CT burden resistor calculation

Dear all,

I am trying to measure Current flowing into CT by converting into analog value.i found example of wring in below link

Can someone suggest me How can i choose the R1,R2,R3,C1 value . Where R3 is burden across CT.

Is there any sample calculation involved.Please let me know explanation with some calculation will be helpful.

They explained about selection but need more clarity. I need better accuracy. why on example voltage devide by 2.5v. Cant we use 5V full scale reading.

R1 and R2 are just a voltage divider to set the ADC scale to center at Vcc/2, which is required for AC. They should be of equal values, 10K to 1Meg, 100K is fine.

C1 is a filter capacitor and is not critical. 0.1 to 10 uF is fine.

R3 should be chosen to match whatever CT you are using (check the data sheet), but somewhere in the range of 10 to 220 Ohms is typical.

You could use a schottky rectifier bridge with the burden resistor on the DC side of the rectifier to make use of the full 5V swing. Value of burden resistor must be chosen depending on the maximum current to be measured and the specific CT. Please note that this configuration with rectifier and burden after rectification sacrifices low range accuracy but may give better performance in the high ranges. Also BOM cost is higher.

AJITnayak:
They explained about selection but need more clarity. I need better accuracy. why on example voltage devide by 2.5v. Cant we use 5V full scale reading.

The analog input can only handle values in to 0.0 to 5.0V range, so for an ac signal you need to bias
the input to 2.5V, then you only have +/-2.5V range for signals.

Or put another way the amplitude or peak voltage of the ac waveform must be 2.5V or lower,
which for sinusoidal wave means the rms value must be 1.76V or less.

So if the max rms current to measure is I, then the burden resistor should be ideally 1.76*N/I,
where N is the turns ratio of the CT.

I would strongly suggest getting a CT with an in-built burden resistor. If a CT is operated with
the burden resistor missing, or with poor connection, the result is catastrophic as massive voltages
will be generated causing arcing and perhaps fire.

Having the burden resistor not directly connected to the CT is a safety risk.

A small point , fit the burden resistor securely - an open circuited CT can generate very high voltages

Hi,
Have you got the CT yet.

The datasheets on most of them have application circuits with burden resistor calculations.

Tom... :slight_smile:

@MarkT do you have any part no for External CT which comes with Burden CT.

How there is chance of file. external CT are used. it read reading in MA which converted into analog voltage internally using combination of resistor.

R1 and R2 are just a voltage divider to set the ADC scale to center at Vcc/2, which is required for AC. They should be of equal values, 10K to 1Meg, 100K is fine.

Can i use divider output in 0-5v range instead of 0-2.5V

You could use a schottky rectifier bridge with the burden resistor on the DC side of the rectifier to make use of the full 5V swing. Value of burden resistor must be chosen depending on the maximum current to be measured and the specific CT. Please note that this configuration with rectifier and burden after rectification sacrifices low range accuracy but may give better performance in the high ranges. Also BOM cost is higher.

Is there any proven circuit available with calculation involved , so i can tested out circuit.

Have you got the CT yet.

The datasheets on most of them have application circuits with burden resistor calculations.

I am waiting for CT. I have informed my supplier to share datasheet.
I need to have better accuracy. weather 10bit resolution ADC from arduino uno is enough for application or need to build external accuracy. +/- 1A to 2 A i need it accuracy.

My external measuring current range is 0.8A-25A . If you can tell me what are specific parameter need to check while selecting CT it wil be more helpful. once i got spec i will share it here.

Hi,
What is your CT that you have ordered?
Can you post a link please?

Thanks.. Tom.. :slight_smile:

If you are buying new , why not get a current transducer - something that gives 4-20mA DC output , rather than a bare CT, solves a lot of problems , both in hardware and software .

Amelec ASC330 for example

The idea with the burden resistor after a schottky rectifier was taken from this site.

Not aware of any proven circuit except the one I built to test it out. Once calibrated with a regular clamp meter my arduino setup displayed accuracy better that the +-1A you require but I only tested it between ~2-8A. Not sure how this circuit will perform under 1A. In my setup I used a Talema AZ-0500 current transformer which has a ratio of 500:1 and then a 100Ohm resistor connected on the DC side of the schottky rectifier bridge.

Here you can see a small spreadsheet I made when testing the circuit. As you can see it does not scale very well in the lower ranges and one should expect even worse accuracy when going lower than 1A. But the performance in the higher range is very good. It all depends on what you are using it for. My intended application was a simple switch that would disconnect a load (fixed, resistive power) when the total load was too high to save the main fuses and then re-connect it again when the system load had dropped to a safe level.

I am using this for motor protection device application. My motor current range will be 0.8A-25A Max. I have done previously with Phase monitoring application . to be safer side considering current parameter its easy to control motor action.

