Why output voltages of precision shunts are so high?

[zlymex]

The output voltage of normal shunts are 50mV, 60mV and 75mV. I thought they may higher at 200mV for precision shunt, but, when comes to precision 100A shunts, why such high voltage of 800mV for Fluke A40B?
http://us.flukecal.com/products/electrical-calibration/electrical-standards/a40b-series-precision-dc-and-ac-current-shunts

Tinsley 4638 @1.0V (100A, 0.01 Ohm), Guildline 9230/100 @1.0V (100A, 0.01 Ohm), even higher.
Transmille 100A @0.7V, Guildline 7340 100A @0.4V.
Ohm Labs on the other hand, produce their 100A shunt at 0.1V only
http://www.ohm-labs.com/precision-shunts/cs-100-300.html

Is that because of the thermal EMF? In theory, reducing the voltage by half will reduce the power and temperature rise also by half, we get the same relative thermal EMF.
Is that because of the thermal EMF at the measurement side? 1 Ohm standard resistor is tested at 0.1V and still
Is that because of the noise or AC measurement considerations?

[TiN]

Perhaps old voltmeters were not good at low ranges, so shunts designed to match them were targeted for higher voltage to fit better accuracy?

[manganin]

The primary level DC shunts up to 100A have usually 0.1V or 1V output at the nominal current.

Before the long scale multimeters the 1V output was easier to measure (and compare with a standard cell) with the existing equipment. And the 0.1V output is also too low for most thermal converters.

The downside of the 1V output is of course the large amount of heat produced. That is why I prefer my Sullivan 0.001ohm 100A shunt instead of the Leeds & Northrup 0.01ohm 100A.

[EEVblog]

Greater signal to noise ratio, less worry about offset voltage in any amp etc. The higher the better, except for the circuit under test.

[zlymex]

I guess that too for noise and measurement reason in the old days. However, Fluke A40B-100A is a relatively news device(old model A40A). And the completely new Transmille's 100A, only come out last year?  It is 0.7V full scale.
http://www.transmille.com/bench-instruments/precision-references-and-standards/ac-dc-shunts

Leeds & Northrup also had a 100A version(0.001 Ohm, type 4223).

[ManateeMafia]

Fluke may have chosen their shunts to match their transfer standard's best range on the 792A? Their rms sensor is in the 2V range. I haven't looked at the 5790A but it is probably the same.

https://www.atecorp.com/ATECorp/media/pdfs/data-sheets/FLUKE-792A_Datasheet.pdf

[manganin]

However, Fluke A40B-100A is a relatively news device (old model A40A). And the completely new Transmille's 100A, only come out last year?  It is 0.7V full scale.

The main use of the shunts mentioned is in AC/DC transfer,  connected to a thermal converter.

Leeds & Northrup also had a 100A version (0.001 Ohm, type 4223).

Because of the small size the 4223 should be considered as a standard resistor rather than a high current shunt. I wouldn't use it for currents higher than 10A. It withstands a 100A current (limited accurary), but only in cooled and effectively circulated oil bath, which is quite impratical.

[acbern]

I think one of the main reasons is that in order to obtain high precision, 0.01ohms is much more convenient than 0.001ohm, so at 100A you end up with 1V if you want to stick to 10mR.

[coppice]

I think they carefully set the voltage drop for most shunts just high enough to ensure the thing is utterly useless for most applications. 

Compared to using a 1millohm shunt, using a 10milliohm shunt certainly makes it easier to get the resistance precise. However, with 10 times the power dissipation you either need to make the shunt have excellent heat dissipation or have a really good temperature coefficient.

[tszaboo]

The temperature coefficient of low value shunts increase, the lower you go in resistance.
Take a look at this datasheet:
http://www.vishaypg.com/docs/63216/VCS1625P.pdf
It is an excellent current shunt, I've used this myself, shipping thousand of product in spec. The values you are looking for are in Table 1, maximum TCR (typical is marketing bllsht). For the resistor above 2 Ohm, it is +/-5ppm, for below 0R03 it is +/-50ppm. And this is with the same technology and everything, only the laser/etching will make a different resistive element.
So with the same technology, while you can create a R resistor x good, R/10 is going to be much worse. And if you are able to handle the heat, by oversizing and cooling, you should. Also, load life stability is very very bad for low value shunts. For example there are Isabellenhutte shunts, which you can order with 0.1% initial accuracy, but just after 2000 hours under load, they can drift 0.2%. Which means you either calibrate very often or you go higher in the resistance.
At 100A, dissipating that 40W is not really an issue.

