design wattmeter

Thread Starter

firend

Joined Feb 18, 2012
44
i am designing a single phase wattmeter. it requires range from 0 to 500 W. U =220 V.
can anyone help me?
 

Brownout

Joined Jan 10, 2012
2,364
Sure. Watts is Volts times Amps. You need to measure the volts and amps and do the multiplication. If you only need accuracy as good as your confidence in the volts, you only need to measure the amps and do the multiplication.
 

DerStrom8

Joined Feb 20, 2011
2,373
I agree with Brownout. It should be quite simple. All you need to do is have something that measures the voltage, something that measures the current flowing between the terminals, and some sort of logic system or microcontroller to do the multiplication. It should be quite easy. Have you ever done any microchip programming?
 

crutschow

Joined Mar 14, 2008
23,157
If you want true watts then you need to multiply the instantaneous current by the voltage at the same instant, do that for many samples over the compete waveform cycle, and then average all the readings. Multiplying the waveform average (RMS) volts by average watts gives volt-amps which gives true power only for a power factor of one.

Alternately you can use a dedicated energy measuring IC such as one of these.
 

Thread Starter

firend

Joined Feb 18, 2012
44
I agree with Brownout. It should be quite simple. All you need to do is have something that measures the voltage, something that measures the current flowing between the terminals, and some sort of logic system or microcontroller to do the multiplication. It should be quite easy. Have you ever done any microchip programming?
i studied programing assembler. i can multiply voltage and current. but i can't measure power factor.
 
Last edited:

crutschow

Joined Mar 14, 2008
23,157
i studied programing assembler. i can multiply voltage and current. but i can't measure power factor.
If you multiply the instantaneous short samples of voltage and current as I described then you don't need to measure the power factor. The power factor is included as part of the measurement.
 

Thread Starter

firend

Joined Feb 18, 2012
44
whether the tolerance of the wattmeter is large?
my english isn't good.
thank you.
 

crutschow

Joined Mar 14, 2008
23,157
The tolerance is determined by the accuracy of the circuits you use to measure the voltage and current.
 

Thread Starter

firend

Joined Feb 18, 2012
44
The tolerance is determined by the accuracy of the circuits you use to measure the voltage and current.
i add DC signal to input signal adc is positive.
it is difficult for me to sample simultaneously two signals at the same time.
what do you have algorithm solve to problem?
thanks in advance!
 

crutschow

Joined Mar 14, 2008
23,157
i add DC signal to input signal adc is positive.
it is difficult for me to sample simultaneously two signals at the same time.
what do you have algorithm solve to problem?
thanks in advance!
You can use two A/D converters sampling at the same time, one for current and one for voltage. Then you read the two values into a microprocessor, multiply the two values, and store them until you have taken samples over a complete cycle of the waveform (perhaps 50-100 samples per waveform). Then you calculate the average of the values over one cycle to get the power.
 

Thread Starter

firend

Joined Feb 18, 2012
44
i worked with two adc. i want to use one adc with 8 input.in order to sample voltage and current.i am going to use algorithm linear interpolation to relugate.where can i reference? can somone help me?
thank very much!
 

Thread Starter

firend

Joined Feb 18, 2012
44
You can use two A/D converters sampling at the same time, one for current and one for voltage. Then you read the two values into a microprocessor, multiply the two values, and store them until you have taken samples over a complete cycle of the waveform (perhaps 50-100 samples per waveform). Then you calculate the average of the values over one cycle to get the power.
hello!
i have a circuit with a 8051, adc0809 and lcd(lm016).the input of the adc is 4V.the output of the adc is (204D) equivalent to 11001100B. i want to display 4 on the lcd. i can only display 204.
can you help me?
thanks a lot.
 

crutschow

Joined Mar 14, 2008
23,157
hello!
i have a circuit with a 8051, adc0809 and lcd(lm016).the input of the adc is 4V.the output of the adc is (204D) equivalent to 11001100B. i want to display 4 on the lcd. i can only display 204.
can you help me?
thanks a lot.
Connect the ADC output to the 8051 and convert the binary by multiplying it by a factor of 400/204.
 

Thread Starter

firend

Joined Feb 18, 2012
44
Connect the ADC output to the 8051 and convert the binary by multiplying it by a factor of 400/204.
excuse me.i use 8051 and lcd(lm032).i want to display generated characteristic. but i can't.
can lm032 display charateristic? which lcd can do?
thanks a lot!
 

Thread Starter

firend

Joined Feb 18, 2012
44
The tolerance is determined by the accuracy of the circuits you use to measure the voltage and current.
I sample voltage and current but the tolerance of result is large.
I think the reason is tolerance of ADC. I use ADC0808.
Do you have the way to improve ?
thanks a lot.
 

crutschow

Joined Mar 14, 2008
23,157
The problem is likely tolerances in your voltage and current sample circuits, not the ADC.

But either way, one way to reduce the error is to calibrate the system. Generate a known accurate voltage and current (or measure it with accurate instruments) and use that to adjust the value you read using software e.g. if the value you read is 5% too high then multiply each reading by 95% or subtract 5% from each reading (whichever would be easier with the software) to generate the correct value.
 

panic mode

Joined Oct 10, 2011
1,775
all comments so far are about designing wattmeter from ground up which is fine. but it may be worth mention that there is plethora of ready chips just for that - measure energy and power, all you need to do is talk to them through I2C or SPI bus and you get all info you will ever need (real power, imaginary power, voltage, current, phase angle, you name it).
 

Thread Starter

firend

Joined Feb 18, 2012
44
If you want true watts then you need to multiply the instantaneous current by the voltage at the same instant, do that for many samples over the compete waveform cycle, and then average all the readings. Multiplying the waveform average (RMS) volts by average watts gives volt-amps which gives true power only for a power factor of one.

Alternately you can use a dedicated energy measuring IC such as one of these.
hello!
I want to communicate between controller and PC. I am going to use visual basic 6.0. the sample ADC takes 100us.the baud rate is 9600.
how long does it take to transmit? can I transmit continuously?
can PC receive right data ?
 
Top