electro-music.com   Dedicated to experimental electro-acoustic
and electronic music
 
    Front Page  |  Articles  |  Radio
 |  Media  |  Forum  |  Wiki  |  Links  |  Store
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
303 filter ... help needed
Post new topic   Reply to topic Moderators: jksuperstar, Scott Stites, Uncle Krunkus
Page 1 of 1 [16 Posts]
View unread posts
View new posts in the last week
Mark the topic unread :: View previous topic :: View next topic
Author Message
3phase



Joined: Jul 27, 2004
Posts: 1169
Location: Berlin
Audio files: 13
G2 patch files: 141

PostPosted: Sun Mar 05, 2006 7:50 am    Post subject: 303 filter ... help needed Reply with quote  Mark this post and the followings unread

Hello,
I try to patch the 303 within the nord modular...biggest problem..
to get the filter wright.

Has any of you more electronic minded guys any insights how the 303 filter actually works in detail? when i look on the circuit diagram it looks like 4 pole design where one pole is one octave up...
maybe i am wrong... i only can read schematics on a pretty surfaced level.

But i need more detail information... how is this filter working..how is feedback applied? any additional filterings?

the waveform gets altred drastical by passing the filter...
a mid boost is happening ...for example...

Any small detail information can be helpfull...

Also when you know somebody that have build the 303 filter and might have studied in detail it would be nice if you could point me there..

thanks for reading this far Wink
Sven

http://machines.hyperreal.org/manufacturers/Roland/TB-303/schematics/
Back to top
View user's profile Send private message Send e-mail Visit poster's website
deknow



Joined: Sep 15, 2004
Posts: 1307
Location: Leominster, MA (USA)
G2 patch files: 15

PostPosted: Sun Mar 05, 2006 12:45 pm    Post subject: Reply with quote  Mark this post and the followings unread

sven, did you find this?
http://machines.hyperreal.org/manufacturers/Roland/TB-303/info/roland.TB-303.secrets.txt

...some things about the filter envelope being bipolar.

deknow
Back to top
View user's profile Send private message Send e-mail
3phase



Joined: Jul 27, 2004
Posts: 1169
Location: Berlin
Audio files: 13
G2 patch files: 141

PostPosted: Sun Mar 05, 2006 12:53 pm    Post subject: Reply with quote  Mark this post and the followings unread

yes..but its not accurat best source i found sofar are the pages of the devilfish maker Robin Whitle. He however dont likes to tell everything...
and he is not touching the filter so there is no deeper information on the circuit design. it needs an expert electronic guy to tell what really goes on there. I am not sure if one pole is one octave up... just one capaciator on the ladder just has half the szise...
Back to top
View user's profile Send private message Send e-mail Visit poster's website
Scott Stites
Janitor
Janitor


Joined: Dec 23, 2005
Posts: 4127
Location: Mount Hope, KS USA
Audio files: 96

PostPosted: Mon Mar 06, 2006 7:33 am    Post subject: Reply with quote  Mark this post and the followings unread

The TB303 filter has a 3 pole, 18 dB filter.

If you've got Power Point, this is a detailed explanation for the filter:

http://wisl.ece.cornell.edu/roth/research/TB-303_Filters.ppt

Cheers,
Scott
Back to top
View user's profile Send private message
3phase



Joined: Jul 27, 2004
Posts: 1169
Location: Berlin
Audio files: 13
G2 patch files: 141

PostPosted: Mon Mar 06, 2006 10:44 am    Post subject: Reply with quote  Mark this post and the followings unread

Thanx Scott,
that document looks like an interesting resource...

Sven
Back to top
View user's profile Send private message Send e-mail Visit poster's website
Scott Stites
Janitor
Janitor


Joined: Dec 23, 2005
Posts: 4127
Location: Mount Hope, KS USA
Audio files: 96

PostPosted: Mon Mar 06, 2006 10:58 am    Post subject: Reply with quote  Mark this post and the followings unread

Hey Sven,

I hope you can understand more of it than I can Very Happy

Cheers,
Scott
Back to top
View user's profile Send private message
ian-s



Joined: Apr 01, 2004
Posts: 2654
Location: Auckland, New Zealand
Audio files: 42
G2 patch files: 625

