Markings on Big shunt - what is R?

[Hawker]

I have two large current shunts a client gave me.
One is marked 25 amps 50mV SE company
The other is 20amps 50mV.

Am I to take it that they read 50mV at rated current?
Or in other words the 25A one is .002 ohms and 20amp one is .0025 ohms?
My VOM isn't very accurate down there.

 

[Steve]

I have two large current shunts a client gave me.
One is marked 25 amps 50mV SE company
The other is 20amps 50mV.

Am I to take it that they read 50mV at rated current?
Or in other words the 25A one is .002 ohms and 20amp one is .0025 ohms?
My VOM isn't very accurate down there.

As far as I know, you are correct. Every shunt I've used is marked
for full scale current, and output voltage at that current.

 

[[email protected]]

I have two large current shunts a client gave me.
One is marked 25 amps 50mV SE company
The other is 20amps 50mV.

Am I to take it that they read 50mV at rated current?
Or in other words the 25A one is .002 ohms and 20amp one is .0025 ohms?
My VOM isn't very accurate down there.

If you want to measure that sort of resistance, you need a four-
terminal (Kelvin) measuring system.

Upmarket multimeters offer this facility - the Thurlby-Thandar 1906
(Farnell order code 724-026) offers four terminal resistance
measurement, but since the resoultion only goes down to 1 milliohm, it
wouldn't do you much good.

Thurlby-Thandar do offer a micro- and milli-ohmeter - the BS407
(Farnell order code 381-2364) which is somewhat more expensive, but
can resolve resistances up to 1.999 milliohm to one micro-ohm, and
19.99 milliohm to 10 uohm.

Top of the line Hewlett-Packard (now Agilent) and Datron multimeters
do better than the TTI 1906, but cost quite a bit more.

 

[John Devereux]

If you want to measure that sort of resistance, you need a four-
terminal (Kelvin) measuring system.

Upmarket multimeters offer this facility - the Thurlby-Thandar 1906
(Farnell order code 724-026) offers four terminal resistance
measurement, but since the resoultion only goes down to 1 milliohm, it
wouldn't do you much good.

Thurlby-Thandar do offer a micro- and milli-ohmeter - the BS407
(Farnell order code 381-2364) which is somewhat more expensive, but
can resolve resistances up to 1.999 milliohm to one micro-ohm, and
19.99 milliohm to 10 uohm.

Top of the line Hewlett-Packard (now Agilent) and Datron multimeters
do better than the TTI 1906, but cost quite a bit more.

But for the OP to simply check his understanding of the markings, he
just needs to stick a few amps through it and measure the voltage
developed across the terminals of the shunt. I usually use a power
supply with an adjustable current limit.

 

[Spehro Pefhany]

But for the OP to simply check his understanding of the markings, he
just needs to stick a few amps through it and measure the voltage
developed across the terminals of the shunt. I usually use a power
supply with an adjustable current limit.

Exactly. Most cheap DMMs have 100uV or better resolution at full
accuracy, so at 3A you have 33 uohms resolution.

Be sure to apply the 3A to the outer (usually larger) terminals of the
shunt and read the voltage from the inner (usually smaller ) set of
terminals.

http://upload.wikimedia.org/wikipedia/commons/1/10/Shuntresistor50A.jpg

 

[[email protected]]

But for the OP to simply check his understanding of the markings, he
just needs to stick a few amps through it and measure the voltage
developed across the terminals of the shunt. I usually use a power
supply with an adjustable current limit.

He'd better measure the voltage twice, reversing the direction of the
current between readings - low level voltage measurements are
bedevilled by thermocouple voltages developed in the junctions between
dissimilar metals, and measuring with AC or at least reversing DC is
the standard way of getting rid of these offsets (or at least of
getting some idea how bad they are).

 

[John Devereux]

He'd better measure the voltage twice, reversing the direction of the
current between readings - low level voltage measurements are
bedevilled by thermocouple voltages developed in the junctions between
dissimilar metals, and measuring with AC or at least reversing DC is
the standard way of getting rid of these offsets (or at least of
getting some idea how bad they are).

Would you not need quite a big temperature differential (between the
ends of the shunt), for that to be significant?

 

[[email protected]]

Several degrees C (more than 5 is my guess) at least just to tick the
LSD. Forget about it. You'd see it anyway when the power supply is
turned off because of the large thermal mass.

The last time I was using a really good DVM to measure low voltages, I
found draft shields were absolutely essential to keep the voltage
stable. Most measuring set-ups have different metals all over the
place, and base metal thermocouples give you about 50uV/C.

http://evitherm.athena.as/default.asp?lan=1&ID=999&Menu1=999

You have to be very careful if you want to get anything useful out of
the uV resolution of a good DVM, particularly when measuring
resistances - which automatically involves dissipating some heat.

