need info AC current meters

Thread Starter

raffter

Joined Feb 28, 2008
113
Hi all,

WHY do current meters have monitoring on BOTH lines (live,neutral)?? Im talking about current transformers... then both "conditioned" signals are sent to 2 ADC inputs(post processor) ....

how does it measure current?? getting the average value of the two? etc??

tia

-Ralph
 

beenthere

Joined Apr 20, 2004
15,819
Can you cite a source for that setup? That is usually confined to protective devices like ground fault interrupters, where a difference in hot and neutral implies a current leakage and forces a trip on the breaker.

In a circuit, the current in the hot lead is the same as the current in the neutral, so measuring one is the same as measuring the other.
 

Thread Starter

raffter

Joined Feb 28, 2008
113
Can you cite a source for that setup? That is usually confined to protective devices like ground fault interrupters, where a difference in hot and neutral implies a current leakage and forces a trip on the breaker.

In a circuit, the current in the hot lead is the same as the current in the neutral, so measuring one is the same as measuring the other.
here is one such source:

http://www.atmel.com/dyn/resources/prod_documents/doc2566.pdf

now I have made my own CT for this(metal core of course) 1:1000 turns ratio with a 100R burden resistor... I have just made the front end , to make things much simpler.. I can make the PIC/firmware later....

my problem is this: measuring resistive loads IS OK (like incandescent bulbs and soldering iron<ceramic heater element>) now with reactive loads(like a TV set, I get erronous readings..registering MORE than its actual consumption...

IS there a trick on HOW to measure with CTs and reactive loads?? or what math equations are involved?
 

leftyretro

Joined Nov 25, 2008
395
here is one such source:

http://www.atmel.com/dyn/resources/prod_documents/doc2566.pdf

now I have made my own CT for this(metal core of course) 1:1000 turns ratio with a 100R burden resistor... I have just made the front end , to make things much simpler.. I can make the PIC/firmware later....

my problem is this: measuring resistive loads IS OK (like incandescent bulbs and soldering iron<ceramic heater element>) now with reactive loads(like a TV set, I get erronous readings..registering MORE than its actual consumption...

IS there a trick on HOW to measure with CTs and reactive loads?? or what math equations are involved?
Yes, reactive loads can skew a measurement that assumes a sine wave RMS type load. There were chips avalible from Analog Devices that would take a low level AC voltage and output a RMS value. It's the same as used on the better DMM that have can read true RMS voltage. So short of adding this kind of input conditioning I can't think of another method for you.
 

Thread Starter

raffter

Joined Feb 28, 2008
113
Yes, reactive loads can skew a measurement that assumes a sine wave RMS type load. There were chips avalible from Analog Devices that would take a low level AC voltage and output a RMS value. It's the same as used on the better DMM that have can read true RMS voltage. So short of adding this kind of input conditioning I can't think of another method for you.
can RMS value be "CODED" ?? say a PIC to process RMS value??
 

beenthere

Joined Apr 20, 2004
15,819
Easily, if you can make the conversion at peak voltage. RMS is pretty close to .707 of peak. It's also pretty simple to make the absolute value circuit from the National Semiconductor op amp collection and present the PIC with a voltage also very close to RMS.
 

Thread Starter

raffter

Joined Feb 28, 2008
113
Easily, if you can make the conversion at peak voltage. RMS is pretty close to .707 of peak. It's also pretty simple to make the absolute value circuit from the National Semiconductor op amp collection and present the PIC with a voltage also very close to RMS.

so uhmmm ADC input will suffice as input for RMS conversion??


regarding NatSemi op-amp collection, what specific circuit should I be looking at?
 
Last edited:

leftyretro

Joined Nov 25, 2008
395
can RMS value be "CODED" ?? say a PIC to process RMS value??
Not very easily, would take lots of processing resources. The input AC voltage wave shape, (actually representing the AC current being consumed) would have to be sampled many times per single AC cycle and then some calculus performed before you would have the data needed to do an accurate conversion to true RMS power being consumed. External special function IC circuit is much more practical I think.

PS: Just answered another thread on a similar topic and found this product. Seems it is possible but I suspect they are using a analog devices RMS converter chip.

http://www.sparkfun.com/commerce/news.php?id=204
 
Last edited:

leftyretro

Joined Nov 25, 2008
395
Easily, if you can make the conversion at peak voltage. RMS is pretty close to .707 of peak. It's also pretty simple to make the absolute value circuit from the National Semiconductor op amp collection and present the PIC with a voltage also very close to RMS.
That's only true when the AC wave form is a sine wave, which only happens with a pure resistive load. When there is a reactive load the AC current wave shape is no longer a sine wave and the .707 Vs peak rule does not apply. Voltage and current are no longer in phase with each other when there is a reactive load.

Lefty
 
my problem is this: measuring resistive loads IS OK (like incandescent bulbs and soldering iron<ceramic heater element>) now with reactive loads(like a TV set, I get erronous readings..registering MORE than its actual consumption...

IS there a trick on HOW to measure with CTs and reactive loads?? or what math equations are involved?
How do you know that your readings are erroneous? How do you know what the true reading is?

Are you trying to measure only current, or are you trying to measure power?
 

Thread Starter

raffter

Joined Feb 28, 2008
113
How do you know that your readings are erroneous? How do you know what the true reading is?

Are you trying to measure only current, or are you trying to measure power?
my analog front end "connects" to a Digital VOM(in DC mode) ...output in mV equals watts... my reference "loads" are a ceramic heater element soldering iron(20W) and a 100W incandescent bulb....

reading of 100mV is 100W...
 

Thread Starter

raffter

Joined Feb 28, 2008
113
Not very easily, would take lots of processing resources. The input AC voltage wave shape, (actually representing the AC current being consumed) would have to be sampled many times per single AC cycle and then some calculus performed before you would have the data needed to do an accurate conversion to true RMS power being consumed. External special function IC circuit is much more practical I think.

PS: Just answered another thread on a similar topic and found this product. Seems it is possible but I suspect they are using a analog devices RMS converter chip.

http://www.sparkfun.com/commerce/news.php?id=204
Leftie,

is THAT an LC filter?? :eek: I can see maybe a 0.001R shunt...

EDIT:

RC filter.. sorry..
 
Last edited:

Thread Starter

raffter

Joined Feb 28, 2008
113
oh I think its NOT an RC filter... more-or-less its an RC line "reducer" :rolleyes:

to supply power to the circuit... now I get it..

so "sensing" is purely an ohmic shunt
 
Top