Hey,
We've been happily using the AD7606 for many years, but now a new batch of PCBs which are supposed to be identical to the old ones is giving us troubles with the AD7606.
Problems
We have two problems/symptoms that may be related (or not):
1) After CONVST the BUSY pin goes high as expected ... but does NOT return to low anymore.
Mysteriously, this depends on the exact specifics of the uC firmware code used to read out the BUSY pin.
If I only send a CONVST and then do nothing, BUSY goes high indefinitely.
If I follow the CONVST with a few reads on the uC pin, then depending on the specifics of the read BUSY may or may not go back to LOW after the conversion time.
2) The conversion time is only 3 us, i.e. *faster* than the min. in the datasheet.
Here's a measurement on a scope:
As visible, the conversion takes 3 us, which is below the 3.45 us min. conversion time from the datasheet for the AD7606.
Because the 3 us would match the typical conversion time of the AD7606-6 I have checked whether the part got swapped.
However, I can confirm that I indeed have the AD7606 since all 8 channels work and give independent readings.
Questions
Have there been any changes/revisions to the AD7606 lately?
Is there anything to consider regarding the BUSY pin (and its drive strength, pull-up/down @ uC, etc.)?
Are there any issues with counterfeit chips?
Upon replacing the AD7606 with the weird behavior with a brand new one from Digikey all the troubles above vanish!
However, our PCB manufacturer also got the AD7606 from reputable sources, so we're very surprised overall ...
