Hi, I am newbie to kicad. While trying to realize my PCB, according to the datasheet, it’s recommended to use a kelvin connection shunt resistor for more precise measurement. The demonstrated footprint used in the datasheet is presented below :
I wanted to know if the shunt resistor presented in the PCB is the same ? is it a 4 wire shunt resistor ? I mean if I use like the shunt resistor shown below, will I have 4 pads or two pads? I can’t understand how the PCB works.
With the picture shown, you only have 2 pads, but resistors with 4 actual pads are becoming obsolete, as it turns out that the thick round solder blobs on the copper pads form pretty stable low ohm connections in themselves. I have even seen SMT resistors with 4 actual terminals, and then soldered on two pads just like shown here.
For KiCad, you have at least two options. The simplest is to combine a standard resistor footprint of the right size, and then add two net-ties in the schematic for the 4-wire / Kelvin feedback points. Another option is to modify the footprints to have 4 pads and integrate the net ties inside the footprint itself.
Actually, I understand the difference between the two kinds of resistors, the 4 wires resistors provide much more precise measurement than the one with two wires. So, because of the measurement pads are separated, this will allow a denoised measurements when compared to the 2 pads shunt resistor! But the main problem here is that in the last figure, I can see 2 pads for measurement while other parts allow the current to flow but they are not separated! So, if there will be someone to help me make the difference in the PCB itself I would be grateful.
I have the problem with understanding what is your problem.
For example parts allow current to flow but are not separated. They (I don’t know about what parts you are saying - there are lot of them) allow current to flow because are not separated.
It looks that you expect to get current in separated parts while I think than if something is separated then there is no current in it. It is basic in your sentences that I don’t understand.
I am even not sure if you know that the 4 pin resistor has pairs of its pins shorted internally. But you say you understand how 4 pin shunt resistor works so you should know.
If we assume (I have never used 4 pin resistors and don’t suppose to use one ever) that soldering has negligible resistance compared to resistor pins then there is small or no difference if pairs of pads are connected internally or at the PCB pad.
I suppose that you name these 2 small pads as being separated and it is the source of me not understanding you. When you have no shunt resistor assembled they are separated, but who cares how PCB behaves without elements.
When you assemble the resistor they are connected. Look at your photo - resistor has two ends, but each of them is made in a form of two fingers (one thin and one thick). So in real circuit (and when we think about how it works we think about real circuit and not empty PCB) these pads are connected and not separated.
But, it looks that you still call them being separated.
If you use shunt resistor intended to be used with this connection made in a form of single pad (like in initial post of this thread) then you have these measurement tracks being connected at PCB both: before shunt resistor is assembled and after it is assembled. But as we don’t care how circuit behaves without elements these two situation (4 pin resistor and 2 pin resistor) are the same - these measurement tracks are connected, not separated.
If the shunt resistor would be something that has its own supply, and according to measured current flowing through thick pads then it issues the right voltage at galvanic isolated output thin pads then we could speak about them being separated. But you said that you know how shunt resistor works so it should be clear that it is not that case.
Yeah I know that, But how does this work on PCBs? I mean there’s two pads allowing the current to flow form + to - and the two pads for current sensing! I’ve found that image of the datasheet below but I can’t figure out how to connect it since it seems to gave only two pads (for allowing the current flow from + to -). Any additional information about the image below and how to use it with my PCB?
Are you sure?
If you had followed the Kelvin connection link to wikipedia, you would have seen an exact example, combined with explanation. I really could not explain it better myself.
If you have a component with 4 pads, then you should have a symbol with 4 pads, and a footprint with 4 pads. I don’t know if the symbol for a shunt reistor with 4 pads exist, but it’s very easy to make your own symbol.
Then you can do your schematic and layout without any issue.
If you have a shunt with only 2 pads, then you connect it as a simple resistor, and you have to be careful in your layout to have the measurement track reach the pads with the proper angle.
Newer versions of kicad allow finer control of “net ties” so this “hack” isn’t needed but this method shows how such kelvin sense points can and should be done
and one of the gitlab issues where kelvin points were the use-case
Thanks, I really appreciate that. Now, I understand better how Kelvin connection works. I can now figure out that in the first two images of my first post, it’s a kelvin connection. But I still can’t understand if in the datasheet, it’s really a kelvin connection since the image on the right half which represents the internal sections of the shunt resistor having 4 pads, two for current sensing and the other ones for current flowing (high voltage). But the figure on the left, there’s no appearing pads! So that confusing! That was the question I want to ask from the beginning.
I still want to figure out if the shunt resistor presented below in the datasheet has a kelvin connection or not! The images below are confusing. So, before I but it, I need to make sure if it as a kelvin connection or not first!