[zlymex]

The temperature coefficient of low value shunts increase, the lower you go in resistance.
Take a look at this datasheet:
http://www.vishaypg.com/docs/63216/VCS1625P.pdf
It is an excellent current shunt, I've used this myself, shipping thousand of product in spec. The values you are looking for are in Table 1, maximum TCR (typical is marketing bllsht). For the resistor above 2 Ohm, it is +/-5ppm, for below 0R03 it is +/-50ppm. And this is with the same technology and everything, only the laser/etching will make a different resistive element.
So with the same technology, while you can create a R resistor x good, R/10 is going to be much worse. And if you are able to handle the heat, by oversizing and cooling, you should. Also, load life stability is very very bad for low value shunts. For example there are Isabellenhutte shunts, which you can order with 0.1% initial accuracy, but just after 2000 hours under load, they can drift 0.2%. Which means you either calibrate very often or you go higher in the resistance.
At 100A, dissipating that 40W is not really an issue.

It's very convincing.
So, high voltage output of a 100A shunt will avoid low value resistors thus avoid bad tempco and poor load life.
I have a source saying that a Fluke A40B-100A is made of 960 pieces of foil resistors(like S102C and S102K) in parallel, value is 7.68 Ohm each. That might be the smallest value for foil elements Fluke think they can use without affecting the performance much.

Further more, why bad performance for small value foil resistors? Is that because the foil is too thick that cannot be compensated by the ceramic substrate effectively? Or thick foil has the tendency of peeling off from the substrate?

[coppice]

The temperature coefficient of low value shunts increase, the lower you go in resistance.
Take a look at this datasheet:
http://www.vishaypg.com/docs/63216/VCS1625P.pdf
It is an excellent current shunt, I've used this myself, shipping thousand of product in spec. The values you are looking for are in Table 1, maximum TCR (typical is marketing bllsht). For the resistor above 2 Ohm, it is +/-5ppm, for below 0R03 it is +/-50ppm. And this is with the same technology and everything, only the laser/etching will make a different resistive element.
So with the same technology, while you can create a R resistor x good, R/10 is going to be much worse. And if you are able to handle the heat, by oversizing and cooling, you should. Also, load life stability is very very bad for low value shunts. For example there are Isabellenhutte shunts, which you can order with 0.1% initial accuracy, but just after 2000 hours under load, they can drift 0.2%. Which means you either calibrate very often or you go higher in the resistance.
At 100A, dissipating that 40W is not really an issue.

The Vishay shunt you referenced is only rated for 5A. If you want a shunt rated for 100A it is physically much larger, and the contraints on achieving good temp coeff and ageing figures are quite different. Voltage drop is a big problem with current measurements when the voltages are low. Dropping 800mV in a 200V system might not affect it much. An 800mV drop inserted in a 5V system might be a little problematic.

[Kleinstein]

There are two current ratings: one on how much current the shunt can stand without getting damaged and one how much much current can be measured without too much error due to self heating or magnetic effects.  The second rating is usually much lower - which value is best also depends on the voltmeter used.  So you may have a shunt that is rated for up 100 A with 0.8 V drop, but it's performance may not be good at more than 20 A.  So if you have to measure a 80 A shunt, one might prefer to used the one rated for 1000 A instead of the 100 A one.

[zlymex]

There are two current ratings: one on how much current the shunt can stand without getting damaged and one how much much current can be measured without too much error due to self heating or magnetic effects.  The second rating is usually much lower - which value is best also depends on the voltmeter used.  So you may have a shunt that is rated for up 100 A with 0.8 V drop, but it's performance may not be good at more than 20 A.  So if you have to measure a 80 A shunt, one might prefer to used the one rated for 1000 A instead of the 100 A one.

That's true for normal shunts. However the topic here is precision shunts which normally designed and specified at full range.

[tszaboo]

The Vishay shunt you referenced is only rated for 5A. If you want a shunt rated for 100A it is physically much larger, and the contraints on achieving good temp coeff and ageing figures are quite different. Voltage drop is a big problem with current measurements when the voltages are low. Dropping 800mV in a 200V system might not affect it much. An 800mV drop inserted in a 5V system might be a little problematic.

I give it as an example. VishayPG (or whatever they call themselves this week in this country) bothers making a proper datasheet. The example is showing it generally for one technology, which is pretty much applicable to most technologies.

The voltage drop is not a problem in your 5V system. If you want to measure 100A accurately, that is your problem, not increasing the 5V to 5.8V. You use system components based on the most difficult one. Eg. it doesnt matter if you want a shunt with 10mV drop at 100A  with 5ppm tempco if they dont make it. Just replace everything else till it works, power supply with sense lines is too easy.

[TiN]

From pictures which I saw on the web, high-value Fluke A40B shunts are array of large amount of VPG foil resistors on PCBs, likely in series-parallel.
So it's not always that shunt consists from single element for precision stuff.