Suggestions for a simple SAFE opto-isolated 230VAC voltage measurement circuit

[bsodmike]

Thanks Fchk. I’ve got some questions.

On X1, I assume that the load is connected before pin 6, which has the live line just before the shunt. Then, pin 1 is the return Neutral line (hence GNDA is the Neutral line).

X1’s pins 2 & 5 are a bit confusing. I would expect the live line connected at the C2 X2 class safety capacitor, with GNDA anyways being the Neutral return.

What does PL_OGD mean for the fuse? You’re using a 4A fuse, any particular specs on this, I.e fast blow?

Please correct me where my assumptions are incorrect!!

Thanks M

[fchk]

PL_ODG is the type of fuse holder I used. The current measurement resistor R6 is sized for 4A, so the fuse must be 4A.

X1: You connect your load to 1(L)/2(N)/3(PE) and your power plug to 6(L)/5(N)/4(PE).  PE and N are directly connected, and L goes through the current shunt R6. If you swap 1/2/3 and 6/5/4 then your current and power readings would be negative..

You may as well swap L and N. GNDA is the internal reference ground for the CS5490 for measurements. If yo swap L and N, you would measure current through N.

C1 (the X2 capacitor) is connected between L and N - which side is GNDA doesn't matter since everything is isolated anyway.

fchk

[Alti]

I want to be able to measure mains-AC voltage safely.

What is wrong with measuring AC mains voltage via resistor divider? The overvoltage category applies to installation location and a voltage divider for 1500V (CAT I) is not much different than the one that deals with 2500V (CAT II), just more resistors. Why MELFs? Standard resistors also come with adequate voltage ratings so for 2500V you just need to use 5 x 1206, nothing crazy. If you need CAT IV - use even more resistors. MELFs shine when you need high pulse power, this application does not require any pulse power.

Then just put adequate count of resistors in series and use an opamp with adequate biasing current. If you want 80M divider then use biasing current that is just much lower than 3uA. Of course the capacitance is going to shift 50Hz response - I am just pointing you won't receive output that is perfectly in phase with input with high value resistors stretched along PCB that form RCRCRCRCRC filter.

Mind opamp differential amplifier has two inputs so you need two strings of resistors actually. If you can solve a problem with 10 resistors and jelly bean opamp then don't overdo with optoisolation. And don't use CAT I equipment in CAT IV environment.

[bsodmike]

The requirement changed to full energy monitoring and therefore we chose a proven IC solution.