PostPosted: Mon Mar 06, 2006 11:41 am    Post subject: Reply with quote  Mark this post and the followings unread

Impulse response graph to show effect of decay? I wonder what grade this got.
Back to top
View user's profile Send private message
3phase



Joined: Jul 27, 2004
Posts: 1169
Location: Berlin
Audio files: 13
G2 patch files: 141

PostPosted: Mon Mar 06, 2006 5:53 pm    Post subject: Reply with quote  Mark this post and the followings unread

Scott Stites wrote:
Hey Sven,

I hope you can understand more of it than I can Very Happy

Cheers,
Scott


no... but the note about rising low boost on accented notes is interesting...
i was lookin thru the internet searching for statements regarding the 303...
some is contradictionary... and it seems that the total technical truth about the 303 never got really explained in public sofar.
Back to top
View user's profile Send private message Send e-mail Visit poster's website
3phase



Joined: Jul 27, 2004
Posts: 1169
Location: Berlin
Audio files: 13
G2 patch files: 141

PostPosted: Mon Mar 06, 2006 6:27 pm    Post subject: Reply with quote  Mark this post and the followings unread

I found some other statements on the net regarding the 303 and its filter...
I wondfer if some of this statements are false ...
if somebody can clearly state as such regarding the schematics it would be nice to let us know.

I also would like to know what the difference between a diode ladder and a transistor ladder is soundwise... and why roland to choose transistors to build a diode ladder filter...
Questions over question s :-/

The found statements: various authors cutup

1) Non-oscillating filter which has CV control over resonance (and large
amounts of resonance at that).

2) An Attack/Release Envelope which is bipolar (or some sort of level
shifter to make a unipolar envelope bi.

Note that the cutoff is lowered at the end of decay - this is due to the
infamous "Gimmick" circuit, 3 transistors which pull this little trick.
WHY, I don't know. Needless to say, it's a pretty feature that you'll have
trouble duplicating w/o a modular. On a typical AR envelope circuit,
you'll want to add a -5VDC offset to the normal 0-10V output.

As for the filter, in Barry Kleins book as well as Electronotes is a Dual OTA filter. This is a simple filter with BP/LP/HP/N outputs. A simple substitution of one resistor in the resonance section turns it into a LI'l monster

The TRICK seems to be the 303's filter has INSANE amounts of resonance
but DOES NOT OSCILLATE. I suspect distortion in the feedback loop has
something to do w/ it's character, but I haven't verified that.

The 303 filter is a diode-ladder filter (only that it uses transistors that
are
connected to act as diodes). So it is more similar to the EMS AKS or
Minisonic design than to the Moog in some way. It is a 4-pole filter; but
due to the pole location, which is quite different from Moog- or
CEM-filters,
it has more of a 18dB/Oct than the 24db/Oct slope in the interesting
range. (Of course, in the infinity, every 4-pole low pass is 24dB).

It is a 4-pole filter, with 3 poles the same and the 4th pole about 2.5
times higher in
frequency than the others. So, it will more resemble a 18dB/oct response.

With the 303 however the first cap has a value half that of the others.
So the first pole is a octave higher.
Because we have a LOW pass filter here this first pole does not contribute
to the filter function.
Not sure how the first pole contributes to the resonance point, with some
other filters the resonant frequency is determent by the last pole only.
However I duno how this is with the moog ladder.

The difference between the 303 ladder and the Moog ladder is how the
resonance feedback is applied. On the 303 the feedback is applied to the
transistor that drives the current through the ladder, which gives extra AM
modulation in the filter, causing that typical 303 filter distortion sound.
On the Moog ladder the resonance feedback is applied to one of the two
inputs of the ladder itself, hence the 'clean' resonance sound.
Back to top
View user's profile Send private message Send e-mail Visit poster's website
ian-s



Joined: Apr 01, 2004
Posts: 2654
Location: Auckland, New Zealand
Audio files: 42
G2 patch files: 625

PostPosted: Mon Mar 06, 2006 9:19 pm    Post subject: Reply with quote  Mark this post and the followings unread

3phase wrote:
The found statements: various authors cutup


1) Non-oscillating filter which has CV control over resonance (and large
amounts of resonance at that).
Rubbish, no CV for resonance and it does oscillate, you just cant switch off the VCO to here it.

