Non-Isolated Shunt based AC Mains Current Sensing

[andig]

Hi Everybody,

I posted this on the microchip forum too, but did not get any answers.

I need to design a low cost AC Over current switch based on PIC16F676. It reads the line voltage (Half Cycle Resistor Divider) and the Current (0.1 Ohms Shunt Resistor). The max load being 4 amps AC/240V

I am using a capacitive power supply due to cost factor. To keep things simple I am using a single 5.1V supply derived from it.

Now my problem is a single supply AC voltage amplifier. The voltage derived from the shunt needs to be amplified and rectified to be read by the ADC. I tried with a LM358 with the non-inverting input biased to 1/2 of the supply voltage followed by a precision rectifier. It did not work as intended. Non-linear response.

Any pointers on the analogue interfacing part ?? That is amplify+rectify voltage derived from the shunt

_____________________________

Best Regards
Sougata

[maddoc]

I saw an interesting approach here: http://www.edcheung.com/automa/power.htm
It uses a flyback transformer core from old TV's to clamp it on the live electrical wire. Work like the clamp am-meters.

[bobcat1]

Hi

The easy way is to use a current transformer there are many low cost type and you cna also do one of your own (it is just a huge coil with the current caring conductor pass trough)

I recommend you to buy one from Asian sources whether you need this for commercial use

The out put of the transformer need to be rectified & amplified before you pass it to the A2D converter

You can also use special IC for measuring the current voltage and power in AC line (see analog site for more information )

All the best  

Bobi

  

[borberk]

At 5V supply you can get only 2.5Vpp undistorted signal from LM358. Use  higher voltage supply or rail-rail opa.

[andig]

Hi,

Let me explain this is more:

1. I cannot spend more than $2.5 on the BOM where the PIC itself contributes to $0.8 so use of a C.T. rules out.

2. As in case of ordinary DMMs the AC current measurement is shunt based. I am just trying to achieve that.

3. Since I am doing half wave measurements on the AC line directly to get RMS voltage reading too. I did made a split supply but need to keep the overall consumption down to a few ma.


Right Now I am using a TLO72 based amp followed by half wave rectification, results are better but accuracy gets lost in the lower side. May be I will compensate that in software.