Page 1 of 1

Posted: 31 Jan 2020, 02:08
I am trying to measure the high current going to a bidirectional an electrolysis cell. The power source is from a transformer with about 9V AC at 50Hz. This goes through a full bridge rectifier. The current should be between 10 and 20 Amps.

At the moment I am using a FET current sensor that outputs a voltage between 0V and 5V, with 2.5V corresponding to zero current. I have put this through a voltage divider and connected it to the analog input of an ESP8266.

The problem is the current varies like a rectified sinewave and the current reading depends on where in the cycle the reading was taken. This means anything between 0A and 20Amps. Now I don't need an RMS reading, just an approximation of this. I have tried using the oversampling feature in ESPEasy, but this doesn't resolve the problem.

Another possibility would be to use a capacitor on the analog input to the ESP8266, thus smoothing out the value. I am not sure if this might distort the result because the values are using this offset values, that is zero current is equal to 2.5V. If this is the best option, what size capacitor should I use?

Any assistance would be appreciated.

Posted: 31 Jan 2020, 13:06
Hello,

several possible ways I think.

First and most simple way:
Use the output of your FET sensor, decouple via a diode and use a capacitor
to get a true DC on your ADC Input. It's same circuit as a simple rectifying
in a power supply.

A second possibility would be a magnetic transformer on the AC side of your bridge rectifier
and a circuit as shown here: viewtopic.php?f=2&t=7353#p41850
Scrooll down to post #8. A transformer up to 20A is needed, watch out it gives an output current,
not an output voltage. R_Load must be dimensioned following the datasheet of the specific transformer.

A nice but more elaborate way if your FET sensor gives the reference voltage of 2,5V:
You might use an external ADC like the ADS1115 and configure it to differential input.
Then put a rectifying curcuit like in the first sugesstion on the positive differential input,
the reference from the FET sensor to the negative diff-input.
This has two advantages: At first, no more calculation is needed as 2,5V- 2,5V gives the correct zero value.
Second, most reference voltages move a little due to temperature etc. This circuit will compensate
a reference voltage drift with no further hassles.