2) An Attack/Release Envelope which is bipolar (or some sort of level
shifter to make a unipolar envelope bi.

Note that the cutoff is lowered at the end of decay - this is due to the
infamous "Gimmick" circuit, 3 transistors which pull this little trick.
WHY, I don't know. Needless to say, it's a pretty feature that you'll have
trouble duplicating w/o a modular. On a typical AR envelope circuit,
you'll want to add a -5VDC offset to the normal 0-10V output.

seems to be verified by the service manual


The 303 filter is a diode-ladder filter (only that it uses transistors that
are
connected to act as diodes). So it is more similar to the EMS AKS or
Minisonic design than to the Moog in some way. It is a 4-pole filter; but
due to the pole location, which is quite different from Moog- or
CEM-filters,
it has more of a 18dB/Oct than the 24db/Oct slope in the interesting
range. (Of course, in the infinity, every 4-pole low pass is 24dB).

It is a 4-pole filter, with 3 poles the same and the 4th pole about 2.5
times higher in
frequency than the others. So, it will more resemble a 18dB/oct response.

With the 303 however the first cap has a value half that of the others.
So the first pole is a octave higher.
Because we have a LOW pass filter here this first pole does not contribute
to the filter function.
Not sure how the first pole contributes to the resonance point, with some
other filters the resonant frequency is determent by the last pole only.
However I duno how this is with the moog ladder.

sounds ok, the transisters are wired different and the moog has a resister in each pole. I think diode ladders just distort more. The 1st pole shift gives the slope a 'knee' so 18dB for first octave, 24dB after. I suspect they wanted a brighter tone without reduced resonance.

The difference between the 303 ladder and the Moog ladder is how the
resonance feedback is applied. On the 303 the feedback is applied to the
transistor that drives the current through the ladder, which gives extra AM
modulation in the filter, causing that typical 303 filter distortion sound.
On the Moog ladder the resonance feedback is applied to one of the two
inputs of the ladder itself, hence the 'clean' resonance sound.

not according to the schematic
Back to top
View user's profile Send private message
ian-s



Joined: Apr 01, 2004
Posts: 2654
Location: Auckland, New Zealand
Audio files: 42
G2 patch files: 625

PostPosted: Tue Mar 07, 2006 2:14 am    Post subject: Reply with quote  Mark this post and the followings unread

Here are some patching exercises.
I don’t have a 303 for tuning, and it has been a long time since I played with one (early 80’s?, I hated the girly handbag carry case btw)

The filter is a variation on one of Robs, I use the classic for the 18dB stage to take advantage of the built in distortion.

The mod envelope switches to 200ms decay on accents, its mod level gets a proportional negative bias as it increases. The mod env is also via the accent amount, added to the main vca and through a home made slew, to the filter frequency.

The slew circuit’s time is inversely proportional to the resonance setting.

The square wave has slew because it is created by saturating the crap out of the sawtooth. the amount of slew varies with frequency and a small constant is added to prevent the square from being to perfect.

The simple sequencer in the FX area uses midi CC to send note, gate, accent and glide. This could be relocated to another slot if you wanted patterns with independent tweaking.


tb303-2002.pch2
 Description:
A fish called "Beelzebub"

Download
 Filename:  tb303-2002.pch2
 Filesize:  4.81 KB
 Downloaded:  1170 Time(s)

Back to top
View user's profile Send private message
3phase



Joined: Jul 27, 2004
Posts: 1169
Location: Berlin
Audio files: 13
G2 patch files: 141

PostPosted: Tue Mar 07, 2006 7:17 am    Post subject: Reply with quote  Mark this post and the followings unread

g2ian wrote:
3phase wrote:
The found statements: various authors cutup



Thanks.. i ve updated the main thread in the G2 section..
i also will check your patch later...
the classical filter distorts internally? interesting...
Back to top
View user's profile Send private message Send e-mail Visit poster's website
3phase



Joined: Jul 27, 2004
Posts: 1169
Location: Berlin
Audio files: 13
G2 patch files: 141

PostPosted: Tue Mar 07, 2006 7:29 am    Post subject: Reply with quote  Mark this post and the followings unread

To g2ian and others

http://machines.hyperreal.org/manufacturers/Roland/TB-303/schematics/roland.TB-303.schem-8.gif

could you retranslate this back to japanese? Wink

I am wondering what the authors want to tell us about the difference of the 303 envelope against "normal" designs...

