checked it. Consider it as the first step of the group effort.

JH.

_________________
"I tell you the truth, if anyone says to this mountain, 'Go, throw yourself into the sea,' and does not doubt in his heart but believes that what he says will happen, it will be done for him. Therefore I tell you, whatever you ask for in prayer, believe that you have received it, and it will be yours." (Mk 11,23f)
Back to top
View user's profile Send private message Send e-mail Visit poster's website
/mr



Joined: Aug 05, 2007
Posts: 223
Location: Elektron City, Sweden
Audio files: 1

PostPosted: Sat May 14, 2011 9:20 am    Post subject: Reply with quote  Mark this post and the followings unread

jhaible wrote:
/mr wrote:
jhaible wrote:
BTW, a single LPF stage is a resonant, 2pole, 6dB/Octave filter.
Only 6 dB/oct, not the ordinary 12?
Increased steepness does make less difference in HPF than in LPF... but anyway... why? Laughing
The resonant peak is nore important than the slopes here, as long as there are _any_ slopes. That's why they gat away whith it.
But why did they *do* it?
I thought it was obvious from the first glance on the circuit. And I think it's probably the most brilliant choice EMS made. Still no idea? Smile
(I'll tell you, of course, but don't want to spoil the grin on your face when you find out for yourself. So?)

Equal-valued capacitors? Smile
Back to top
View user's profile Send private message
decaying.sine



Joined: Aug 31, 2009
Posts: 92
Location: New Haven, CT, USA

PostPosted: Sat May 14, 2011 12:14 pm    Post subject: Reply with quote  Mark this post and the followings unread

I got my 1% caps from Newark already. They shipped very quickly. The 47nf shipped with part blue and part gray color. Now this just upsets my delicate sensibilities. Wink
Back to top
View user's profile Send private message
jhaible



Joined: May 25, 2007
Posts: 2014
Location: Germany
Audio files: 24

PostPosted: Sat May 14, 2011 12:15 pm    Post subject: Reply with quote  Mark this post and the followings unread

/mr wrote:
jhaible wrote:
/mr wrote:
jhaible wrote:
BTW, a single LPF stage is a resonant, 2pole, 6dB/Octave filter.
Only 6 dB/oct, not the ordinary 12?
Increased steepness does make less difference in HPF than in LPF... but anyway... why? Laughing
The resonant peak is nore important than the slopes here, as long as there are _any_ slopes. That's why they gat away whith it.
But why did they *do* it?
I thought it was obvious from the first glance on the circuit. And I think it's probably the most brilliant choice EMS made. Still no idea? Smile
(I'll tell you, of course, but don't want to spoil the grin on your face when you find out for yourself. So?)

Equal-valued capacitors? Smile


Precisely.
Q factor is set by resistors, instead of capacitors. So all caps can be equal, and bought in bulk.

JH.

_________________
"I tell you the truth, if anyone says to this mountain, 'Go, throw yourself into the sea,' and does not doubt in his heart but believes that what he says will happen, it will be done for him. Therefore I tell you, whatever you ask for in prayer, believe that you have received it, and it will be yours." (Mk 11,23f)
Back to top
View user's profile Send private message Send e-mail Visit poster's website
jhaible



Joined: May 25, 2007
Posts: 2014
Location: Germany
Audio files: 24

PostPosted: Sat May 14, 2011 12:21 pm    Post subject: Reply with quote  Mark this post and the followings unread

BTW, the Korg MS-20 HPF is just the same principle. But here you want equal resistors (for changing them equally), and thus set the Q by capacitor values. So you get a 6dB/Oct slope, but still a resonant and even self oscillating HPF.

JH.

_________________
"I tell you the truth, if anyone says to this mountain, 'Go, throw yourself into the sea,' and does not doubt in his heart but believes that what he says will happen, it will be done for him. Therefore I tell you, whatever you ask for in prayer, believe that you have received it, and it will be yours." (Mk 11,23f)
Back to top
View user's profile Send private message Send e-mail Visit poster's website
decaying.sine



Joined: Aug 31, 2009
Posts: 92
Location: New Haven, CT, USA

PostPosted: Sun May 22, 2011 2:51 pm    Post subject: Reply with quote  Mark this post and the followings unread

I wanted to let folks know that I am building a bill of materials to check against Oscar's BOM. I believe that I saw Jürgen post that this would be a good idea. I am almost done at this point. I'll also work on mouser order codes (or other sources) for the parts as well. This would probably only work for United States builders.

I started typing out all the parts from the schematics and quickly fell asleep. Then I opened up the PDF's and copy every bit of text, pasted it into a text document, and then just adjusted the spaces with tab delimits. It takes about 3 minutes per schematic (about 27 of them if I recall). I am sure everyone else already had this figured out long ago, but it was a Idea moment for me!
Back to top
View user's profile Send private message
The Bad Producer



Joined: Mar 08, 2009
Posts: 282
Location: The Manhole

PostPosted: Sun May 22, 2011 3:52 pm    Post subject: Reply with quote  Mark this post and the followings unread

Nice one decaying.sine, seeing as you've done the hard work Very Happy I'd be up for translating your BOM into a UK friendly one with parts number from Rapid/RS/Farnell etc...

