4 wire measurement of a 0402 chip resistor

[DeepLink]

I need to measure the resistance of a 0402 chip resistor - before and after reflow soldering
The resistor will not be soldered to a PCB, just go through the oven to be heated

I need it done by a 4 wire measurement, as the resistance is low and I need the accuracy

My kelvin clips are to big for a reliably connection on such a small component

Are there any fixtures that can be used for this task?

[Psi]

Do you want like a ZIF socket for a 0402 part with 4wire measurement ?

I suspect any difference you detect caused by the oven process will be smaller than the difference you get just having it mounted vs not mounted on a PCB, ignoring the heating involved in doing that.

[wraper]

Make PCB test pads for 4 wire measurement.

[PartialDischarge]

Make PCB test pads for 4 wire measurement.

Exactly, in the final application the resistor will be soldered so best to mimic the same conditions

[DeepLink]

Do you want like a ZIF socket for a 0402 part with 4wire measurement ?

I suspect any difference you detect caused by the oven process will be smaller than the difference you get just having it mounted vs not mounted on a PCB, ignoring the heating involved in doing that.

Yes, a "ZIF" socket would fit my needs, but I can't find any
It is not possible to solder the resistor to a PCB to measure, as I need to measure before soldering

The resistor is a 49.9 Ohm 0.1% TC10
According to the datasheet "Resistance to Soldering Heat - ΔR±0.2%"

In production we are seeing some PCB's where the resistors are outside these tolerances, so we need to figure out weather it is the initial resistance or the reflow process that are giving us problems

We have the temperature under control and the equipment to measure correctly, but simply need a fixture

[mendip_discovery]

Make smaller kelvin clips.

Or buy something like this,
https://uk.rs-online.com/web/p/lcr-meter-accessories/1235980
It is for an LCR meter but should also do the job neatly for you.

I have seen some of Ali for less and from Rohde for a lot more.

[coppice]

If your resistor is a low enough value to worry about the effects of the copper traces try looking at https://www.analog.com/en/analog-dialogue/articles/optimize-high-current-sensing-accuracy.html

[Nanitamuscen]

It is possible to make a 4-wire seat on the printed circuit board, covered with gold, on which the investigated resistor is easily pressed from above.

[mawyatt]

Might be possible with something like this.

https://www.eevblog.com/forum/projects/different-type-lcr-smd-fixture/

Altho we haven't measured any 0402 components, works OK for larger SMD devices.

Best,

[Psi]

Maybe you could modify this 0402 flip top capacitor aging test jig to be a 4 wire resistor tester by linking one of the other 0402 channels together at the right spot to be the measurement circuit.

https://www.aliexpress.com/item/4000129421329.html

Or maybe message that seller and ask if he has any 0402 4-wire measurement jigs. He look to be in that industry of 0402 jigs.  Only 85.7% positive feedback though, so YMMV

[PartialDischarge]

In production we are seeing some PCB's where the resistors are outside these tolerances, so we need to figure out weather it is the initial resistance or the reflow process that are giving us problems

This is actually an interesting problem that probably few have taken a look at. Maybe even the small size of 0402 makes them more prone to variations compared to 0603 or 0805.
Let us know what you find.

[guenthert]

[..]
According to the datasheet "Resistance to Soldering Heat - ΔR±0.2%"
[..]

What does that mean?  Is here with 'R' truly the electrical resistance meant?

[DeepLink]

Make smaller kelvin clips.

Or buy something like this,
https://uk.rs-online.com/web/p/lcr-meter-accessories/1235980
It is for an LCR meter but should also do the job neatly for you.

I have seen some of Ali for less and from Rohde for a lot more.

We do have such a fixture for our LCR meter, that is a 2 wire connection to the DUT

[DeepLink]

[..]
According to the datasheet "Resistance to Soldering Heat - ΔR±0.2%"
[..]

What does that mean?  Is here with 'R' truly the electrical resistance meant?

This simply means that the resistance can shift +/- 0.2% from its initial value, after being reflowed

[Psi]

Could you make a 4-wire test PCB that you solder the part to before hand using ultra low melt solder, like 118°C stuff.
Then expose to full reflow temp and retest after.

[JodelJonny]

Wouldn't all kinds of clamping pressure affect the resistance as well?

Mercury would qualify as ultra low melting point solder, wouldn't it?

[Psi]

Wouldn't all kinds of clamping pressure affect the resistance as well?

Mercury would qualify as ultra low melting point solder, wouldn't it?

Gallium maybe, i think you're pushing it with mercury. 

[Kleinstein]

Getting a proper 2 wire contact to the small pads is tricky. This applies also to the PCB the resistors are used in. The solder and exact position on the PCB can effect where exactly the current flows or voltage is taken from. The same applies to the test jigs , the one at the manufacturer to trim / test the parts and the 2nd jig for the test before soldering.

The idea with essentially all solders is getting some alloying to the material. So it can be undone fully. Mercury and also gallium are quite keen to react with other metals.

[Conrad Hoffman]

I'd just make the connections very close to the probe tip and not worry about trying to get separate connections right on the resistor.

[mawyatt]

An 0402 will be a challenge for 4 wire resistance measurements, and maybe the best approach is to use the SMD fixture for LCR meters.

Best,

[jbb]

It is possible to make a 4-wire seat on the printed circuit board, covered with gold, on which the investigated resistor is easily pressed from above.

Wouldn't all kinds of clamping pressure affect the resistance as well?
…

Hmm, sounds like some kind of squishy conductor might help. Maybe some Z-axis tape would help? These have little metal conductors and conduct through the tape (Z direction). But not across the tape (XY directions).

Possible option: https://www.adafruit.com/product/1656

[coromonadalix]

https://www.ebay.co.uk/itm/Four-Wire-Kelvin-Test-Tweezers-with-4x-BNC-for-LCR-Meters-Component-Testers-/264074739210

[mawyatt]

https://www.ebay.co.uk/itm/Four-Wire-Kelvin-Test-Tweezers-with-4x-BNC-for-LCR-Meters-Component-Testers-/264074739210

We got an equivalent of these long ago, they tended to be very difficult to use. Later acquired the Tonghui TH26009B style, these were much easier to use, but more expensive.

Best

[mawyatt]

Wouldn't all kinds of clamping pressure affect the resistance as well?

That's a good question and one we had awhile back wrt to measuring SMD very low impedance values. The LCR fixtures and tweezers we have all suffer form the variations in applied pressure, including the "Zero" ohm calibration.

When we set out to develop the "Plunger" type SMD fixture shown (see above link), one of the goals was to achieve better repeatability than what we had seen with the LCR SMD fixture and tweezers (even the better Tonghui ones mentioned) with low Z DUTs.

This is still an ongoing developement, but highly encouraging results have been achieved. We've been able to measure quality low impedance (milliohm levels) with very good repeatability, even into the sub-milliohm regions. The main reason for this level of repeatability is that the Force and Sense leads from the LCR meter aren't terminated at the fixture connection point like conventional fixtures, but the connection between Force and Sense is made with the DUT SMD end terminations. So in effect the DUT itself is making the connection between the Force and Sense which dramatically reduces the contact impedance, pressure, surface, and orientation sensitivity of the fixture to the DUT.

Still need to perform more work and get gold plated PCBs, will provide more details if folks are interested.

Best,

[DeepLink]

https://www.ebay.co.uk/itm/Four-Wire-Kelvin-Test-Tweezers-with-4x-BNC-for-LCR-Meters-Component-Testers-/264074739210

Unfortunately, this is still a 2 wire connection to the DUT. We do have such a tweezer, but readings are quite unstable and very dependent on applied pressure