Why Current sensor to be placed in Positive Rail for DC current measurement?

[kalaianand]

I am using ACS709 (Hall Effect) current sensor for measuring DC link current for BLDC motor. Is it mandatory that, Current sensors to be placed in the Positive Rail? Why NOT Return path. Any disadvantage?

 

[KlausST]

Hi,

You can place it in gnd line also.

Klaus

 

[kalaianand]

Hi,

You can place it in gnd line also.

Klaus

NO dear friend, There is some disadvantage implementing the current sensing in Return Path.

 

[KlausST]

Hi,

NO dear friend, There is some disadvantage implementing the current sensing in Return Path.

Hey, great explanation "some disadvantge"...


In a DC path all current going through the positive line must come back through the GND line. In same time and same magnitude.
So where exactely do you see your "some disadvantage"?


Klaus

added: And why do you ask, when you have your answer already?

 

[kalaianand]

Hi,



added: And why do you ask, when you have your answer already?

LOL!!

One expert from TI (Texas Instruments) asked this question to me. He knew answer for this. But he didn't tell me answer. But i am trying to find answer.

It may be due to small potential difference occur due to very small amount of resistor used in any current sensor inside. Or due to Ground noise will occur in Noisy environment/ Application. But i am not Sure about this.

 

[red_alert]

One (minor) disadvantage could be the small (and variable) voltage drop across the measuring shunt which may perturb some other ground referenced measurements like output feedback voltage (especially for low voltage/high output current designs).

Later edit: We were posting at the same time.

 

[FvM]

I fear the question is meaningless without referring to a specific circuit and the answers aren't but guesses.

 

[kalaianand]

Hello Mr. Fvm,

Question is simple. For any current sensing circuit using Hall effect sensor, All most all designers choose to put current sensor in +ve rail not on the Return Path (-Ve). Still you did't get my question?

 

[FvM]

Without additional circuit requirements, there's no reason to place a current sensor not in the return path.

But the return path may be shared with a driver circuit (not mentioned in your question) and its supply current could e.g. disturb the current measurement.

 

[KlausST]

Hi,

what are you talking about?

One time you speak of hall sensor another time you speak of shunt current measurment..

A hall sensor needs no shunt. It even can used with superconductor material, it just reacts on the magnetic field.

****
My answer relies on the use of hall sensors.
For sure a shunt gives additional ground offset caused by resistance and even worse with it´s inductivity.


Klaus

 

[FvM]

Only post #6 did talk about shunt measurements.

 

[colin55]

"A hall sensor needs no shunt."

The hall sensor measures the magnetic flux ACROSS THE SHUNT.

 

[ahsan_i_h]

Question is simple. For any current sensing circuit using Hall effect sensor, All most all designers choose to put current sensor in +ve rail not on the Return Path (-Ve). Still you did't get my question?

The question is not simple. The +Ve and -Ve rails are nodes. You can not put current sensor in a node. You can put it in a branch. So you need to specify the branch.

 

[KlausST]

Hi,

@colin55.

Could you explain more detailed, with links to datasheets or application notes.

In my eyes the classic current measurement with a shunt is to use the voltage drop across a known resistor (shunt).
The usually very small voltage is amplified. Here one needs the voltage drop but on the other hand this voltage drop may case problems.

With the simple hall sensors there is no need for a voltage drop and therefore no need for a shunt. As you say it measures the magnetic field/flux.

The pitfall with hall sensors is the geometry and distance from trace to sensor and the influence of other currents/ magnetic fields. To improve on this some vendors build special hall current measurement devices with a fixed built in current path. But this path i don't call a shunt. But maybe one could call it shunt - without desired significant voltage drop.

Klaus