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.

INA181: IC overheating

Part Number: INA181
Other Parts Discussed in Thread: CC2640R2F, CC2640

Hello,

for a bidirectional current sensing application I'm using an INA181.
I'm sensing the current in- and outflow of a leadacid-battery, so approx. 12V. As a µC I use a CC2640R2F.
I applied 3.3V to VS, ~12V to IN+ and IN- and gave it as a reference the VDDR pin of the CC2640, which is ~1.7V, the output goes directly to an analog pin of the µC.
There are two problems occuring:

1. When my board is completely switched off, but the 12V on IN+ and IN- still applied, there is a Voltage on VDDS present, which I don't want. If switched off, the board is required to carry no voltage at all  (but the 12V) .

2. After a while the INA181 heats up (you can't touch it with the finger), and eventually it broke due to overheating.

So my question is, is there anything I'm missing out on? I sticked to the datasheet as close as possible, but the INA181 shows the stated behaviour. 
Maybe using the VDDR pin of the CC2640 as a reference was a mistake, would that explain it? Any suggestions on how to make it work on the prototype?

Any help is highly appreciated, thank you. Alex

  • Hey Alexander,

    Thanks for choosing TI. We will look into this, but could you answer a couple questions. What are you using to power the INA181/VDDS? Also, what supply voltage configuration are you putting the CC2640R2F in? Basically what configuration are you choosing below?

    Sincerely,

    Peter Iliya

    Current Sensing Applications

  • Hey Alexander,

    I have a few questions for you as well.

    1. What voltage are you seeing on VDDS when it is powered off?
    2. Are you also powering off the microcontroller?
    3. Is the INA181 heating up when everything but the common mode is powered on or off?
    4. Could you send us a photo of your setup?
  • Hi Peter,
    thanks for your reply.
    To power the INA181, I'm using a LDO-Regulator 3.3V from ST "LDK320AM33R".
    The chosen supply configuration is the first one from the left (the one with the 10µH inductor).

    Regards, Alex
  • Hi Mitch,

    thank you too for your reply.
    1.) I currently desoldered the INA181, so I can't recreate the voltage, but I remember it to be something around 1.6V-1.7V, maybe ~1.3V. I'm afraid to destroy my last INA181, so unfortunately I can't come up with a more precise answer at the moment.
    2.) Yes, the µC is powered off too. Only the 12V on IN+ and IN- are applied to the INA181. The rest of the board hardware is separated via a switch from the batterie voltage.
    3.) I don't exactly understand what you mean by "common mode". The INA181 heats up, when I switch on the rest of the boardhardware. At some point I most likely had a short circuit through the INA 181, because the LDO was heating up very much as well and at some point the thermal protection triggered (shutdown at 160°C).
    4.) What exactly do you refer to by setup? My desk?;)

    Regards Alex
  • Hello Alex,

    Thank you for the information. What I mean when I ask for a picture of your setup is a picture of how you hook up your board to the supplies and battery. You don't have to turn it on or have the INA181 populated to get this picture, just tell me which voltage each supply would be set at.

    Also, could you send me you're whole schematic? That would be very helpful.

    For your information, common mode is the voltage across in+ and in-. In this case, it is your 12V.
  • Hi Mitch,

    my board is hooked on a laboratory DC power supply with 12V. There's not yet a batterie involved (That would be P5/BB on the schematic). As a load there's only the µC + peripherals on the board.
    As for question 3.):

    When I power the µC with the CC2640 LaunchPad via the JTAG-plug and common-mode is turned off (DC-power supply off), it didn't heat up.

    It went like this:
    As I was programming, my board powered by the LaunchPad, everything was fine. Then I switched on the DC power supply ~12V and unplugged the LaunchPad and realized a voltage on VDDS, even though the µC was switched off (SW4). I followed this and it was coming from the common mode voltage. I then desoldered the INA181 to make sure it was the reason for the voltage and it was. I stopped then looking into this matter and didn't solder the INA181 back on.

    Later I was programming a second board of the same kind which had the INA181 still soldered on. I switched on the DC power supply and everything seemed to be okay until after (I have to lie) ~5-10min my board shut off. That was when the thermal protection of the 3.3V LDO (U1) was triggered. From then on, if I switched on the board with SW4 after 2-3seconds it would shut down again. I then realized how hot the INA181 and the LDO were, desoldered the INA181 and everything was functioning fine again and the INA181 most likely gone.

    Maybe this is of help to you. Thank you for looking into this, find the schematic attached. Regards Alex

  • Hello Alex,

    It looks like you are driving the output of your LDO (U1) while powering up with the JTAG. It's possible that this is where the initial damage is coming from. Could you try powering up a board normally that you have not connected to the JTAG and see if it also is having heating/failing issues?
  • Hi Mitch,

    thank you for the hint. I still have one board, that I didn't connect to the JTAG, but it's not finished yet. I will check this out the next days.
    Could you please explain further what your basic assumption is, how driving the LDO on Vout could destroy the INA181, but not having any effect on the e.g. CC2640 since both are on VDDS?

