Regarding unstable ADCs with PWM signals ...
I'm sure you've already done this, but "just in case":
Take your first ADC reading and discard it.
Take an odd number of sequential ADC readings, adding them together.
When done, divide by the ODD number you measured
Read ADC and Discard
Read 9 ADC values, and add them together (x=x+ADCValue)
x/9 = 9-averaged ADC Value.
The purpose of discarding the first ADC value is the ADC will have some
random value held by the input capacitor. When you read, this voltage
gets pulled to zero, ready for the next ADC read.
I always have deleted the first ADC value, and run an odd-number average
with ADCs. This worked very well in an industrial, very high current
environment (measuring 1000 - 1600 Amps in under 100 mS).
If you've already implemented this, you will want to include a low pass
filter with the 3dB point 1/t of your measurement period. There is no
reason to have a higher frequency than this, and a lower frequency
filtering will cause your ADC values to lag actual values.
Also, be careful of high crest values (sharp peaks and dips) ... NEVER
assume your ADC input is a pretty sine wave or a SLOWLY changing DC
voltage/current.
This work was performed on an Isopod Microcontroller board (New Micros,
no longer in business) using a Motorola DSP56F800 series 16-bit Digital
Signal Process (DSP) processor running at around 80 MHz.
Yea, that long ago!
Hope this helps.
Regards,
Rick
_______________________________________________
Triangle, NC Embedded Computing mailing list
To post message: [email protected]
List info: http://mail.triembed.org/mailman/listinfo/triembed_triembed.org
TriEmbed web site: http://TriEmbed.org
To unsubscribe, click link and send a blank message:
mailto:[email protected]?subject=unsubscribe
