ODP: quantizer question

[Roman Sowa]

8 bit DACs DO have poor tolerance, as every other converter.
But ussually it is limited to the few LSBs. So, the wider DAC input,
the better linearity within MSB range.

1/2 LSB quantization noise mentioned by Colin has nothing to do with
DAC linearity. They're made just as precise as needed to maintain
differential error less than 1LSB (the code higher by 1LSB gives
output voltage no less than with code smaller by 1LSB). And that's not
even a rule for higher resolution DACs (14 bits and more).
That's what makes'em 8-bitters cheeper.

If you had no problem with 8 bit DACs, then maybe you were lucky
and got a better piece, or used it as a 6 or 5 bit converter.

I was lucky too - i bought 12bit AD7548 in radio shack for about
$1.50 a piece.

Now, with new IC prices falling down (at least few of them) it's easy
to find affordable, yet well performing DAC

Roman

> -----Oryginalna wiadomość-----
> Od:	Fraser, Colin J [SMTP:Colin.Fraser at scottishpower.plc.uk]
> Wysłano:	31 sierpnia 1998 11:19
> Do:	Synth-DIY
> Temat:	RE: quantizer question
> 
> > >In looking at Chris list's ARP style quantizer, I can't help but
> > >ask if one could get away with using a cheeper 8 bit D to A
> > >converter. One that is like $3 insted of like $23? Since only 7
> > >bits are used anyway.
> 
> I have used various 8 bit DACs (ZN428, ZN558) to output CVs in various
> units and never had any problem with 'tolerance'.
> You just have to apply a bit of gain to the output of the DAC so that
> a
> 1 bit step = 1 semitone (83.3mV for 1v/oct).
> 
> A 12 bit DAC allows 3 or 4 octaves of Hz/V type output if you use all
> the bits (and a lookup table).
> 
> The average error in a quantised *continuous* signal will be at least
> 1/2 LSB - maybe this is where people get the idea that an 8 bit DAC
> has
> poor tolerance.
> 
> 
> Colin f