View previous topic :: View next topic |
Author |
Message |
hayee
Joined: 05 Sep 2007 Posts: 252
|
What resolution is required |
Posted: Tue Jan 31, 2012 12:20 am |
|
|
Hi,
I am planning to read 3-phase voltage and current in a microcontroller and calculate its power. Is 10-bit resolution is enough to read the parameters and calculate power ? Can I calculate percentage error (when using 10-bit adc as compared to 12-bit adc) ?
Thanks |
|
|
Ttelmah
Joined: 11 Mar 2010 Posts: 19516
|
|
Posted: Tue Jan 31, 2012 3:02 am |
|
|
Seriously, 10bit is _more_ than enough. Few voltmeters even get close to this when measuring AC....
However, you need more data than the voltage and current. You need also the phase relationship between these to actually calculate 'power'. Power is only the product of voltage and current, when feeding a load which has a power factor of '1'. Into any other load there will be a phase difference between the signals, and power will change.
Best Wishes |
|
|
temtronic
Joined: 01 Jul 2010 Posts: 9228 Location: Greensville,Ontario
|
|
Posted: Tue Jan 31, 2012 6:30 am |
|
|
Also signal conditioning, how 'clean' the readings are, is more important that ADC bits.
It's far better to have 8 'clean' bits than 10 or 12 'dirty' bits of data !
Capturing 12 bits of data when 3 or 4 are spikes,glitches,etc. defeats the purpose of a high bit ADC.
You don't say what frequency this 3 phase power is. 50Hz,60Hz, 400Hz,20KHz but that will be a huge factor in designing the correct signal conditioning section of your project.There's no math you can do inside the PIC to compensate for 'bad' raw ADC data. |
|
|
SherpaDoug
Joined: 07 Sep 2003 Posts: 1640 Location: Cape Cod Mass USA
|
|
Posted: Tue Jan 31, 2012 6:58 am |
|
|
Consider that 10 bits gives you 1 part in 1024. That is slightly better than 0.1% Unless you take great care with the rest of your system you will likely have other errors much bigger than 0.1%. _________________ The search for better is endless. Instead simply find very good and get the job done. |
|
|
Battery David
Joined: 01 Feb 2010 Posts: 25
|
|
Posted: Tue Jan 31, 2012 7:56 am |
|
|
Following up on SherpaDoug's comment, if you are using 1% resistors to divide down a higher voltage, 10-bits is now more accurate than your voltage.
After I figured that out I started using 0.1% resistors for scaling. |
|
|
Mike Walne
Joined: 19 Feb 2004 Posts: 1785 Location: Boston Spa UK
|
Current sensing |
Posted: Sat Mar 03, 2012 4:39 pm |
|
|
Is your three phase supply star or delta connected?
How are you proposing to measure current?
Quote: |
Can I calculate percentage error (when using 10-bit adc as compared to 12-bit adc) ?
|
Percentage error in which value?
Mike |
|
|
|