On 02/03/2014 04:24 PM, RogerN wrote:
"Tim Wescott" wrote in message
...

On Sat, 01 Feb 2014 17:19:56 -0600, RogerN wrote:
snip
Anyone here mess with signal conditioning for current transformers?
Seems that I need a "precision rectifier" circuit and a capacitor to
charge up to the peak, then scale to RMS amps.

Do your A to D conversion at the 3rd or higher harmonic of line frequency
(higher really is better), square, and average. Then take the square
root, and you have RMS.

--
Tim Wescott
Control system and signal processing consulting
www.wescottdesign.com

I did some checking, the Arduino's AnalogRead() function is supposed to take
around 100uS. I'm thinking I can take a A/D reading ~641uS apart and get 26
readings in 16.666mS. Then do the RMS calculations from the readings. This
should be great because I should be able to take the voltage from the burden
resistor to provide 2V P-P at full scale. Note I'm putting a 100A CT on a
15 A circuit so I should have around 1/10 amp resolution per count (10 bit).
If this doesn't read well, I'll get a lower amperage CT.

So thanks for the idea! That saves an Op Amp rectifier and filter! I never
thought of doing the calculations right from the AC signal.

RogerN


I'd be interested in the actual USEFUL resolution you can get in this
noisy environment. I tried to do 10 bits on a PIC in a quiet
environment and was very disappointed.

Might be interesting to use an X-10 appliance module to control the
heater. Don't know about Arduino. PIC Basic has library support for
X-10 commands. Or just take apart a controller and wire to a button.