Question re. shunts to measure amperage

[Wayne Lundberg]

I just learned that you can measure the change in resistance in a shunt
connector between source and load as the voltage or amperage changes due to
on/off, and load characteristics. I don't understand why this happens. If
you power a machine like a vacuum cleaner, does the resistance of the
extension cord change when turning off to on? and when tackling a tough
spot?

If this is true, I should have learned this fifty years ago and it would
have saved me a lot of time and would have helped in trial and error
experiments.

Anybody?

Wayne

 

[Eeyore]

I just learned that you can measure the change in resistance in a shunt
connector between source and load as the voltage or amperage changes due to
on/off, and load characteristics. I don't understand why this happens. If
you power a machine like a vacuum cleaner, does the resistance of the
extension cord change when turning off to on? and when tackling a tough
spot?
Yes.

If this is true, I should have learned this fifty years ago and it would
have saved me a lot of time and would have helped in trial and error
experiments.

Whatever it is, be it a shunt resistor or a cable, has a temperature coefficient
of resistance. As the current increases, so does the temperature from I2R
heating effects, so the resistance changes ( upwards ) and in the reverse
direction as the load reduces. High accuracy shunts may be made of materials
with negligible tempco to minimise the effect though.

Incidentally, the same happens in loudspeakers and causes thermal 'power
compression' at high volumes as the voice coil resistance increases.

There is also a power compression effect in loudspeakers if the voice coil ends
up 'excursing' out of the magnetic gap or at least the linear part of it. A
popular way of reducing the effect is to use long voice coils.

Graham

 

[Chris]

I just learned that you can measure the change in resistance in a shunt
connector between source and load as the voltage or amperage changes due to
on/off, and load characteristics. I don't understand why this happens. If
you power a machine like a vacuum cleaner, does the resistance of the
extension cord change when turning off to on? and when tackling a tough
spot?

If this is true, I should have learned this fifty years ago and it would
have saved me a lot of time and would have helped in trial and error
experiments.

Anybody?

Wayne

Hi, Wayne. Placing a small series resistor, or shunt, in series with
the load is an easy way to measure current. You just measure the
voltage across the shunt, which is proportional to the current.

The problem is that resistances change with temperature. Copper in
particular is pretty bad here -- its temperature coefficient of
resistance is about +0.4% per degree C.

Let's say you've got a 1 milliohm copper shunt, and you're measuring 0
to 50 Amps = 0 to 50mV. At 50 amps, the shunt will be dissipating 2.5
watts. Let's assume that 2.5 watts causes the temperature of the
copper to rise 20 degrees C. That means the actual resistance of the
shunt will be more like 1.05 milliohms, leading to a 5% error in your
reading.

This is a limitation in measuring current by measuring the voltage drop
across an extension cord. Yes, as the current increases, the extension
cord heats up, and its resistance does change. But the change in
resistance is due to the change in temperature. You can put a shunt in
an oven, with no current, and measure resistance change just the same
with an ohmmeter.

A plain old resistor has a temperature coefficient of around 300
ppm/degree C, or 0.03% per degree C. Good wirewound resistors have
temperature coefficients better than 25 ppm/degree C, or 0.002% per
degree C.

Most high current measurement shunts are made of alloys of metals whose
postive and negative temperature coefficients cancel out, leading to a
more accurate reading at high current/power (close to 20ppm/degree C).

Does this answer your question? If not, feel free to post again.

Cheers
Chris

 

[Wayne Lundberg]

Hi, Wayne. Placing a small series resistor, or shunt, in series with
the load is an easy way to measure current. You just measure the
voltage across the shunt, which is proportional to the current.

The problem is that resistances change with temperature. Copper in
particular is pretty bad here -- its temperature coefficient of
resistance is about +0.4% per degree C.

Let's say you've got a 1 milliohm copper shunt, and you're measuring 0
to 50 Amps = 0 to 50mV. At 50 amps, the shunt will be dissipating 2.5
watts. Let's assume that 2.5 watts causes the temperature of the
copper to rise 20 degrees C. That means the actual resistance of the
shunt will be more like 1.05 milliohms, leading to a 5% error in your
reading.

This is a limitation in measuring current by measuring the voltage drop
across an extension cord. Yes, as the current increases, the extension
cord heats up, and its resistance does change. But the change in
resistance is due to the change in temperature. You can put a shunt in
an oven, with no current, and measure resistance change just the same
with an ohmmeter.

A plain old resistor has a temperature coefficient of around 300
ppm/degree C, or 0.03% per degree C. Good wirewound resistors have
temperature coefficients better than 25 ppm/degree C, or 0.002% per
degree C.

Most high current measurement shunts are made of alloys of metals whose
postive and negative temperature coefficients cancel out, leading to a
more accurate reading at high current/power (close to 20ppm/degree C).

Does this answer your question? If not, feel free to post again.

