how to test PowerMOS Rds on?

[pianomania]

rdson measurement

all we know Rds-on is to measure the power mos turn-on effective resistance on saturation region, it is typical 0.1ohm around.

but include measurement instrument internal resistance, wire resistance , and bonding wire all contribute a certain amount R.

how to eliminate these influence ?

or how to alleviate the effect ?

 

[jluhzw]

rdson

Hello, as far as I know, the power mosfet works in the triode region ,not the saturature region. Yes , we can acquire rdson from the spec which contains the
bonding wire resistance.so when designing the power mosfet size, we need know
how much the bonding wire resistance is. For another aspect, we should consider the Power dissipation in the Power MOSFET, and calculate how many rising temperture from the ambient temperature.
Maybe you can know the inoformation about the bonding wire from the chip's package information. So you can estimate
the length of the bonding wire.Also you can know the maximum load current,
you can estimate bonding wire's diameter. For large load current, it maybe more
than one bonding wire for one PIN to PADS.Sometimes, it could not be known how many bonding wire for the PIN, we can seek help to the X-Ray equipments.
Now you know the bonding wires' length and diameter, you can estimate the resistance of the bonding wire.

Added after 12 minutes:

PS: Generally,for the bonding wire diameter 1mil , the maximum current is 1A
The following is how to calculate the resistance of bonding wire.

 

[Tryon]

rds on

May be bond wire resistance we can control it but how about the measurement instrument internal resistance, wire resistance ? That is so complicated.

 

[haff99]

bond wire resistance

There is no way to eliminate these effects. If you have a very low RDSon, you have to consider the bond wire resistance, metal resistance etc. when sizing your device. Obviously the more bond wires you use the smaller the resistance. So you have to account for these effects when sizing your device. They can contribute a significant portion of your rdson if your rdson is very small.

 

[timof]

mos rdson calculation

There is no way to eliminate these effects. If you have a very low RDSon, you have to consider the bond wire resistance, metal resistance etc. when sizing your device. Obviously the more bond wires you use the smaller the resistance. So you have to account for these effects when sizing your device. They can contribute a significant portion of your rdson if your rdson is very small.

Parasitic resistance of the measurement setup (probes, wires,...) can be eliminated by using Kelvin measurement setup (separate Force and Sense pads for both source and drain).

Parasitics due to contacts, vias, and mainly metal interconnects, and wirebonds/balls are intrinsic components of large area power device Rdson, and should be taken into account. Optimal metal layouts are too complicated to allow simple calculation of their resistance - and they can contribute a lot to total Rdson. There are some software tools that can calculate Rdson from power device array layout (plus analyze current density distribution in metals/vias).

 

[rikie_rizza]

rdson test

Some times Power Mosfet wiring is not easy to look with X-ray because the wire is not gold but it uses Aluminum.

What you must know is that RDSON value is tested after the device is fully assembled in a package. So, the value that you see or know is the value from the wafer to the lead while the device is turned on.

If the value is 100mohm, then that include the lead and its plating, the wire bond (of coure if its more than 1 wire or lead you must measure them all at once), and the wafer property.

 

[timof]

rds(on) typical powermos

What you must know is that RDSON value is tested after the device is fully assembled in a package. So, the value that you see or know is the value from the wafer to the lead while the device is turned on.

Rdson can also be measured on wafer, before packaging - so Kelvin measurements may help to eliminate certain parasitic components of Rdson.

 

[Novichok]

I work at a semiconductor production plant for MOSFETS, and the way we do it is by Kelvin measurement (using a Tektronix curve tracer for lab measurements and a Tesec standalone for production) once the product is complete. Does not matter what type of tester you are using, while making a Kelvin measurement, you eliminate most of the resistance on the probes (just make sure Force and Sense are the closes to each other and to the body package of the device). RDSON can also be measured directly at the die before it is assembled but it is not practical, because you can not force very high levels of current in to the device (we have devices rated up to 200A) and even if you could, some packages are designed with multi wires to reduce resistance which would be hard to implement on a bare die.
I don't know if there is a better way to do it, but that is the way we do it and we get very close to the datasheet values.