Here my application demand higher current not small range.In the circuit there is no Burden resistor used.
Is there equation you have derived for Arduino read the voltage.

The burden resistor is the one designated as R12 in the circuit. From what I understand this kind of circuit and especially when used with a burden resistor selected to achieve a large voltage swing may introduce a large phase shift which may not be the best choice when measuring an inductive load (electric motor), but this is above my level of knowledge. The voltages in the spreadsheet was taken from my oscilloscope because I was also curious what the wave form would look like after schottky rectification.

I have pasted #1 http://www.the-diy-life.com/simple-arduino-home-energy-meter/
Its similar to my application for measurement of current. I have confusion with few theory mentioned.

  1. dropping voltgae to 2.5v , why not considered 5V
    2)Calculating RMS current
    double RMSCurrent=((maxCurrent -516)X0.707)/11.8337;
    How 0.707 amd 11.8337 derived?
  1. The reason for this is that unless you use some kind of rectification you get AC-signal to your arduino analog input. Negative voltages on those inputs are bad. So you either use a voltage divider and use half of the maximum signal input voltage (2.5V) OR you can use a bridge rectifier (or half wave if you prefer it over full wave) and use the entire 0-5V swing.

  2. Since I had too many unknowns I skipped the calculations and simply just mapped the values using a good clamp meter as base. Assuming the your clamp meter is accurate, your arduino will be close to that as well.

How 0.707 and 11.8337 derived?

The 0.707 the the relationship between the peak value of a perfect sine wave to the RMS value. Note as the waveform departs from a perfect sine wave this value is less accurate.

Iā€™m not 100% sure but I believe the 11.8337 is a combination of the CT transfer function and the divider.

If this is just for motor protection , why not use a simple over load relay with fuses to give full range protection - cheap , reliable and widely used , and do a good job .

If this is just for motor protection , why not use a simple over load relay with fuses to give full range protection - cheap , reliable and widely used , and do a good job .

MY application demand lot of feature like under /over current ,current unblance, short circuit current, single phase failure,rotor rock current,dry run .earth fault detection

Where for all current in base reference. I am planned to use 3 CT in 3 phases. 1 CBCT for earth fault. thats reason going with external CT not with overload relay

I am also looking for implementation of this feature .if any one guide me how it can be done it will be more helpful.

I have attached image for reference. It shows current Vs time characteristic . Is there any equation available.

higher current lower the trip time. lower current ,higher the setting time. Its inversely proportion. how it can be generated for various curve 2C/5C/10C

IT_char.jpg

Motor Protection Principles.pdf (680 KB)

I am far from an expert but I doubt that a 328P will do what you want. Considering you would like to protect 3-phase-motors that may very well be driven by a variable frequency drive up to about 400Hz and factoring in all signals you need to measure and use for determining status, I don't think the 328P has enough power and inputs.

If I were you I would start with measuring current and maybe phase shift using three current transformers. If you need best ADC performance possible I would use a 4.096V voltage reference for the 328P and a 2.048V voltage reference to create a ground point for the current transformers. That way you can use only a burden resistor and nothing else to read the motor current and the measurement should be very accurate.

If I were you I would start with measuring current and maybe phase shift using three current transformers. If you need best ADC performance possible I would use a 4.096V voltage reference for the 328P and a 2.048V voltage reference to create a ground point for the current transformers. That way you can use only a burden resistor and nothing else to read the motor current and the measurement should be very accurate.

Thats only i am asking i need better accuracy . its has 10 bit resolution. 0.8A-30A current measuring range.
Considering CT of https://www.promelec.ru/pdf/AC-1030_Jun-06.pdf

give me best example of schematic for better accuracy and calculation involved.

Large electric motors can generate very high currents when starting up. I have absolutely no idea HOW large these can be but assuming 2 time normal running current (normal running current being 30A RMS) startup current can reach 60A RMS or 85A peak.

Again these are just numbers based on fictive numbers to explain the concept.

85A peak with a ratio of 1000:1 (the CT you posted the data sheet for) and a maximum voltage swing of 2V means we should try a and use a burden resistor of about ~23 Ohm

U=2V
I=85mA (1000:1 of 85A primary current)
R=U/I=2/0,085=23,5

Since the signal is AC we need to create a fixed, "fake" ground level for the ADC conversion. Using two voltage references, one at 2.048V for the ground level and one at 4.096V to make absolutely sure that the ground level is exactly 50% of maximum input voltage you get 2.048V or 9bits theoretical resolution. Using 85A as maximum peak current 85/512=0,166 so maximum theoretical resolution with this setup is ~0.17A

Note that the CT you linked may work best with burden resistor around 100 Ohm but that would generate too high input signal for the arduino ADC inputs. You may want to look at another CT. Also, you need to make measurements using an oscilloscope first to verify the value of the burden resistor before you connect it to an input that may be damaged if you exceed 5V referenced to ADC ground.