Cheers
Chris

Chris, Eyeore... You guys have promted me to take out my highly accurate VM
and nest a thermocouple to a resistor between motor and source as a means
to measure the temperature differences in load characteristics for my
experiment. I don't really need to know actual current demand until we come
up with the optimum design and then we will need to determine actual power
requirements for battery life and the rest for the autonomous robot being
designed to save lifes from the incredibly nasty IEDs in the field.

 

[Wayne Lundberg]

Whatever it is, be it a shunt resistor or a cable, has a temperature coefficient
of resistance. As the current increases, so does the temperature from I2R
heating effects, so the resistance changes ( upwards ) and in the reverse
direction as the load reduces. High accuracy shunts may be made of materials
with negligible tempco to minimise the effect though.

Incidentally, the same happens in loudspeakers and causes thermal 'power
compression' at high volumes as the voice coil resistance increases.

There is also a power compression effect in loudspeakers if the voice coil ends
up 'excursing' out of the magnetic gap or at least the linear part of it. A
popular way of reducing the effect is to use long voice coils.

Graham

Graham, Cris - I have some high resistance wire used in electric kilns. If I
took a couple of inches of this stuff and connected it between my source and
load, then measured what... voltage? ... resistance? along the heating wire?
Or... should I bandage some thermocouple pairs and measure voltage changes
from temperature changes?

Wayne

 

[jasen]

I just learned that you can measure the change in resistance in a shunt
connector between source and load as the voltage or amperage changes due to
on/off, and load characteristics. I don't understand why this happens. If
you power a machine like a vacuum cleaner, does the resistance of the
extension cord change when turning off to on? and when tackling a tough
spot?

If this is true,

you can measure current by the voltage drop in a series resistance as
the load changes and/or is switched.

Dunno about what you said.

"clamp meters" are neat too.

 

[Chris]

Chris, Eyeore... You guys have promted me to take out my highly accurate VM
and nest a thermocouple to a resistor between motor and source as a means
to measure the temperature differences in load characteristics for my
experiment. I don't really need to know actual current demand until we come
up with the optimum design and then we will need to determine actual power
requirements for battery life and the rest for the autonomous robot being
designed to save lifes from the incredibly nasty IEDs in the field.

Hi, Wayne. You know, some milliohmmeters made for the winding industry
have a built-in T/C for measuring ambient temp and compensating.

If you're determining motor current for a battery application, a shunt
probably isn't going to cut it. The shunt will affect the performance
of the motor. And you're going to be very inaccurate in trying to
infer current from motor temp.

You might be better off looking at a hall effect DC current sensor to
measure current. These output a DC voltage proportional to the
current, and do this without affecting the load. Some of these have
frequency response over 10KHz, which is sufficient to read a quickly
changing motor load with a scope. You can also use these with a
satndard DMM for average readings.

You can home brew one of these, or purchase it from DMM or scope
accessory manufacturers.

I'd keep the T/C, though. Motor temp in high ambient temp conditions
might be good to know.

Good luck
Chris

 

[Chris]

Chris, Eyeore... You guys have promted me to take out my highly accurate VM
and nest a thermocouple to a resistor between motor and source as a means
to measure the temperature differences in load characteristics for my
experiment. I don't really need to know actual current demand until we come
up with the optimum design and then we will need to determine actual power
requirements for battery life and the rest for the autonomous robot being
designed to save lifes from the incredibly nasty IEDs in the field.

For purposes of obtaining earmarks for military contracts, the
investment of a minimum of $12K or 0.1% of projected sales in
transactional lobbying expense (whichever is greater) is far more
important than any technical expertise.

And since there's a lot of competition for IED R/C car contracts, you'd
best engage a Senator as well as at least one House member. Of course,
the transactional lobbying expense noted above applies to each
honorable representative. Incorporate your business in a state and
district with representatives of the appropriate political party.

You can always make it work after you get the contract. Just keep your
priorities in order. ;-)

Cheers
Chris

 

[Wayne Lundberg]

For purposes of obtaining earmarks for military contracts, the
investment of a minimum of $12K or 0.1% of projected sales in
transactional lobbying expense (whichever is greater) is far more
important than any technical expertise.

And since there's a lot of competition for IED R/C car contracts, you'd
best engage a Senator as well as at least one House member. Of course,
the transactional lobbying expense noted above applies to each
honorable representative. Incorporate your business in a state and
district with representatives of the appropriate political party.

You can always make it work after you get the contract. Just keep your
priorities in order. ;-)

Cheers
Chris

Thanks Chris... our good fortune is to have a Washington insider on our team
along with a couple of experienced hands at this kind of funding and
transfer of technology later on to the commercial world. Also good fortune
to have angel financing for startup costs and first serious prototypes.

And thanks to your previous note on the technical side of the issue. I have
a lot to learn, and hopefully as our seed money comes in will be able to
contract occasional moonlight engineering talent to help us through the
technical challenges.