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 » Developers' Corner
Poly synth VCF trimmers
Post new topic   Reply to topic Moderators: DrJustice
Page 1 of 1 [15 Posts]
View unread posts
View new posts in the last week
Mark the topic unread :: View previous topic :: View next topic
Author Message
VA1



Joined: Aug 20, 2018
Posts: 98
Location: Nederland

PostPosted: Thu Aug 01, 2019 8:19 am    Post subject: Poly synth VCF trimmers Reply with quote  Mark this post and the followings unread

How would a Moog VCF look for a poly synth ?
i am intrested in the trimmers, to have all voices the same sound.

I like the moog sound for my polysynth,
i found the simplest moog vcf schematic in the taurus,
it uses 1 ota for comparator + resonance + volume adjust + output,
very clean schematic.
Will it be suited in combination with trimmers for a polysynth ?

Here is a list of what i can think of :
The frequency needs to be the same
The cutoff need to have the same range
The reso need to be the same
The volume need to be the same

I dont want to build a synth with different sounding voices,
and i dont want trimmers, or not so many.

Maybe volume and reso can be done with 1 trimmer ?

Anyone building poly synths ?

thanks
Back to top
View user's profile Send private message
Electric Druid



Joined: Mar 13, 2012
Posts: 44
Location: UK

PostPosted: Fri Oct 11, 2019 11:55 am    Post subject: Reply with quote  Mark this post and the followings unread

Are you using a microprocessor? A lot of polysynths include auto-tuning routines for the filter (Oberheim Xpander is one example - there are many). Ironically since you're looking at Moog filters, the MemoryMoog is a very poor example, since it has a dozen trimmers on everything. But it went out the door half-finished, so we'll give Moog a free pass on that one. I bet the Moog One doesn't need many/any trimmers.

You've got basically two CVs - one for cutoff, one for resonance. You want the same input to give the same output for both CVs. For each of these you need to check various points (each octave is common) and find the DAC output that gives the result you require. Store these values in a tuning table. In-between output values can then be interpolated from the octave values in the table. Go to half-octaves if you can be bothered or are fussy.

No trimmers required! Magic! Just a bit of time doing some firmware....

_________________
Electric Druid Synth and Pedal DIY website
Back to top
View user's profile Send private message
VA1



Joined: Aug 20, 2018
Posts: 98
Location: Nederland

PostPosted: Sat Oct 12, 2019 7:22 am    Post subject: Reply with quote  Mark this post and the followings unread

Thank you for the reply.
I use a 8 bit DAC, and use the full range, i could add another CV for finetuning, thank you.

There is depth from the cutoff signal that might differ,
so it will be 2 finetuning CV, and 2 trimmers for the depth of the reso and cutoff mixed control voltage.
Back to top
View user's profile Send private message
Electric Druid



Joined: Mar 13, 2012
Posts: 44
Location: UK

PostPosted: Sat Oct 12, 2019 10:25 am    Post subject: Reply with quote  Mark this post and the followings unread

An 8 bit DAC is going to produce steps when you modulate it. Unless you're just using it for the basic "Note CV" part of the Filter cutoff CV, and then adding the other modulations (envelope, LFO, whatever) in the analog domain afterwards.

Say the filter covers 10 octaves, 20Hz to 20KHz. That gives you 256/10 = 25.6 values per octave. 25.6 per octave is 25.6/12 = 2.133 per semitone, which is one every 46 cents. Given that the typical figure used for "minimum perceptible pitch step" is 6 cents, that's very definitely an audible step.

_________________
Electric Druid Synth and Pedal DIY website
Back to top
View user's profile Send private message
VA1



Joined: Aug 20, 2018
Posts: 98
Location: Nederland

PostPosted: Sun Oct 13, 2019 5:47 am    Post subject: Reply with quote  Mark this post and the followings unread

I have a input for everything,
the cutoff has a CV on the filter,
the note has a different CV on the filter,
if i have modulation it will also have a seperate CV.
For my mono synth i have a LFO hacked into the 8 bit cutoff,
the bad thing is it is limited to the cutoff range,
the good thing is it is very smooth with 4,7U,
gives fast decay if the LFO was still high on a note retrigger, can give a nice pattern.

