Removing ripple from DC current measurment??

[PaulM]

I'm working on an off-grid solar system for work that is going to power some measurement equipment. It consists of 6 240W solar panels, a 60 amp Morningstar MPPT charge controller, an Outback 2.5 kW inverter and a 24 volt, 1820 Ah battery. To keep tabs on the system, we have installed a datalogger with remote communication that sends data via email three times a day. We only have four analog channels and are measuring battery voltage, battery temp, solar charge current, and inverter current. The logger works fine except the inverter current reading fluctuates between 0 and 20 amps with an actual steady load of 10 amps. We are using a 100A hall effect current sensor (offers a higher output voltage than a shunt) and the voltage output reads perfectly steady on a DMM. Today I put a scope on, and found a ripple on the signal, about 300 mV peak to peak (signal is around 2.6 VDC or so), approx 120 Hz. The solar current signal is perfectly smooth so I know it's not the power supply for the hall sensors. I am assuming this ripple is caused by the inverter and the logger's sampling frequency must almost match the 120 Hz or some multiple.


Does anyone have any suggestions on how to remove this ripple? I tried a small cap today and it helped a bit, but I'm afraid of damaging the hall sensor if I go too big on the cap.

[liveforphysics]

You can go huge on the cap with no risk to the sensor if you stick a 100ohm resistor between the sensors output and the cap, and then read your voltage off the cap. You just sacrifice the ability to see any spikes or transients if you go that route.

[PaulM]

Thanks Luke, I'll give that a try today.

We're not too concerned with measuring spikes or transients, it's mainly to get an idea of the power consumption and see how it compares to our solar input.

[bigmoose]

What Luke described is the essence of a low pass filter. That is what you are after. Play with it in spice to tune your parameters. Or do it in analog (design) and put the knee of the low pass at say 10 to 30 Hz.

[PaulM]

Thanks again Luke, I used a 100 ohm resistor and grabbed a decent size cap from my assortment (2200 uF, 35V) and it seems to have worked great. The current graph is now very smooth and the data is actually usable now.

[liveforphysics]

Thanks again Luke, I used a 100 ohm resistor and grabbed a decent size cap from my assortment (2200 uF, 35V) and it seems to have worked great. The current graph is now very smooth and the data is actually usable now.

Right on. Glad it did what you were looking for.