Do i have to understand it this way that there is just a triger pulse that defines the maximum point with a constant rate dicharging elko?

thanks for any info..

Sven
Back to top
View user's profile Send private message Send e-mail Visit poster's website
3phase



Joined: Jul 27, 2004
Posts: 1169
Location: Berlin
Audio files: 13
G2 patch files: 141

PostPosted: Wed Mar 08, 2006 10:23 am    Post subject: Reply with quote  Mark this post and the followings unread

My newest idea is that the 303 has 2 filters working...one resonating the other not...

Can anybody refer on this by cheking the circuit diagram?
Am i on the wrong path?

thanks
Sven
Back to top
View user's profile Send private message Send e-mail Visit poster's website
ian-s



Joined: Apr 01, 2004
Posts: 2654
Location: Auckland, New Zealand
Audio files: 42
G2 patch files: 625

PostPosted: Wed Mar 08, 2006 12:04 pm    Post subject: Reply with quote  Mark this post and the followings unread

3phase wrote:
To g2ian and others

http://machines.hyperreal.org/manufacturers/Roland/TB-303/schematics/roland.TB-303.schem-8.gif

could you retranslate this back to japanese? Wink

I am wondering what the authors want to tell us about the difference of the 303 envelope against "normal" designs...

Do i have to understand it this way that there is just a triger pulse that defines the maximum point with a constant rate dicharging elko?

thanks for any info..

Sven



That description covers the env.mod to VCF expo converter connection. If the diagram is to scale, adding 1/3 negative offset to envelope signal should do the trick. The accented envelopes also get connected to the same point but via a lower value resister so the accent has about twice the max sweep of the env mod?
The accent mod env signal (with no negative bias) is also added to the VCA.
Back to top
View user's profile Send private message
3phase



Joined: Jul 27, 2004
Posts: 1169
Location: Berlin
Audio files: 13
G2 patch files: 141

PostPosted: Wed Mar 08, 2006 2:58 pm    Post subject: Reply with quote  Mark this post and the followings unread

I have understood that when i have read the manual in the first place... but there is more about it...
My emulation from 2 years ago was based on this idea about the 303..
And the info on Robin Whitles site...he knows probably more about the 303 as anybody else but his opinion is that he has revealed enough of the 303 secrets allready and that i should sit with an oscilloscope on one finding my own clues...I however cant do that to a lent machine...
So lets see how far we get with listening and tghe opinions of other synth specialists

I like to win this challange... not just for me..it will be good promotion for the G2 when its shown that a digital synth can do it.

After all the tests I am sure its possible to get close enough..
Its just necessary to get the wright structure patched... It maybe wont sound 100% like a 303 but like an elastic nord...thats good enough for me...just...sofar all 303 emulations on the nord are not going deep enough.
They behave rather stiff because whats really going on in a 303 is sofar not published in detail...its like a puzzle with lots of missinformation and overseen details.
The envelope thing is different to a fixed offset...seems to react dynamic...
or the envelope has constant rate and therefore needs longer when the fixed offset is down... but when the offset is lowering as more as more the envelope gets down...it would also appear to get alonger decay time..

I am rather curious why they state that theire envelope is differnet to others where the shape of the envelope is performed regardless of its input... Is this just babelfishing or do we really have here a constant rate envelope while the glide is constant time? just the opposite like its usually done?
Also some mysteries about the filter question..i recorded some audio snippets direct after the filter that sound indeed strange... either there are 2 filters working in parallel or the resonance is somehow not affecting all poles...
Have you the electronic knowledge to see such things in the circuit diagram?
I started to add some audio examples to the 303 cloning thread in the nord modular forum.
Thanx anyway for your input allready..and when you have the knowledge..the complete circuit diagram is also to find on at the posted link.

Sven
Back to top
View user's profile Send private message Send e-mail Visit poster's website
Display posts from previous:   
Post new topic   Reply to topic Moderators: jksuperstar, Scott Stites, Uncle Krunkus
Page 1 of 1 [16 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
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
e-m mkii

Please support our site. If you click through and buy from
our affiliate partners, we earn a small commission.


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