The worse thing is i use a PIC16 internal DAC, it is not linear,
for the note i have a 6-bit resistor DAC ( 303 style ),
from 1000 resistors matched 17 per set.
enough sets for a poly synth.

My 40-pin DIP PIC16 has not enough pins for a poly synth.
Still figuring out how to do a poly synth.

My monosynth is almost done.
Back to top
View user's profile Send private message
Electric Druid



Joined: Mar 13, 2012
Posts: 44
Location: UK

PostPosted: Sun Oct 13, 2019 6:35 am    Post subject: Reply with quote  Mark this post and the followings unread

Ok, so the DAC only has to produce the Note CV. That makes life a lot easier, since even if you covered the full MIDI range, you could still get away with a 7-bit DAC assuming it was calibrated correctly! In fact, the Sequential Pro-One did pretty much exactly this.

For a polysynth, you need to think about how you're going to communicate with the voices. The "most vintage" way is to have the central processor producing Gate/Trigger and Note CVs for each voice. This gets less simple when you start adding things like velocity sensitivity and aftertouch and so forth, since the main processor has to produce a lot of CVs.

An alternative scheme would be to include a processor in each voice, and then communicate with the voices using an SPI connection or a UART or something. This is the approach Olivier Gillet took with the Ambika polysynth, for example. While a multi-processor system is always going to be a bit more complicated, it has some advantages, one of which is that each processor is doing less, so can be smaller and cheaper.

One way or another you're going to need to be able to produce a lot of CVs though - start looking at external DACs. I've used the 12-bit MCP 4822 (or MCP4922) which are dual SPI DACs and pretty easy to use and get running.

HTH,
Tom

_________________
Electric Druid Synth and Pedal DIY website
Back to top
View user's profile Send private message
VA1



Joined: Aug 20, 2018
Posts: 98
Location: Nederland

PostPosted: Sun Oct 13, 2019 9:55 am    Post subject: Reply with quote  Mark this post and the followings unread

I am using a 4051 and sample & holds.
Every 4051 (8 CVs) costs 1 more pin.
My monosynth has currently 15 CVs.

Indeed a MCU per voice can be done with UART communication if you individually program each voice with own id number,
SPI costs to much pins compared to UART.
Currently i have a 6 bit (pre VCF) digital chorus getting UART values and sets the internal DAC for a single OTA crossfader ( wich i got inspired by your website ) all working good when tested, almost finished.

Downside is i use 39 row experiment board, a whole voice wont fit in 1 card.
For the mono synth i have :
VCF + exponential converter card
VCA + 6 bit velocity + (pre VCA) distortion card
6-bit DAC card
Chorus card
2 VCO + waveshaper + exponential cards.
Mixer + modulation card.
CV sample & hold card ( for the ones that dont fit in other cards )

This all for a poly synth can not be done, getting to big,
my first poly synth will have DSPIC oscillators for all 6 voices in 1 card.
Hoping to combine the note DAC and VCF, and a 6xVCA card.
Total of 8 cards.

For the next one with analog VCOs i need the multiprocessor setup as you say, still have to count pins for that.
Back to top
View user's profile Send private message
VA1



Joined: Aug 20, 2018
Posts: 98
Location: Nederland

PostPosted: Wed Oct 16, 2019 2:02 am    Post subject: Reply with quote  Mark this post and the followings unread

Electric Druid wrote:
you could still get away with a 7-bit DAC assuming it was calibrated correctly! In fact, the Sequential Pro-One did pretty much exactly this.


How can you know it is calibrated correctly ?
The resistors inside have to be matched better then 0,05% to beat my own measured resistors.
Back to top
View user's profile Send private message
Electric Druid



Joined: Mar 13, 2012
Posts: 44
Location: UK

PostPosted: Wed Oct 16, 2019 2:21 am    Post subject: Reply with quote  Mark this post and the followings unread

VA1 wrote:
Electric Druid wrote:
you could still get away with a 7-bit DAC assuming it was calibrated correctly! In fact, the Sequential Pro-One did pretty much exactly this.


