I am trying to use an AD5680 nanoDAC in a control loop whose output needs to be updated on the order of every 10 microseconds. Unfortunately, I'm finding two problems:
1) The output change does not occur for about 250 microseconds after the serial data transmission ends.
2) If serial data are sent more frequently than about every 200 microseconds, the analog output switches between multiple states in a cyclical pattern (the pattern depends upon the input data word, and the temperature).
I could not find a specification in the data sheet to describe either of these phenomena. The listed settling time of 4 microseconds and serial clock rate of 30 MHz suggest much higher update rates should be possible.
Any suggestions for workarounds? Best alternate part?
The captures seems to correlate the behavior described.
I personally believe the devices was not designed to operate in this way (multiples updates), and probably there is an internal race condition…