Quite often we have AC signals with negative excursions we need to measure/digitize
with a processor. But processor does not tolerate negative inputs*. A simple method that
stays DC coupled source to processor in is to add a R divider referenced to UP supply
rail. This eliminates the need for a OpAmp or coupling cap with a bias network in design.
* There are processors out there that can handle an A/D conversion ~ 100 mV outside UP rails
and still be accurate in conversion. That allows high and low side current measurement for example.
Attached is excel spreadsheet to calc R values for a given supply and shift.
Note accuracy dependent on positive supply rail tolerance, as well as R ratio. A workaround
would be to use a Vref to power processor to eliminate Vsupply impact on accuracy.
Another downside is you throw away some G, but many applications that would not be
an issue for basic simple measurements.
Of course this lowers input Z as seen by source of Vin, which can be alleviated
somewhat by raising the values in the divider. Those in turn limited by leakage
currents. Trade-offs, always trade-offs.
I saw this originally in an EDN article.
Regards, Dana.
with a processor. But processor does not tolerate negative inputs*. A simple method that
stays DC coupled source to processor in is to add a R divider referenced to UP supply
rail. This eliminates the need for a OpAmp or coupling cap with a bias network in design.
* There are processors out there that can handle an A/D conversion ~ 100 mV outside UP rails
and still be accurate in conversion. That allows high and low side current measurement for example.
Attached is excel spreadsheet to calc R values for a given supply and shift.
Note accuracy dependent on positive supply rail tolerance, as well as R ratio. A workaround
would be to use a Vref to power processor to eliminate Vsupply impact on accuracy.
Another downside is you throw away some G, but many applications that would not be
an issue for basic simple measurements.
Of course this lowers input Z as seen by source of Vin, which can be alleviated
somewhat by raising the values in the divider. Those in turn limited by leakage
currents. Trade-offs, always trade-offs.
I saw this originally in an EDN article.
Regards, Dana.
Attachments
-
220.9 KB Views: 1
Last edited: