[sdiy] ADS8326 problem & sdiy-list not working properly

Magnus Danielson magnus at rubidium.dyndns.org
Wed May 25 00:37:48 CEST 2011


On 05/24/2011 05:52 PM, karl ekdahl wrote:
> Seemingly i can't send any e-mails to SDIY lately and i had to
> subscribe using this address, please moderators fix it! I like this
> list!
>
> Anyway
>
> Per previous discussion i picked up a ADS8326 and built a circuit
> around it, it seems to work - somewhat - but the results i get are
> *seriously* strange and unusable. What i get is a "stepped" output,
> only a few numbers show up - it's like it's quantized. The result i
> get seems to be exactly what this person got;
> http://e2e.ti.com/support/data_converters/precision_data_converters/f/73/p/2871/9884.aspx
> - but i don't know how they fixed it. While checking the data stream
> it seems like all '1's are always in groups of five somehow...
>
> Just like in the datasheet i'm using 22 clocks, i've tried both
> putting in and removing various delays but it doesn't seem to change
> anything. Is there any other possibility that this is *not* the result
> of a broken IC, and what would it be in that case? I'm using it at
> 3.3V if that makes any difference...

Try letting the CS input stay high between the conversions rather than 
just "flipping" it just prior to clocking. The CS pin doubles for 
power-down. Also see if not the clock should stay high between the 
conversions. There is some timing relationships there to observe.

Could be that the digital section is being upset by an incomplete 
shutdown or something.

Try variants around that.

Cheers,
Magnus



More information about the Synth-diy mailing list