Measurement of a Dual 220V

[Silent_Thunder]

Hi everyone;
I have a problem regarding measurement of a Dual 220V .
I hope someone could have the time a help me out, thanks in advance

the circuit attached has 2 inputs , variable from 130 ~ 240V, both have neutral in common.
the output from each branch (AN1 and AN2) goes to 2 analouge input of a microcontroller.

when I apply input only to (L_IN and neutral) I get reading from both AN1 and AN2,
almost the same value for both, which is a problem.

and when I apply input for both (L_IN and OUTPUT) with neutral in common,
I get unexpected reading cuz some times <OUTPUT> is much higher than <L_IN>.

I want to separate the out from the circuit (AN1 and AN2).

any idea how I would do that ( putting in mind that the both input voltages share the neutral)

thank you for your time, and for the help

Regards

[pickit2]

Is this in real mode or in simulator ie: Proteus

[Silent_Thunder]

this is in real mode, in Proteus everything works fine, Proteus does not sense the cross talk between the 2 inputs.

[pickit2]

I would suspect your wiring, if your cable is running close to each other, it might be a voltage pick up.
I remember a project that worked well in the workshop, but on site where it was placed, there was a standing voltage on the signal leads. a few ferret beads and twisted pair wiring sorted the problem, but it was puzzle as to why.

[Silent_Thunder]

I wish it was that simple
even if I applied single AC input, I get reading on both AN1 and AN2 channels

[pickit2]

are you getting a reading on both AN1 and AN2 if you take the  microcontroller out of the circuit.
if you are, then check your wiring, I still thing its mains pickup.

[Top]

When L_IN goes negative current flows through D10.

Use half-wave rectifiers and connect GND to Neutral.

[kreutz]

Please, make sure you let the capacitors discharge completely between tests. Even if you don't use half wave rectifiers, as suggested above, this circuit should work exactly as when simulated.
Check for conductive solder paste leftovers between AN1 and AN2 input pins of the micro-controller chip. Wiring might also be a problem as suggested by pickit2. Taking the micro controller out of the circuit is also a good troubleshooting hint. Recheck the wiring again... ground loops will also drive you crazy, check the micro controller power supply grounding and use a battery to replace it just in case.

[Silent_Thunder]

Thank you all for your support;
I'll modify it to half wave and see what that gonna lead me, and I will tell u what happens.

[Silent_Thunder]

I'd like to thank you all once again
I modified the input to half-wave rectifier, and now it works perfectly, with a small modification in calculation formula.
 Attached the new modification in case anyone came up with the same issue.


kind regards to all of you 

[pickit2]

Safety note: putting an earth on the mains negative will not pass, electrical standards and approved codes of practice.

[enzine]

In a similar case in the past I used custom measurement transformers (TV).

If you do not have stringent precision specifications, you could simply use two small power supply transformers, for example 220V / 6V ac.
Safety problems and cross-talk are solved.

Ciao

[Signal]

First step was to reject bridge rectifiers to allow common ground as Top pointed.

Second step is to reject C7 and C10 peak holders leaving C8 and C11 (at least 100uF each) as calculators of average amplitude that is more adequate measure than peak that is usually disturbed in real life.

Next step is to reject C8 and C11 to allow uC to calculate half-wave-true RMS. (positive side effect: much faster reaction on input changes)

Then to measure both half-waves you can use ...
... transformers as enzine suggested
... or HPF capacitor to shift "zero level to Vdd/2 before ADC inputs. (As positive side effect that variant is free from diodes as rectifier so measurement could be more precise.)

Code:

                   5VDC
                    |
                   6.8k
127VAC>--1uF--280k--|-----,---1k-->ADC
                   6.8k  10nF
                   _|_   _|_

... or opamp shifter without serial capacitor. (that variant is not justified because of next)
... or true-RMS specialized IC


Of course all above is only a theoretical exercise because I do not know the real target of that measurement.

[jumulab]

Connect the neutral line to common GND,  in this way you shouldhave the  reference unified
at the AC/DC rectifier and  get the good result in the A/D converter.



Posted on: February 08, 2018, 02:41:32 14:41 - Automerged
see this example of connection of a real 3 phase meter inputs. In the example we are taking measures over the AC
signal,  in you case yoy are rectifiing the ac and taking samples over the DC, but the connection schema is equivalent.
Only the reference pint  NEUTRAL is the key in the schema.


Posted on: February 08, 2018, 02:50:04 14:50 - Automerged
the second part of schematic

[zielpunkt]

Sorry, I assume you've solved your problem and are not looking for safety advice, but if you use your circuit in a professional way, you should isolate your inputs. If transformers are unsuitable for you, you can try using insulator chips for these measurements. There are several types of them at relatively low costs. Good luck.