This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

INA253: Feedback for INA253 and INA250 schematic and layouting

Part Number: INA253
Other Parts Discussed in Thread: INA250

Hello,

We selected INA253 and INA250 for current measurements in a HIL test rack application.

Can I ask you have a look at the schematic and layout below and give a feedback why the old setup did not work and if the new approach should work?

 

The old approach was like this:

Applying DC currents well in the appropriate ranges (10A) destroyed the sensors. I shorted a power supply directly with the INAs. 12/16 sensors are completely gone: sensor supply current is around 50mA instead of the max quiescent current of 2.4mA in the datasheet. The analog output stays at 0V instead of the reference 2.5V +/-current reading. 4 sensors still look ok in terms of supply current, but the output voltage is off by roughly a factor of two. Seems to be the case for INA253A2 and INA253A3. For example the INA253A2 outputs ~3V when I apply a current of 1A. We identified three areas where the Speedgoat schematic/layouting varies from the TI evaluation board that handles the currents well.

 

  1. NC connected to GND:

As a matter of fact, pin 8, 12, and 15 are connected to GND on the eval board.

 

  1. SH and IN connected to IS:

We layouted a big polygon (signal plane) that includes SH and IN as shown in blue. This is a mistake that might lead to a bigger measurement error, but also doesn’t explain the destruction.

 

  1. We used an external 2.5V voltage reference.

 

The new setup will look like this, the 0Ohm resistors will eventually be replaced by a regular wire :

 

Thanks in advance for your feedback. Let me know if you would like to know more details.

Regards

  • Hi,

    You’re correct that a big polygon to include SH and IN pins will cause large error, but doesn’t explain the total destruction.

    Have you looked at the 5V supply to make sure that it is stable and doesn’t have spikes larger than 6V?

    There is no problem with supplying the REF pins directly with 2.5V; Pin 8 and 12 have no internal connection and they don’t have to be shorted to ground.

    Regards, Guang

  • Hi Alen,

    oops, why do you connect 5V across the shunt? This will immediately destroy the shunt :-(

    Kai

  • Hi Kai,

    5V are not directly connected to the shunt. I was just too lazy to add another connector for the 5V supply and used the remaining pin.

    The 5V should be reasonably stable as we have a Traco (U1) onboard for supply regulation. 5V should not be above 5V + 50mV ripple.

     

    Thanks for the remarks about pin 8 and 12. Do you have an explanation about pin 15? I couldn’t find a more detailed description in the datasheet why this should be connected to GND.

    Do you have a rough estimation on what the reference max current could be?

  • Hi,

    Pin 15 is an input used during production test. It should be connected to ground. However leaving it floating will not cause catastrophic destruction as you observed.

    I have no doubt that the 5V supply is stable and within spec, however I would still observe its transient behavior(on and off), just to make sure it is as expected.

    The reference max current should be less than a couple of uA.

    Regards, Guang