Current sensor and di/dt sensor for ADE7763

https://www.analog.com/static/imported-files/data_sheets/ADE7763.pdf
The above link is the datasheet for the ADE7763.
In pg 12, they have given two circuits involving the use of Current sensors - Current Transformer (1800:1), di/dt sensor

I found the current transformer to be very costly. What is the di/dt sensor ? Is there any other method for measuring current (20A or 10A)

Re: Current sensor ADE7763

hello

the di/dt is the curent transformer itself !
do you need to measure so big current 1800 Amps ?
in this case the cost could be high!

Define your AC current range to define the current transformer.
Supose you have a 50:1 transformer with a load of 1 ohm
you will get 1 Amps into 1 ohms => 1 V efficace or more ,
when you have ana AC current of 50 Amps passing trough the primary of transformer.

Never use a cureent transformer without a low load at the secondary.
In the industry a direct ferromagnetic analog ampmeter is connected directly to the secondary.

Another method is to use Hall effect transducer.. see LEM compagny or others..

Re: Current sensor ADE7763

Ah,
CT as current transformer (manuf. : VACUUMSCHMELZE. etc)
https://www.vacuumschmelze.com/fileadmin/documents/broschueren/kbbrosch/Curr_Transf_neu_engl.pdf
is well known device
- primar: wire trough the hole
- secundar: has to be ended with resistor typ. 10-30 Ohms (each type of CT has own load value resistor see pdf)
- sec. signal: voltage on load resistor is exactly copy of the current !! sinus in, sinus out !!
why the CT is expensive ? it has to be linear from 0-max currents, believe me this is very hard to make.
load resistor signal goes to the differential input on chip, calculate that you need about 20% of margin too.
CT is galvanic isolated! CT signal from res. load need 1 order anti alias filter before goes to the chip !

di/dt : means the sensor is Rogovsky coil !
- well,here are many types of this coils, benefit is: it is very cheap, but is sensit. on external magn. fileds, needs shield, use Anysis sim. to define it!!
- usually 2 coils are in serial, midd. point is on gnd, first coil is measuring coil, second coil is for eliminate non signal ext. mag. fields.
- both end of coils go over second order anti alias filter direct to the differential input on the chip, switch on integrator in the signal processor !
- this type of sensor need a compensation (or has a problems) with very small meas. currents, sensitivity ... 10-40uV/A use intern. multiplay !

CT is expensive but good
Rogovsky is cheap but has problems and need add, help
chose what you need ..

regards Keen

Re: Current sensor ADE7763

the ACS712 is simple to use with an ADC
**broken link removed**

Re: Current sensor ADE7763

horace,
here we are talking about energy metering class 0.1/0.2/0.5
ADE7763 is complete DSP metering engine & support classes above, this is profesional dedicated chip.
ACS712 is hall current senzor only, 2kV isolated, for metering under IEC standard we need 6kV minimum, don't mix apples and plums.

Re: Current sensor ADE7763

This meter is for home application, maybe a max of 20 A measurement will suffice. pls recommend me a CT or give direction to design one...:?:

Re: Current sensor ADE7763

Buy something in Hoffer or in the Lidle store for 10 Euro, they are in class 2% (China coffee product), it is good for you for control partial electr. consumer loads. Meters build in home are under control of electricity distributors , you can not use home made device for payment el. meter !
CT which one? VAC definitive! but which part number? CT+ADE7763+PCB+shield+influence of other components=measuring system,
remember: measuring system has to be linear from 0-max current, 0-max voltage, but you are measuring energy not current, volts! means you have to be in frame of acceptable error of energy measuring! you need energy reference source device (ref. volt/amp supplay + independent energy calibrated system). you will measure something with your measuring system than you will compare this with reference, difference has to be: let say 0.1% if your system is in class 0.2, of course after tuning registers in ADE7763 .. of course 0.1% must be on full current scale 0-max current and here is a big problem, before you will achieve energy accuracy on full current scale you will build some tens PCB's and change some CT's. that is money game.
If you will achieve acceptable accuracy under all meters standard you can start to produce meters. Start now , you will learn a lot.

Re: Current sensor ADE7763

Most cheap single phase energy meters with 16 or 20A current range are using shunts, just a piece of Nichrome wire. An isolated current measurement doesn't bring an advange as long as the voltage measurement doesn't implement isolation.

