Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

What is Kelvin Connection and its application in IC Design & Layout?

tomsawyer3.14

Newbie
Joined
May 9, 2021
Messages
1
Helped
0
Reputation
0
Reaction score
0
Trophy points
1
Activity points
17
Hi,

I am doing IC layout and I saw on one schematic some comments like "Kelvin Connection" & "Kelvin Ground.

Is this the same with the Kelvin Connection on Instrumentation where you measure a resistor by passing current then measuring the voltage across?

I don't get its application/imporatance in the IC/ micro level.

Can you guys help me understand it?

Also hat are the layout consideration when there is a Kelvin Connection on the schematic?

Regards,
Tom
 

KlausST

Super Moderator
Staff member
Joined
Apr 17, 2014
Messages
19,715
Helped
4,341
Reputation
8,691
Reaction score
4,303
Trophy points
1,393
Activity points
130,510
Hi,

There are so many discussions and informations about it. Thus just in short:
It does not matter if wiring, PCB layout, or IC design....always the same.
You simply split one one wire into two:
* one carries the current (it has voltage drop)
* while the other senses the voltage at the point of interest..with about zero current thus avoiding errors caused by voltage drop.

For more informations please do an internet search.

Klaus
 

danadakk

Advanced Member level 3
Joined
Mar 26, 2018
Messages
998
Helped
194
Reputation
392
Reaction score
206
Trophy points
43
Activity points
4,679
This might help, attached. Keithley 7'th edition low level measurement handbook.


Regards, Dana.
 

Attachments

  • LowLevelHandbook_7Ed.pdf
    3.7 MB · Views: 5

dick_freebird

Advanced Member level 5
Joined
Mar 4, 2008
Messages
7,285
Helped
2,127
Reputation
4,259
Reaction score
1,983
Trophy points
1,393
Location
USA
Activity points
58,436
Products such as voltage references / regulators
will see degraded load rejection if the output
"force" (pass transistor) and "sense" (feedback)
are made to share a common bond wire and
package lead.

Designing it so the feedback comes back
from the pad, post or PCB by another route
lets the error amplifier "kill" the ohmic I*R error
by a bit, a bit more, altogether.

This also applies at die probe where probe tip
and contact resistance can be high and highly
variable.
 

crutschow

Advanced Member level 5
Joined
Feb 22, 2012
Messages
3,858
Helped
920
Reputation
1,838
Reaction score
907
Trophy points
1,393
Location
L.A. USA Zulu -8
Activity points
21,795
Products such as voltage references / regulators
will see degraded load rejection if the output
"force" (pass transistor) and "sense" (feedback)
are made to share a common bond wire and
package lead.
Say you have a voltage regulator or power supply and you want a stable voltage at the load independent of the load current.
For that you would have a separate set of wires from the regulator, besides the wires carrying the current, to measure the voltage directly at the load, so any resistive voltage drop to the load can be cancelled (ignored).
That is a 4-wire Kelvin connection.

You would do the same thing when designing a regulator at the chip level.
The IC traces that carry the load current to the output power and ground pins would be separate from the IC traces carrying the output voltage to the sense circuitry.
 

timof

Advanced Member level 2
Joined
Feb 21, 2008
Messages
652
Helped
213
Reputation
426
Reaction score
209
Trophy points
1,323
Location
San Jose, CA, USA
Activity points
6,907
The basic idea behind Kelvin (or four terminal) sensing is very simple, and is explained here (or in the posts above):


The devil is in the detail, though.

When you have IC design, for example - power transistor, you want to measure the current or voltage on it.
Now, power FET usually uses multiple metal layers (from 1-2-3 in old technologies, to ~20 in latest technologies), and multiple "fingers".
Where you "connect" in this massive metal network, is very important.
The sense point should be selected so that it provides kind of an average of the whole device.
The information about where this average is, can be obtained from proper simulation.

Such methodology for selecting an optimal sense point - where you connect sense or replica (small) device - to measure the current in the large device - is very important for sensing accuracy.
 

dick_freebird

Advanced Member level 5
Joined
Mar 4, 2008
Messages
7,285
Helped
2,127
Reputation
4,259
Reaction score
1,983
Trophy points
1,393
Location
USA
Activity points
58,436
If you have a very cellular "big transistor" layout it can
be effective to distribute sense "cells" throughout.
Provided you have the high-enough metal stack to
do it.

But sometimes the real quantity of interest is some
external voltage, which needs to be brought back on
chip without accruing error. And you may not be
allowed the extra pin to do that cleanly, so have to
do the best you can with pads and wires and routing
(what you can control, as a designer).
 

LaTeX Commands Quick-Menu:

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Top