The fact that the manufacturers data sheets derate resistors linearly
against ambient temperature doesn't means that the temperature rise of
a resistor is a linear function of temperature - a low dissipations
the Maclaurin number is below 500 and you don't get any significant
convective cooling at all, so the the resistor is a lot warmer at low
power dissipations than you'd expect from linear extrapolation.

 

[Spehro Pefhany]

Would you not need quite a big temperature differential (between the
ends of the shunt), for that to be significant?

Several degrees C (more than 5 is my guess) at least just to tick the
LSD. Forget about it. You'd see it anyway when the power supply is
turned off because of the large thermal mass.


Best regards,
Spehro Pefhany

 

[John Larkin]

Several degrees C (more than 5 is my guess) at least just to tick the
LSD. Forget about it. You'd see it anyway when the power supply is
turned off because of the large thermal mass.

Right. And manganin has a low thermoelectric potential relative to
copper, just a few uV per K.

John

 

[John Devereux]

The last time I was using a really good DVM to measure low voltages, I
found draft shields were absolutely essential to keep the voltage
stable. Most measuring set-ups have different metals all over the
place, and base metal thermocouples give you about 50uV/C.

So use a really bad DVM

 

[Spehro Pefhany]

So use a really bad DVM

The pairs for base metal thermocouples are chosen to give a relatively
high voltage for a given temperature difference (among other things).

Typical connection material pairs are something like 5:1~10:1 better,
and you'll very seldom see anything other than a "0.0" mV if you go
around probing random bits of metal that have just been handled by
30°C fingers. A DC shunt will be made symmetrical in part so the
substantial self-heating at rated current won't cause thermocouple
voltages to affect the reading. Immediately when the current is shut
off, the meter should go to 0. If it doesn't, then the reading can be
corrected by that factor, but it will not be a problem with such a
setup and a 100uV resolution meter.

If you want to calibrate a really high current shunt (not just check
it) at a current orders of magnitude less than the rated current, then
such things would come into play (and perhaps you'd be using a meter
with 100nV resolution rather than 100uV), but the OP just wants to
assure himself that it's actually 50mV at rated current as marked.

We've supplied many, many, high current (up to 15,000A) DC measurement
systems using such shunts, BTW. They're usually between 50mV and 150mV
output at rated current.



Best regards,
Spehro Pefhany

 

[Spehro Pefhany]

The last time I was using a really good DVM to measure low voltages, I
found draft shields were absolutely essential to keep the voltage
stable.

If you're down in low uV DC territory or below, for sure. But we can
get 30 micro-ohm resolution out of a 100uV resolution measurement @3A,
which requires no special care. That's a resolution of ~1.5% of the
expected value, which is fine for the intended purpose.

Do you have a reference on techniques for nanovolt DC measurements?
I'd be interested in that.

You have to be very careful if you want to get anything useful out of
the uV resolution of a good DVM, particularly when measuring
resistances - which automatically involves dissipating some heat.

Heat in itself is not a problem. Nor even are thermal gradients. It
has to be an asymmetrical thermal gradient with dissimilar metals.

The fact that the manufacturers data sheets derate resistors linearly
against ambient temperature doesn't means that the temperature rise of
a resistor is a linear function of temperature - a low dissipations
the Maclaurin number is below 500 and you don't get any significant
convective cooling at all, so the the resistor is a lot warmer at low
power dissipations than you'd expect from linear extrapolation.

Maclaurin number? Do you mean the Reynolds number? or maybe the
Nusselt number?

Keep in mind that shunts typically dissipate a fair bit of heat at
full rated current. The OP's wee 25A one will dissipate in excess of
1W in normal use. Larger ones are in the hundreds of watts. The lack
of significant dissipation might affect the reading a bit.


Best regards,
Spehro Pefhany

 

[[email protected]]

If you're down in low uV DC territory or below, for sure. But we can
get 30 micro-ohm resolution out of a 100uV resolution measurement @3A,
which requires no special care. That's a resolution of ~1.5% of the
expected value, which is fine for the intended purpose.

Do you have a reference on techniques for nanovolt DC measurements?
I'd be interested in that.

No. My impression is that any such reference would start off by
recommending that you immersed the active part of the experiment in
liquid helium and go on from there. Microvolt DC measurements are
already tricky enough.

The English national standards laboratory at Teddington does offer "A
guide to measuring resistance and impedance below 1MHz" ISBN 0 9044557
31.1

http://www.npl.co.uk/cgi-bin/guide_info.pl?guide=105

I've got a copy, but can't recommend it - it doesn't say anything
silly, but it doesn't help you understand what is going on.

Heat in itself is not a problem. Nor even are thermal gradients. It
has to be an asymmetrical thermal gradient with dissimilar metals.

You can't dissipate heat without creating a thermal gradient. Ad hoc
connections are always asymmetric.

Maclaurin number? Do you mean the Reynolds number? or maybe the
Nusselt number?

