ACS712T Max current

Hello Friends!

I need to monitor current in my solar system but i'm in doubt how to choose proper current sensor. I found ACS712T 30 A sensor at many places to buy but i'm not sure that it can handle 20 amps at 24v constant load? it is marked in datasheet as 30 amp sensor but when i look at the PCB those leads are too thin for such a strong current.. So i ask about your opinion, what is the maximum current that i can pass through this sensor? somebody sure knows the answer...

Regards!

The PCB has nothing to do with the chip. The chip - assuming you get the right dash number is rated to measure 30 A, and has a 1.2 mOhm "internal conductor resistance", so at 30 A it will dissipate 1W. The junction to ambient on the Allegro board is 23 C/W, so if you are at 25 C, the chip will get to about 50 C, a bit warm, but fine.

Not all boards have the 30 A chip, the sparkfun boards use the 5 A chip.

Ok, so i can use it for measuring current up to 20A at 24V? when you say 25C you mean ambient temperature? because at summer at that place where sensor will be temp can get up to 40+C.

Regards!

This is all spelled out in the data sheet, but TJmax is 165 C.

so the 40 C ambient + the 25 C rise puts you at 65 C, well below the max. At 20 A, you will run cooler. Just make sure you have the same heatsinking as spelled out in the reference design.

KeithRB: This is all spelled out in the data sheet, but TJmax is 165 C.

so the 40 C ambient + the 25 C rise puts you at 65 C, well below the max. At 20 A, you will run cooler. Just make sure you have the same heatsinking as spelled out in the reference design.

OK, thank You for the quick answer!

The chip is not connected directly to your source, there are inductive tracks to measure the current flowing, since it’s more like a ct not a shunt, heat disspation is not an issue.

The magnetic coupling is inside the package, the full current flows through the leads.

Inside, outside...

It's still magnetically coupled like a ct is.... but acs do have an interesting video about the ic.

KeithRB:
This is all spelled out in the data sheet, but TJmax is 165 C.

so the 40 C ambient + the 25 C rise puts you at 65 C, well below the max. At 20 A, you will run cooler. Just make sure you have the same heatsinking as spelled out in the reference design.

Note the case temprature is not the same a the junction temprature, you have to take into consideration the thermal resistance between junction and case to calculate the junction temprature at any given case temprature.

Likewise case to ambient also has a thermal gradient.

A temprature of 25C in the data sheet means case temprature not ambient. This will only be the same thing if you have an infinite heat sink.

In the post up higher I note that the Theta Junction to ambient on the Allegro test board is 23 deg C per Watt. I know how to do thermal calculations.

I know how to do thermal calculations.

In case the OP doesn't.

http://www.thebox.myzen.co.uk/Tutorial/Power.html

http://www.thebox.myzen.co.uk/Tutorial/Power_Examples.html

Then there is my M/A-com app note about thermal calculations for RF power transistors:
“How to avoid getting burned by junction temperature”

(Which I can no longer find because of the brain dead website.)

I found ACS712T 30 A sensor at many places to buy but i'm not sure that it can handle 20 amps at 24v constant load?

It sure is a popular sensor with many PCB variations available. Disregarding the capabilities of the IC , the only PCB I would trust to handle continuous current at 20A is the evaluation board, which uses 4-oz copper and nickel plated binding posts.

The terminals used on other boards can be of low quality. It would be good to get the termination part#s and PCB manufacturing details ... I've seen a few burnished terminals like these with other products used in applications @ 20A.

ACS712-Demo-Board-Guide

cjdelphi: The chip is not connected directly to your source, there are inductive tracks to measure the current flowing, since it's more like a ct not a shunt, heat disspation is not an issue.

Heat dissipation is always an issue with high currents as thermal losses go up as the square of the current... 1.2 milliohms in a SOIC8 package is struggling at 20A without good cooling.

Solder the thick wires directly to the PCB traces on top of the package pins, then they serve as heatsinking as well as preventing damage to the PCB). For 20A you'll have something like 4 mm^2 of copper or more I'd guess, that's going to conduct quite a lot of heat.

For larger currents the ACS756 and ACS758 are the way to go, massive copper tabs on the high current circuit.