How can you know it is calibrated correctly ?
The resistors inside have to be matched better then 0,05% to beat my own measured resistors.


The Pro-One uses the 6 MSBs of an 8-bit AD558 DAC with a "DAC scale" trimmer. The DAC is reasonable but not perfect. The datasheet tells you everything you need to know about how well you can expect it to be calibrated.

https://www.analog.com/media/en/technical-documentation/data-sheets/AD558.pdf

_________________
Electric Druid Synth and Pedal DIY website
Back to top
View user's profile Send private message
VA1



Joined: Aug 20, 2018
Posts: 98
Location: Nederland

PostPosted: Wed Oct 16, 2019 7:16 am    Post subject: Reply with quote  Mark this post and the followings unread

I found it :

the absolute tolerance of these resistors is
typically ±20%

That is a very false sounding synth.
Back to top
View user's profile Send private message
Electric Druid



Joined: Mar 13, 2012
Posts: 44
Location: UK

PostPosted: Wed Oct 16, 2019 7:48 am    Post subject: Reply with quote  Mark this post and the followings unread

VA1 wrote:
I found it :

the absolute tolerance of these resistors is
typically ±20%

That is a very false sounding synth.


No, that's a mis-reading of the datasheet. +/-0.5 LSB is more like it. The synth sounds fine.

_________________
Electric Druid Synth and Pedal DIY website
Back to top
View user's profile Send private message
blue hell
Site Admin


Joined: Apr 03, 2004
Posts: 24079
Location: The Netherlands, Enschede
Audio files: 278
G2 patch files: 320

PostPosted: Wed Oct 16, 2019 7:52 am    Post subject: Reply with quote  Mark this post and the followings unread

But it is the relative accuracy between the steps which is interesting here, not the absolute one. That, and maybe the full scale accuracy .. although that can be trimmed away with a scaling pot.

edit: what Electric Druid said .. the 1/2 LSB.

_________________
Jan
also .. could someone please turn down the thermostat a bit.
Posted Image, might have been reduced in size. Click Image to view fullscreen.
Back to top
View user's profile Send private message Visit poster's website
VA1



Joined: Aug 20, 2018
Posts: 98
Location: Nederland

PostPosted: Wed Oct 16, 2019 8:48 am    Post subject: Reply with quote  Mark this post and the followings unread

It is something i dont understand.
Datasheets arent my best skill.

LSB is for bits, not ohm ?
Back to top
View user's profile Send private message
Electric Druid



Joined: Mar 13, 2012
Posts: 44
Location: UK

PostPosted: Wed Oct 16, 2019 10:01 am    Post subject: Reply with quote  Mark this post and the followings unread

VA1 wrote:
It is something i dont understand.
LSB is for bits, not ohm ?



Yes. LSB is "least significant bit". The accuracy and/or linearity of a DAC is usually specified in terms of its smallest step - the least significant bit.

Note that just because a DAC has a lot of resolution doesn't mean it has a lot of accuracy. A 12-bit DAC which specifies "+/-8 LSB" is only really accurate to 8 bits. Sometimes that doesn't matter, for example if I was programming a digital LFO - I might want a smoother output, so lots of tiny steps are better than a few big ones, but the accuracy is not that crucial - it'll only distort the LFO wave shape a fraction.

It all depends on the job you're trying to do!

_________________
Electric Druid Synth and Pedal DIY website
Back to top
View user's profile Send private message
VA1



Joined: Aug 20, 2018
Posts: 98
Location: Nederland

PostPosted: Thu Oct 17, 2019 6:47 am    Post subject: Reply with quote  Mark this post and the followings unread

Ok thank you.
How do i compare it with my resistor DAC for note accuracy.
I bought 1000 x 180K 1% resistors, and measured it matching to exact 180,0K
My multimeter is not more accurate then this.

Where do i look to see if the resistors in any DAC are close matched like this ?
The PIC16 internal 8-bit DAC is not linear and useless for note frequency.
Back to top
View user's profile Send private message
Display posts from previous:   
Post new topic   Reply to topic Moderators: DrJustice
Page 1 of 1 [15 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 » Developers' Corner
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