Spikes in measured Power when powering on emonTX

When turning on my emonTX I get a spike in the measure power - way more that can be supplied by the circuit breaker.  Any suggestions why and how it can be avoided? 

I am using the emonTxV3_4_3Phase.ino.

By the way I have merged the emonTxV4_3Phase with the temperature measuring code from emonTxV3_4_DiscreteSampling. If anyone wants it they can get it. I am not very good at coding so there is no guaranty. 

Robert Wall's picture

Re: Spikes in measured Power when powering on emonTX

That should not happen - it looks as if the filters are not being initialised correctly, or your centre bias supply is way off. Do you have a line like this (+ 4 similar ones for each current input?)

static double offsetV = ADC_COUNTS>>1; //Low-pass filter output - start at half-rail, or zero if CT was not detected.

That sets the filter to what should be the normal quiescent working point of the input circuit (i.e. 0 if there is no CT, half-way if there is. If for some reason that is not working, the easy way is to not send the output for the first minute or so while the filters settle.

The temperature measuring should be in the next version, when I have time to fully test it! At present, I'm looking at the problem of retaining the sensor identities if one is replaced, and it's that which has held up progress.

larsthiesson's picture

Re: Spikes in measured Power when powering on emonTX

Yes I have lines like in the sketch like you describe. I think I will implement a waiting if the output is more than 13 amp.

Thanks again.

Robert Wall's picture

Re: Spikes in measured Power when powering on emonTX

You could do some deep debugging to find out why. "offset..." starts out at 512 and should end up at 512 if the bias network is absolutely spot on. It won't be of course. On my voltage channel, its been up to 519.7 - I guess that's due to a spike or asymmetry in the waveform - and seems to settle at around 513.3, while the current input (only 1 CT plugged in and not on a cable, so zero input) settles to 510.7, taking about 15 readings to get there.

It will make a difference to the power value of course, but it's the only way to compensate for varying bias conditions (due to temperature drift etc).

But your real problem is, if you're only reading up to 13 A with a 100 A input, you should really do something about that so that you use more of the available range, such as have a multi-turn primary winding for your CT, and then adjust the current calibration accordingly.

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.