Re: Current sensor ADE7763

Sorry, I can't understand your second point. :-( pls elaborate the second point...

Re: Current sensor ADE7763

ADE7763 is usually directly connected to power supply, see application notes. You don't need isolated current measurement.

Re: Current sensor ADE7763

hello,

FvM said:

ADE7763 is usually directly connected to power supply, see application notes. You don't need isolated current measurement.

Click to expand...

i read also the datasheet and i will never use a direct shunt on phase , to measure current .
because there is no galvanic insulation to avoid lethal accident .

Even all is in floating situation with insulate power supply , what appends if you connect an external device on SPI Bus ?.. some flash !
For human safety it is better to use Current transformer.

or you can assume yourself using a Shunt and support all risk.
In this case, is the device compliant with the law ?

Re: Current sensor ADE7763

i read also the datasheet

Click to expand...

Yes, but apparently didn't understand the point about galvanical connected voltage measurement...

In a professional design, the isolation will be provided on the digital side (SPI bus). Of course this is a problem during development, particularly for a hobbyist without suitable safety transformer. Or you have to use an instrument transformer for the voltage measurement in addition to an isolated current tranducer. Unfortunately, it will involve a considerable measurement error.

For this reason energy meters with specified accuracy class will use galvanical voltage measurement if ever possible.

Re: Current sensor ADE7763

From the IEC standard for el. meters:

1. L1, L2, L3, N power lines
2. power supplay ==>> +3.3, GND

N line must be connected direct to the GND !!
--------------------------------------------
shunt:
(just a piece of Nichrome wire) <<== wrong !
Shunt is very sensitive senzor, it can change parameters trough the time (current overload, time etc.)
You must stabilize its parameters under the heat, like cheramic material! see & study this: **broken link removed**
If from unknown reason 4kV appears on shunt (L1/2/3), .. the meter must survive (no smoke app. ) even more, it has to be as 'nothing happened',
high voltage disturbance must not change energy registers ! The standard says 4kV, meter manufacturers are testing on 6kV.

I did not say galvanic isolate must to be, but if some comp (ACS712 chip etc.) has brake down voltage 2kV it is not good parameter

Voltage is measuring like current over differential input on ADE7763. Of course you have to use resistor divider, (watch out! 6kV test)

The same story is with CT/Rogovsky senzor. Everything depends on schematic construction.

- - - Updated - - -

IEC:
The meters are professional elec. devices.

* They must work in their class * 24h/day 25 y !
Main tests: 8kV => new standard IEC 61000
1. Surge: 8kV 1.2/50micro s on all power lines both polarity 30times every minute, the same on all I/O incl. communication lines .. meter must work
2. Burst: 8kV 1n/50n, 75pulses in 5kHz every 300ms both polarity on all power lines, 4kV on com. and I/O .. meter must work
3. UHF electromagnetic field 80MHz-1GHz 30V/m .. meter must work
4. Radiation: in country frame
etc.

Re: Current sensor ADE7763

Most cheap single phase energy meters with 16 or 20A current range are using shunts, just a piece of Nichrome wire.

Click to expand...

(just a piece of Nichrome wire) <<== wrong !

Click to expand...

Good that some people still know about right and wrong :smile:

Seriously speaking, I'm reporting a fact, not discussing expectable accuracy. But as you are asking about it, a few remarks:
- Nichrome has a TC of about 50 ppm, not mind blowing but at least good for class 1. A thick wire (e.g. 1 - 2 mm diameter) will have excellent mechanical stability and usually not need a support.
- If you're concerned about shunt wire stability, don't even consider ACS712, which is worse by a factor of ten.
- Cheap current transformers (e.g. Talema products) don't perform so well when e.g. looking at linearity error

Re: Current sensor ADE7763


The first rule in mass electronic devices manufact.: do not trust to anybody, specially not to the components distributors even components manufacturers.
Make tests and statistic all the time, trough the time the truth will appear. Every mistake goes from your budget.

Re: Current sensor ADE7763

So is there any conclusion..... :?:

Re: Current sensor ADE7763

Different conclusions have been posted during the discussion, you have to decide.

it is easy to use current sensing resistor or hall effect IC. i think more accurate one is using a current sensing IC.