440v ac measurement atmega8

[microkid]

Hi

any body done this before?
I need basic circuit (steppin down) and interface to mc and calculations.
I cant use a step down transformer.

MicroKid

[vovchik02]

You can try optoisolated amplifier HCPL78XX (HCPL7800 e.g.) for decouple High voltag and mesure cicuit.

Bat transformer is more simple and inexpensive solution.

[jzaghal]

HI,

You can use two 220V:12V transformers in series to obtain 12V + 12V (i.e. 24Va.c out)

Rectify with bridge and filter with say 100uF/100V cap.

Use a resistive divider say to obtain 4.4Vdc .

Read with A/D converter 1V is equiv to 100Va.c

Hope this will help.

Bye.

[microkid]

thanks to jzaghal  and  vovchik02

The space limits me not to use transformers and miniature transformers are much costlier than ordinary.

I will search availability of hcpl7800.

Is there any circuit using direct voltage divider?

MicroKid

[beque]

Look at this doc below:

http://www.lem.com/docs/products/LV%2025-P%20SP2%20E.pdf

Voltage transducer isolated using Closed Loop Hall Effect

[Walkura]

This is the setup i recently used .
http://saturn.uni-mb.si/~bojan/pdf/app50.pdf
Worked very well for me .
(i needed to measure by microcontroller the input voltages while keeping a total optical isolation)

Good luck

[DTiziano]

A solution can be built by using two microcontroller, the first make the measure directly from the line and send data via optocoupler to the second.
You need an isolated power DC/DC converter to supply the line micro and use the supply voltage as the reference for the ADC.
A simpler voltage divider (attention that 440 vac means 620 Vpp) can be used.

[Walkura]

If you don't need optical isolation there is no objection to just use a diodebridge + resistor devider .
(meaning that the microcontroller is under all circomstances referenced to the same ground as the 440 Volt)
In the past i used setups like that and it worked fine .
With voltages like that ALWAYS keep at least 2 resistors in serie (no more then 200 Volt over 1 resistor)
The il300 solution works very nice but if you don't need optical isolation ,resistors would do .
Also with the non inverting setup of the il300 isolation amplifier you need + & - powersupply or you will have a small offset .
But i can assure you that it works nice lineair and accurate for optocoupled measuring of high voltages .
Other options are ofcourse the lm311 voltage to frequency converters and vice versa but thats more parts and trouble .
Also there are isolation amplifiers from texas instruments the iso122 but they are expensive and from what i heard noisy .

[dideco]

Another Idea (not tested, but should work according to my previous experiences)

You can use a miniature toroid which are commonly used in filters in computer power supplies to reduce EMI. The material used in this transformer toroid is specifically designed for low frequency applications. I have tested those filter transformer as a current transformer in 220 Volt, just winding only two turns of a cable which was on the 220 Volt side. I was able to measure a few miliamperes linearly with these toroid cores (ex. 15 watts lamp on 220V AC). Other toroid transformer cores did not worked as good as that toroid filter transformer at the input of a power supply such as computer power supply, probably because of their magnetic material designed for high frequency applications.

So my idea-circuit is as this: These transformers has two similar windings which can be used as primary and secondary. On the 440 Volt side, use one side of winding connected with one or two capacitors in series . The capacitor values should be something like 47 nf to 220 nf but high voltage ones. There should also be a resistor in series to protect capacitor(s) from high currents as used in transformerless power supplies designed with capacitors in series (47 ohs to 220 ohms according to the max current limit of the capacitor). This is the 440 volt side.

For the 5 volt side, use the second winding. Voltage should be divided with resistors to get 0-5 Volts range. Use an osciloscope. Be careful for the 5 Volt side because there should be low current but high voltage as two of the winding is same on each side of the core. This high voltage can damage input of an osciloscope. Always use resistors to divide voltage before measurement. A diode and a capacitor should filter to get DC.

Advantages: Small in size, isolated, linear, cheap, all with passive components.

Good luck.

[jzaghal]

Hi,

Walkura, did you mean lm331 ( lm311 is a comparator!).

Bye.

[microkid]

Thanks for every one replied.

My application already stated is verymuch limited by space and money. I cannot add any more components.

While searching, i found an application note from atmel for an energy measurement using mega 48 uses no optical isolation but only resistors.

But I now think using dideco idea, using a toroid to isolate.

I thank every one again for sharing their time.

MicroKid

[picavr]

all you need is there
http://www.sonsivri.com/forum/index.php?topic=12303.0