• State Answer Suggested
  • Locked Locked
  • Replies 8 replies
  • Answers 1 answer
  • Subscribers 49 subscribers
  • Views 1247 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

INA301 excessive power if applying common mode voltage of 20V at 5 kHz

Silvio Ziegler
Prodigy 110 points
Other Parts Discussed in Thread: INA301, INA200

Hello

I want to measure current in half-bridge configuration using INA301 (sense resistor at switching node). If switching 20V at 5 kHz the INA301 already heats up to 70°C. First I thought that the Common-Mode requirement (GND-0.3V) may be violated but even after adding an offset to the switching node voltage (now switching from 8V to 20V) the device is still getting way to hot. At the same time the output voltage is ok even if the device runs at 70°C, so it seems fully functional.

The problem is that I want to switch at up to 50kHz at 30V and since power loss seems to be a function of frequency and voltage the device temp completely is out of range. Can somebody confirm that behaviour or on the other hand has successfully run such a configuration?

Regards

Silvio

  • 0
    TI__Mastermind 26445 points
    Hi Silvio,

    Please allow us some time to verify on the bench. A question I have in the mean time is if you have measured the slew rate of you signal at the INA301? If you have can you please let me know what it is.

    Regards,
    Javier
  • 0 in reply to Javier Contreras4
    Prodigy 110 points
    Rising and falling edges are in range 400V/us. There is no load connected which is why I get such a fast edge.
  • 0 in reply to Silvio Ziegler
    Prodigy 110 points
    Any news on that one? Meanwhile I tested with the INA200 and that works good. So is there a problem with the INA301?
  • 0 in reply to Silvio Ziegler
    TI__Mastermind 26445 points
    Hi Silvio,
    Sorry for the delay. The slew rate at the input is to high. You can add a filter to slow down the slew rate. For instance an input filter with 5ohm resitors and and a 0.1uF capacitor between the inputs.(see figure 39 in datasheet of INA301) The slew rate may not be the same as when you actually have a load connected and may not need the input filter.

    Regards,
    Javier
  • 0 in reply to Javier Contreras4
    Prodigy 110 points

    Hello Javier

    Can you state what slew rate is acceptable for the INA301? I did not find any information in the datasheet and on the other side the INA200 seems not to suffer from this limitation.

    Actually I think your proposed solution does not work since such a filter will only limit rise time between IN+ and IN- and not between IN+/- and GND. Since in my application the problem is rise time between inputs and GND this will not work. BTW rise time is still a problem even if load is connected.

    Regards

    Silvio

  • 0 in reply to Silvio Ziegler
    TI__Mastermind 26445 points

    Hi Silvio,

    An acceptable input slew rate by design should be 1V/us.  You can achieve this by placing the 5ohm resistor and capacitors to ground instead of between the input like I mistakenly mentioned.  The capacitor to ground should be placed on each input along with a 5 ohm resistor.

    The INA200 has a different input circuit that does not have this exact limitation on the input.

    Regards,

    Javier

  • 0 in reply to Javier Contreras4
    Prodigy 110 points
    Now that I spent quite some time to debug this problem I want to share my findings here:

    -The INA301 does not really like fast edges and is getting hot. Mitigations like using capacitors to slow down the switching edges is not really practical considering additional power loss just from the added capacitor itself (C*V^2*f --> e.g. 10nF*30V^2*50kHz = 450mW). So the INA301 is definitively no option for such application.
    - The INA200 does not get hot but the output gets disturbed considerable by the fast common mode voltage change. After all, this is not acceptable for applications that need certain precision as mine. Interestingly TI does not state CMRR as a function of frequency in the datasheet as many competitors do. I think that should be added to the datasheet. In addition, you sometimes find in datasheets scope pictures of output voltage vs common mode step changes that gives a good idea if that chip is suitable for such application.
    - Finally I was testing a lot of current monitors from different manufacturers and found one that really works.
  • 0 in reply to Silvio Ziegler
    TI__Mastermind 27290 points


    Hi Silvio,

    Can you share the other manufacturer's part number?


    We have a new device nearing RTM that is designed specifically for this type of application, with 4 gain options. I think you’ll find its performance superior. The samples are available, if you can provide me with your contact information/shipping address; I'll coordinate to have some free samples shipped to you. I can also provide a free EVM so that you can easily wire a device up in your application.


    My email address is g-zhou@ti.com; you can contact me directly through email if you prefer,

    Regards,

    Guang-Apps CS Products