[sdiy] modern AD/DA chips?

Jerry Gray-Eskue jerryge at cableone.net
Wed May 6 18:42:50 CEST 2009


I have used a lot of D2A and A2D converters in industrial test equipment.
You can get high precision out of just about any converter. For high
precision you will however have to design in a 0 Offset (null circuit) and a
Span adjust. In addition the Reference voltage for your span adjust needs to
be very stable and insensitive to temperature.

The most important specification of the converter is how well its output
tracks (linear and no missing codes) If the output deviates from the ideal
you are stuck with that error, or have to try to recurve it on the software
side. It is also possible to make the Null and Span adjusts on the Software
side but any software adjustment is limited to the value of the LSB (least
significant Bit) for the "tweak" value.

Using an 8 bit converter you will have 256 steps ( 0-255), for a 10 Octave
(1V per Octave) range Midi to CV you need 12 semitones per Octave * 10
Octaves or 120 steps, so you can set this up as 2 bits per semitone for 240
steps with 16 steps left over. Now the Trick is that the value of your LSB
Must be dead on or you will Know it is off. With 2 bits per semitone a 1 bit
error is 50 cents.

You reference voltage for an 8 BIT DAC using an output Gain stage of 4X
would need to be 2.667 . This setup would output 0 - 10.669 at 0.0417 per
Bit or 12 Bits per volt out ( all numbers are rounded )

At 50 cents per Bit you cannot make any software corrections, that is one of
the advantages of high resolution converts.


-----Original Message-----
From: synth-diy-bounces at dropmix.xs4all.nl
[mailto:synth-diy-bounces at dropmix.xs4all.nl]On Behalf Of Dan Snazelle
Sent: Wednesday, May 06, 2009 5:38 AM
To: pfperry at melbpc.org.au; sdiy
Subject: RE: [sdiy] modern AD/DA chips?



i am confused as the Converter in the PAIA midi2cv doesnt seem like a super
high precision type.

i have never had any problems with that unit

anyway thanks everybody for the advice.

i will most likely look into some of the 8bit models!!

thanks



--------------------------------------------
check out various dan music at:

http://www.myspace.com/lossnyc

(updated monthly)

http://www.soundclick.com/lossnyc.htm



http://www.indie911.com/dan-snazelle

(or for techno) http://www.myspace.com/snazelle

ALSO check out Dan synth/Fx projects:

AUDIO ARK:

www.youtube.com/watch?v=TJRpvaOcUic

www.youtube.com/watch?v=BqIa_lXQNTA&feature=channel_page

www.youtube.com/watch?v=V4nJPjGgOcU&feature=channel_page

and soundtrack/design work:

NEW: check out Dan's sound design from the 1998 award winning film SAFARI by
catherine chalmers

http://www.catherinechalmers.com/videos.cfm






----------------------------------------
> From: pfperry at melbpc.org.au
> To: synth-diy at dropmix.xs4all.nl
> Subject: Re: [sdiy] modern AD/DA chips?
> Date: Wed, 6 May 2009 18:29:13 +1000
>
> ---- Original Message ----- From: "Andre Majorel"
>>The CS4398 is supposed to have a THD + noise of -107 dB. Doesn't
>>that imply that the error is at 10**(-107/20) = 4.47E-06 ? If so,
>>that part is 18 times as precise as you need it to be.
>>Am I missing something ?
>
> One thing to watch: for DA &AD converters intended for audio use - and
most
> with
> 16 or more bits are - there is usually no tight guarantee regarding zero
> offset, or
> gain constancy over temperature.
> Neither of these things matter at all for audio. But for a say a MIDI-CV
> converter,
> they are essential.
> paul perry Melbourne Australia
>
> _______________________________________________
> Synth-diy mailing list
> Synth-diy at dropmix.xs4all.nl
> http://dropmix.xs4all.nl/mailman/listinfo/synth-diy

_______________________________________________
Synth-diy mailing list
Synth-diy at dropmix.xs4all.nl
http://dropmix.xs4all.nl/mailman/listinfo/synth-diy




More information about the Synth-diy mailing list