    Regards, Alex
  • Hey Alex

    It's hard to know exactly what the problem is without seeing the setup and board layout. It's also possible that the devices have ESD damage, or that they got overheated during assembly. What I recommend is testing the INA181 before you put in on the board to ensure that it is working properly. This way you can know if it is a problem on the board, or a problem with the device. It would also be a good idea to measure the resistance between the various pins of the INA181 before and after you power it up. If you need more devices to test, you can request samples from the TI product page at www.ti.com/.../samplebuy.
  • Hi Mitch,

    I will get some more devices and do the testing you suggested. When I have more data, I'll update this thread.

    I have one more question:
    - When only common mode is powered on (in my case 12V), and no voltage applied to VS (only IN+ and IN- carry the common mode voltage), is there any voltage measurable on the Pins "OUT, VS and REF" or should it be 0V? (Assuming the device and everything else is working properly)

    If not, are there other devices with the same functionality that are completely isolated (no voltage passing through from common mode)?

    So far thank you very much for your effort, you guys are really helpful and I appreciate that a lot!
    Regards Alex
  • Hey Alex,

    That is correct, when there is only voltage at IN+ and IN- then OUT, VS, and REF should all be 0V. I look forward to hearing how those tests go.

  • Hi Mitch,

    today I received the order of some more INA181 devices from TI-Store. I soldered a device on my PCB with no parts on it at all, but the INA181 and shunt. I then applied 12V to IN+ and IN- and measured the outputs. 

    VS = 1.468V
    OUT = 1.468V
    REF = 11.19V

    None of the pins have a physical connection to each other on the board in this setup.
    I then measured the Pin-to-Pin resistance.

    Soldered on the board:

    OUT - IN+/IN- --> ~500 kOhm
    REF - IN+/IN- --> ~500 kOhm
    VS - IN+/IN- --> ~9.6 MOhm

    The IC just itself NOT soldered on the board:

    OUT - IN+ --> ~500 kOhm
    OUT - IN- --> ~500 kOhm
    REF - IN+ --> ~500 kOhm
    REF - IN- --> ~500 kOhm
    VS - IN+ --> ~9 MOhm
    VS - IN- --> ~9 MOhm
    IN+ - IN- --> ~2.5 kOhm

    As you stated above the Pins VS, REF and OUT are supposed to show 0V, which is not the case right out of the box.
    The functional block diagramm in Figure 39 in Section 8.2 of the datasheet shows probably the resistance I measured.

    I'm confused, maybe this information tells you something.
    Regards, Alex

  • Hey Alex,

    I'll look over this information. Meanwhile, could you also try applying the 12V to IN+ and IN- on a new part not connected to your board and measure the voltage on VS, OUT, and REF? This way we will know for sure there is no short on your board.
  • Hi Mitch,

    I checked all the connections on the board and there was no short. But to be 100% sure I took a brandnew INA181 and applied 12V directly to the pins and measured the OUT-Pin. In order to achieve this I soldered small wires to the IC, which I then connected to my DC power supply and voltmeter.

    The result is exactly 1.468V on OUT at 12V on common mode. So it can be definately said that my board is not involved since the upper results could be reproduced without it. I hope that helps.

    Regards Alex
  • Hey Alex,

    Looking into it a little bit farther, it turns out that a small parasitic voltage on these pins is possible as explained in this forum thread:
    e2e.ti.com/.../1785831

    However, this still does not explain the overheating. You mentioned that the LDO was overheating as well, but does the LDO still overheat when the INA181 has been removed? Also, could you send me some waveforms of the voltage at each and every pin of the INA181 while it is in the operating condition that leads to increased temperature?
  • Hi Mitch,

    it's good to know, that the parasitic voltage is normal. I now approached the intended configuration step by step by soldering a INA181 to a fully functional board. I first applied 12V to common mode, which was fine, then I switchen on the whole board, that worked out as well. I then applied 3.3V from the Launchpad via JTAG, that was fine too.

    The measured voltage on OUT fits perfectly to the applied current to the shunt.
    So my guess is, that probably either the first IC that I used was broken, some ESD issues occured or I messed up something when soldering it.

    It's a bit unsatisfying, because I can't recreate the overheating and thus never know what actually happened. But on the other hand two of my boards now work as intended for some hours without any problems.
    I'd like to thank you for your help, that is much appreciated. And after all, it didn't work in the beginning and now it does and you were a great support on the way!
    Thank you!

    Regards Alex

  • Hey Alex,

    I'm glad that you were able to get it working!