AC line voltage measurements

[prasha920]

Hi All,

i want the efficient way to measure the ac line voltage of the amplitude 250Vrms,
tried several methods,
1. line coltage --> full wave bridge rectifire --> filter caps --> resistor divider network...
almost all worked up to the measurements of amplitude. problem is that my project is having keys for overvoltage and undervoltage settings, so there should not be the risk of electric shock.

2. ac voltage after stepdown transformer --> rectifier--> filter caps -- > resistor divider network.
i am using relay to switch another circuit on and off, this relay is driven by the same supply which is powering the uC, so when line voltage is out of the range relay is energised which sinks more current from regulator ic(this changes the adc voltage and ultimately the line voltage reading)

any help would b greatly appreciated as i am working on these issues from 3 days.

Prashant

[Benj]

Hello,

With the second approach can you simply use a separate DC PSU to drive the electronics and supply power to the relay. This should then leave the AC voltage completely untouched so you get an accurate reading no matter what.

[prasha920]

hi Ben,

so for separate DC PSU, i should use separate transformer?
it may make the unit so bulky..
actualy i was using capacitive power supply for the electronics part as it working on very less current, but this capacitive PS is unable to drive Relay as relay's current requirement is much higher.

to solve this issue, i made two power supplies but used common step down transformer
i.e. ac mains --> 24V transformer --> 2 separate bridge rectifiers --> separate filtering caps --> Regulator ICs

is it ok?
i made sure that grounds of these two supplies are not common. in this case all is working ok if Line and Neutral are at right place, if they are exchanged risk of electric shock generates.

m i right at using common transformer for two separate PSU's? or m i going wrong anywhere else?

[medelec35]

Hi prasha920
Food for thought:
If it was me measuring AC mains, I would probably use opto-isolator with a fixed CTR.
This would provide total isolation. I would use something like 2 x 120K resistors in series from say live to one of the inputs, then the other i/p would go to neutral.
you have to stick with AC type opto-isolator. you can use a DC type and connect a diode e.g 1n4148 anode to cathode of opto, and cathode of diode to anode of opto thus making it suitable for ac.
Using this method you can directly interface the output of opto-isolator directly into a microcontroller with either a pull down resistor from opto emitter and ADC i/p of microcontroller to 0V and opto collector connected to +5V.
Or
Emitter of opto connected to 0V, collector connected to microcontroller i/p and via pull-up resistor to +5V. This way the ADC in rises as AC voltage falls so you will need to use something like ADC = 1023- ADC
I won't go into too much detail at this stage in case you think my idea is not that good or practical


Martin

[prasha920]

Hi Martin,

If it was me measuring AC mains, I would probably use opto-isolator with a fixed CTR.

can you mention the optoisolator name or no.?
i practically implemented this technique, i used mct2e optosiolator and same config. as u stated

Using this method you can directly interface the output of opto-isolator directly into a microcontroller with either a pull down resistor from opto emitter and ADC i/p of microcontroller to 0V and opto collector connected to +5V.

with this setupi got the expected output but within a range, for ex. from 15Vac to 190Vac i get proprer readings and after 190Vac i even dont get a single digit change..
i guess MCT2E is not linear type opto so the obtained result.

which opto u would have used? also i didnt get what is CTR?



Prashant

[medelec35]

I would use any that has high linearity
How about something like
http://uk.rs-online.com/web/p/optocouplers/0450806/

CTR is the Current Transfer Ratio.

If you want to make a production item then fixed ctr will be ideal.

after 190Vac i even dont get a single digit change

Sounds like either input has reached maximum current or output is saturating so no more increase in voltage on the i/p will make any difference on the voltage of the o/p
you can try to either increased the the resistance value connected to the input or use a fixed resistor across the i/p of the opto so its not driven to saturation,
or change the resistor connected to the output so the transistor o/p is not fully on.
It is probably the latter that's causing the issue.
Have you got a circuit diagram of your hardware?
I maybe able to help you further, but I'm not an expert with this so a solution is not guaranteed.

[prasha920]

hi Martin,

i checked the optocoupler, it serves the purpose but it will be a costly solution.

glad to have help from you.

one more thing Martin, i was thinking to use V to F converter as replacement for costly Linear opto. is it possible?
soon i upload the circuit diagram,

[medelec35]

i was thinking to use V to F converter as replacement for costly Linear opto. is it possible?

Nota voltage to frequency converter before, but I'm sure we can modify your existing circuit to work as soon as I know all the parts your using.
As soon as you have posted your circuit, i will see if I can assist you further.

Martin

[prasha920]

Hi Martin,

i am extremely sorry for such long delay to update you..

i have attached the image of power supply used for microcontroller and voltage divider circuit to read from mains.

if live and neutral connections as exchanged then there is a chance of shock hazard in this circuit as power supply is of capacitive type.
how can we prevent this shock hazard?

Prashant