I am interested in low resistance shunts (tens / hundreds of microohms range) used in high voltage / high current setting. Typically, those are busbar-mounted shunt resistors like this one:

enter image description here

The current is passing through the busbar and the shunt, while voltage drop is measured by the two small sense pins pointing upwards. However, some of such shunts are designed with four sense pins instead of two and I cannot understand why. Could someone please point out what could be the advantage of using 4 sense pins instead of 2? Could it be related to measurement accuracy, or perhaps rather with reliability, etc.?

To be clear: I am talking not about the typical 4 terminal shunt connection, but specifically about 4 sense pin connection.

enter image description here


4 Answers 4


By pure chance I found the source of your image (2 pins):


And another product from the same manufacturer (4 pins):


The main difference seems to be that the 4-pin version is described as having "fully redundant sense pins". So one would guess that this is done for safety — being able to diagnose a broken connection, when the two measurements don't agree with each other.

It's easy to imagine situations in which a failure of the current monitoring system could lead to fireworks.

  • \$\begingroup\$ Thanks. Yes, this is kind of obvious reason for safety critical applications. I am still interested if there is any advantage regarding the measurement accuracy, but it seems that there is not. \$\endgroup\$ Dec 30, 2021 at 9:28

I'm not an expert in this area and don't claim to know the answer. Nevertheless, my guesses are that four pins offer:

  • improved mechanical advantage over two pins in applications where there is high g-loading or high/sustained vibration or both,

  • attachments for two redundant current sense circuits in a high-reliability current sense application,

  • attachments for high range and low range current sensing using a single shunt (e.g., measuring the DC amps signal as well as the noise that's riding on the DC amps signal),

  • and as mentioned in tobalt's answer to your question, thermoelectric compensation.

  • 3
    \$\begingroup\$ I suspect mechanical reasons as well as the article linked in one comment below shows a PCB being mounted onto the bus bar sense pins. isabellenhuetteusa.com/wp-content/uploads/2021/03/… \$\endgroup\$
    – DamienD
    Dec 23, 2021 at 8:45
  • \$\begingroup\$ @DamienD, the URL you've provided seems incorrect; I get a "page not found" error when I try to load that URL. \$\endgroup\$ Dec 23, 2021 at 9:17
  • 1
    \$\begingroup\$ isabellenhuetteusa.com/wp-content/uploads/2021/03/… \$\endgroup\$
    – DamienD
    Dec 23, 2021 at 9:23
  • \$\begingroup\$ Sorry, but I cannot find tobalt's answer to this question. Could you please elaborate on how four sense pins might help to mitigate the thermoelectric effect? It is very important to my application and that's why I had this question in the first place. However, I do not see how four pins might help to mitigate it, but maybe I am missing something. \$\endgroup\$ Dec 30, 2021 at 9:25
  • \$\begingroup\$ @Dainius, as I recall from tobalt's answer (which has mysteriously disappeared), uneven heating within the shunt might cause unequal voltage potentials at the two pairs of sense terminals. Measuring separately the potential across each pair of sense terminals and averaging these two measurements could help compensate for this effect. Again, I'm not an expert, but I suspect the two voltage potential measurements could be made in parallel '=' (top left, top right; bottom left, bottom right) or crosswise 'X' (top left, bottom right; bottom left, top right) across the four shunt terminals. \$\endgroup\$ Jan 1, 2022 at 17:44

The two most likely reasons are

  1. mechanical
  2. redundancy

The mechanical aspect has been covered ( https://electronics.stackexchange.com/a/601504/16222) and it is plausible since under high currents there will be mechanical stress experienced by the connections, be it due to thermal expansion or due to mechanical movement as a result of the high magnetic fields

Redundancy however shouldn't be dismissed. From an FMEA perspective there maybe a need of "independent sensing" and while it can be argued from a failure perspective that a mechanical device (such as this shunt) has a 1e-12 failure rate, the same cannot be said for the acquisition and processing. A totally independent analogue path, sampling and reporting may be required for hazard mitigation once the fault tree of a higher system is considered


From this, (sorry, in french), a picture of relative distribution of current in bar type conductor.

One can see that current must be measured at some points only (For measuring the current, you need 2 sections along the bar) -> 4 points of measurements well placed, if one want a correct value.

NB : this appear when measuring alternating currents (50 or 60 Hz), but also when measuring transient currents.

This bar is very large, but the problem is the same, whatever dimension. A study with flux3D, for example, could confirm this.

enter image description here

As a consequence, it will be noted that the edges of the bar will be hotter than at the "center".

This will be particularly surprising when this is the case with transient currents.

After a few "seconds", the heat will of course be distributed throughout the bar which will be at an almost "uniform" temperature ... Fairly "fast" propagation of heat in a metal.

  • \$\begingroup\$ Where do you see 4 points? \$\endgroup\$
    – eckes
    Dec 24, 2021 at 3:01
  • \$\begingroup\$ This is one "transverse view" of a section. For measuring the current, you need another section along the bar .... \$\endgroup\$
    – Antonio51
    Dec 24, 2021 at 9:01
  • \$\begingroup\$ I don't know if this is The Answer, but it is certainly interesting! \$\endgroup\$
    – user57037
    Jan 1, 2022 at 17:59

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.