Charlie

(not sure whereabouts all ten peeps are based, I'm in UK)

_________________
http://loudestwarning.tumblr.com/
Back to top
View user's profile Send private message Visit poster's website
decaying.sine



Joined: Aug 31, 2009
Posts: 92
Location: New Haven, CT, USA

PostPosted: Sun May 22, 2011 7:39 pm    Post subject: Reply with quote  Mark this post and the followings unread

This may have been mentioned before but I'll post anyways. I apologize if it has already been reported. These duplicate resistor identifiers came up when making my BOM to compare to Oscar's work.

The duplicates...
Synthesis Amp, Compander, Noise, and S-Generator Schematic
R2663 1k
R2663 1k

Synthesis Amp, Compander, Noise, and S-Generator Schematic
R2634 10R
R2634 100K

Slew Freeze Schematic
R2521 10k
R2521 27k

Silence Bridging Filterbank Control Schematic
U2402 is CD4016 on schematic. On Oscar's BOM there is a CD4066 but no CD4016. They're both Quad Bilateral Switches so either is okay, but 4066 better for audio switching? This may have been addressed earlier, if so, nothing to see here Smile
Back to top
View user's profile Send private message
jhaible



Joined: May 25, 2007
Posts: 2014
Location: Germany
Audio files: 24

PostPosted: Mon May 23, 2011 9:50 pm    Post subject: Reply with quote  Mark this post and the followings unread

4016 and 4066 will both work. They are used as switches (B, D) and as Inverter (A) and Schmitt trigger (C) in the AR envelope generator of the Silence Bridging.

JH.

_________________
"I tell you the truth, if anyone says to this mountain, 'Go, throw yourself into the sea,' and does not doubt in his heart but believes that what he says will happen, it will be done for him. Therefore I tell you, whatever you ask for in prayer, believe that you have received it, and it will be yours." (Mk 11,23f)
Back to top
View user's profile Send private message Send e-mail Visit poster's website
JoeMorris



Joined: Apr 26, 2009
Posts: 161
Location: Brighton

PostPosted: Tue May 24, 2011 4:55 am    Post subject: Reply with quote  Mark this post and the followings unread

Quote:
Nice one decaying.sine, seeing as you've done the hard work I'd be up for translating your BOM into a UK friendly one with parts number from Rapid/RS/Farnell etc...

Charlie

(not sure whereabouts all ten peeps are based, I'm in UK)
_________________


I often find mouser UK cheaper than farnell if you're ordering enough to get free shipping... (takes a couple of days longer of course)
Back to top
View user's profile Send private message
The Bad Producer



Joined: Mar 08, 2009
Posts: 282
Location: The Manhole

PostPosted: Tue May 24, 2011 6:19 am    Post subject: Reply with quote  Mark this post and the followings unread

JoeMorris wrote:

I often find mouser UK cheaper than farnell if you're ordering enough to get free shipping... (takes a couple of days longer of course)


Cool, I'll have to look into that, thanks!

_________________
http://loudestwarning.tumblr.com/
Back to top
View user's profile Send private message Visit poster's website
JoeMorris



Joined: Apr 26, 2009
Posts: 161
Location: Brighton

PostPosted: Tue May 24, 2011 6:22 am    Post subject: Reply with quote  Mark this post and the followings unread

prefer the usability of their website to farnell's too... (auto-BOM import etc..)
Back to top
View user's profile Send private message
/mr



Joined: Aug 05, 2007
Posts: 223
Location: Elektron City, Sweden
Audio files: 1

PostPosted: Tue May 24, 2011 6:48 am    Post subject: Reply with quote  Mark this post and the followings unread

JoeMorris wrote:
prefer the usability of their website to farnell's too... (auto-BOM import etc..)

Both have their pros and cons. Farnell's shopping cart have seemed generally less cooperative and less stable for me during the years, sometimes throwing it all away after deciding my session was over, etc. On the other hand, Farnell's product filtering parameters include Lead Spacing for capacitors, whereas Mouser's for some reason don't... I found this to be an unexpectedly important comfortability matter in this big vocoder project. Smile

Speaking of Farnell, a couple of years ago they offered resistors in high-volume pricing even if the resistance values were different. I.e., if I ordered 10 pc each of 50 different values, I got them for the 500pc price. Very nice! I don't see that feature any more... does anybody else offer this kind of pricing?
Back to top
View user's profile Send private message
decaying.sine



Joined: Aug 31, 2009
Posts: 92
Location: New Haven, CT, USA

PostPosted: Tue May 24, 2011 6:57 am    Post subject: Reply with quote  Mark this post and the followings unread

I have all the resistors done for the BOM Mouser order. I included an extra column that recommends buying the number of pieces specified for the project or the next available price break--depending on the one that is cheaper. As you might guess sometimes it is cheaper to buy at the next available price break than the actual amount that you need.

I am working on caps now, but I'll wait and post everything together.

Also, my BOM of materials is identical to Oscar's original BOM. However, I am not 100% done checking yet. My suspicion is that his is accurate and it also has lead spacing for caps and that is awesome!
Back to top
View user's profile Send private message
Display posts from previous:   
Post new topic   Reply to topic Moderators: fonik, Scott Stites
Page 1 of 7 [167 Posts]
View unread posts
View new posts in the last week
Goto page: 1, 2, 3, 4, 5, 6, 7 Next
Mark the topic unread :: View previous topic :: View next topic
 Forum index » DIY Hardware and Software » Jürgen Haible designs
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