Oops. Raleigh number - the Reynolds number applies to flow, the
Raleigh number applies to convection. Both show up in my Ph.D. thesis.
Why I keep on thinking the Rayeigh number is called the Maclaurin
number I'll never know. Check out

http://www.ldeo.columbia.edu/users/jcm/Topic3/Topic3.html

if you want a bit more detail.

Keep in mind that shunts typically dissipate a fair bit of heat at
full rated current. The OP's wee 25A one will dissipate in excess of
1W in normal use. Larger ones are in the hundreds of watts. The lack
of significant dissipation might affect the reading a bit.

Heat dissipation is proportional to the square of the current - 3A is
going to generate about 1.44% of the heat dissipated at 25A. 3mV isn't
too hard to measure, unless you expect the voltage to be accurate to a
couple of uV, but reversing the current is a useful check.

 

[[email protected]]

Right. And manganin has a low thermoelectric potential relative to
copper, just a few uV per K.

Never had any manganin voltage probes for any voltmeter I ever used.
Most of them looked like chromium-plated steel.

 

[Spehro Pefhany]

Never had any manganin voltage probes for any voltmeter I ever used.
Most of them looked like chromium-plated steel.

The shunt is probably manganin. The wires you put under the screws are
probably plated copper, (AWG22 or something like that) so copper for
T/C purposes, so the number is applicable.

BTW, the probes (if they are the probe type, and cheap) are most
likely nickel-plated brass.


Best regards,
Spehro Pefhany

 

[joseph2k]

If you're down in low uV DC territory or below, for sure. But we can
get 30 micro-ohm resolution out of a 100uV resolution measurement @3A,
which requires no special care. That's a resolution of ~1.5% of the
expected value, which is fine for the intended purpose.

Do you have a reference on techniques for nanovolt DC measurements?
I'd be interested in that.


Heat in itself is not a problem. Nor even are thermal gradients. It
has to be an asymmetrical thermal gradient with dissimilar metals.


Maclaurin number? Do you mean the Reynolds number? or maybe the
Nusselt number?

Keep in mind that shunts typically dissipate a fair bit of heat at
full rated current. The OP's wee 25A one will dissipate in excess of
1W in normal use. Larger ones are in the hundreds of watts. The lack
of significant dissipation might affect the reading a bit.


Best regards,
Spehro Pefhany

Go Sphero, i breifly googled and did not find any appropriate Mclaurin
number application, Reynolds number, as i had learned in school, is pretty
strictly related to aerodynamic drag of various shapes, but Nusselt number
appears to be applicable. Well placed question, i learned something as a
result.

 

[Steve]

If you want to measure that sort of resistance, you need a four-
terminal (Kelvin) measuring system.

Upmarket multimeters offer this facility - the Thurlby-Thandar 1906
(Farnell order code 724-026) offers four terminal resistance
measurement, but since the resoultion only goes down to 1 milliohm, it
wouldn't do you much good.

Thurlby-Thandar do offer a micro- and milli-ohmeter - the BS407
(Farnell order code 381-2364) which is somewhat more expensive, but
can resolve resistances up to 1.999 milliohm to one micro-ohm, and
19.99 milliohm to 10 uohm.

Top of the line Hewlett-Packard (now Agilent) and Datron multimeters
do better than the TTI 1906, but cost quite a bit more.

http://www.discovercircuits.com/H-Corner/1ampcurrent.htm

Probably not the most accurate thing, but may be helpful for smaller
shunt measurements.

 

[Spehro Pefhany]

Go Sphero, i breifly googled and did not find any appropriate Mclaurin
number application, Reynolds number, as i had learned in school, is pretty
strictly related to aerodynamic drag of various shapes, but Nusselt number
appears to be applicable. Well placed question, i learned something as a
result.

Thanks. We use the Reynolds number (and the Nusselt number) in mold
cooling calculations- you want a Reynolds number high enough to assure
turbulent flow and therefore efficient heat transfer from the metal to
the cooling medium. A small number like 1,000 means you have laminar
flow for sure, whereas a larger number such as 10,000 or larger means
you have turbulent flow in the cooling passages. In between is
transitional. I wish the powers that be had deemed a basic education
in fluid mechanics to be of more importance to us EEs.


Best regards,
Spehro Pefhany

 

[John Larkin]

http://www.discovercircuits.com/H-Corner/1ampcurrent.htm

Probably not the most accurate thing, but may be helpful for smaller
shunt measurements.

We have a similar 1-amp home-made box. We used a good voltage
reference, a 4-lead Vishay power resistor as the shunt, and a 10-turn
pot to trim current. Once it warms up it is short-term stable to a few
PPM. It has a load-short switch to keep it warm. We'll typically drive
a precision 1-ohm oil-filled resistor in series with an unknown,
measure the drops, and compute the unknown.

John