Electronics > Metrology

Shunt resistor issue

<< < (8/8)

The Soulman:
Not for the OP to worry about, just to get some brainwaves flowing and to get the feet of the ground...  :-+


--- Quote from: Gyro on July 13, 2016, 07:32:29 pm ---
"HINT: It is important to connect the I (current) leads outside the U (potential) leads"



--- End quote ---

Now imagine a straight bare copper wire of 10 meters long with a end to end resistance of 30 Ohms and we have a connection terminal on each end called A and a set of connections spaced 1 meter apart somewhere along the line called B.
Now if we force 1 amp of current from A to A we get a total voltage drop of 30 volts across the wire and at our measurement terminals B We measure
3 volts, all is well.
If we put the same 1 amp of current thru terminals B and measure voltage at A, we´d measure 3 volts  :o the same.

And what if we swap one A terminal for one B terminal, 3 volts again....  :popcorn:

StillTrying:
Right...

All my posts are carefully considered believe it or not, shocked was/is the right word.

You keep asking why the resistance at 1-1 is different from 2-2/3-3 for some reason, it doesn't matter what it is at 1-1, you can make it anything you want because the lugs are not measuring there. But..

The only way to measure resistance is to force a known steady current between 2 carefully selected points and then measure how much force(voltage) was needed to do it. The more force(voltage) that was needed the higher the resistance is.

When you force this known steady current though 1-1 or 2-2 the current does not flow like this. shunt1.jpg

If you could see the current flowing between 2 points on a surface it would look more like this. shunt2.jpg

Most of the current total flows directly through the middle line between the 2 points, but there are smaller currents taking longer curved paths, each side of the center line, and even smaller currents flowing all over the place. If you need a very accurate measurement, then all these smaller currents add up to quite a few % of the total current.

Drawing these currents on your shunts would look vaguely like this. shunt3.jpg

Right away, hopefully, you can see that some of the lower parts of the 1-1's current paths are missing - cut off, by the cutout, compared with 2-2. So when you come to force the steady known current between 1-1 it takes a bit more force(voltage) to get it to flow though the paths which are still there. And of course this bit of more force means 1-1 has a bit more resistance.

Here's one you prepared earlier. shuntg.jpg

If you moved points 1-1 down towards the bottom edge even slightly, then more of 1-1's current paths would be missing, and show up as more resistance. If you moved points 1-1 up a bit it would give 1-1 more current paths and less resistance, you'll get a different resistance nearly every time.

The 4 wire measurement bit to avoid all these errors has already been done.. and done, so just do it for now.

All this won't be 100%, but close enough for me!

LukaszRed:
@up StillTrying
Thanks! This is sort of answer that puts more light into some corners. So what you are trying to say that the current density affects the resistance? We agree with that and we did a simulation of that phenomenon and we can confirm that resistance readings change across the shunt.
So it looks like when we measure the resistance with cropico we are applying the 10A current to the pins which potentially makes the current density relatively high across the measuring path. That's why we are getting high resistance results. Of course ideally we want to apply the current on the component ends to provide that current flow and only voltage leads to the pins.

If I am getting this right I think we are nearly there.

Alex Nikitin:

--- Quote from: LukaszRed on July 14, 2016, 09:06:59 am ---So it looks like when we measure the resistance with cropico we are applying the 10A current to the pins which potentially makes the current density relatively high across the measuring path. That's why we are getting high resistance results.
--- End quote ---

When you measure the resistance with Cropico on the pins you add the resistance of the pins to the measured value. A simple estimate is that each pin has about 25-30uOhm resistance, hence +50uOhm reading.

Cheers

Alex

Navigation

[0] Message Index

[*] Previous page