• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Subscribers 48 subscribers
  • Views 1051 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

Selecting Current Sense Amplifier for PSFB Secondary Side Load Current Sensing

JSly
Intellectual 410 points

Other Parts Discussed in Thread: INA180, AMC1300, AMC3301, AMC1302, INA239, INA226, ISO7842, INA260, INA240

Hello,

I'm working on designing a 3 kW PSFB and could use some help with selecting the current sensor + amplifier for the secondary side load current sensing. 

A few relevant specs about the system - 

  • 3 kW output (3000W/12V = 250A output current)
  • 100 kHz switching frequency
  • MCU control 

Based on this white paper, it seems like a shunt would be better in terms of accuracy, max temp, and bandwidth capability. I'm working on selecting a current sense shunt amplifier and was wondering if you have a recommended part here. I'm not sure what the bandwidth requirement is here, but I imagine it may be a function of switching frequency? If you could help with that as well, that'd be great.

I was looking at the INA180 datasheet and had trouble identifying what the maximum input voltage is without clipping. Where can I find this information?

Thanks!

J

Thanks!  

  • 0
    TI__Mastermind 26415 points

    J,

    Thanks for using E2E! A couple of questions here.

    Where in the power chain are you planning to place the shunt? Will the shunt see any pulse width modulation from the full bridge? Also, what is the planned common mode across the shunt, i.e, are you planning to implement on the high side or the low side, and what DC bias will be present at the shunt input pins?

    Off the cuff, I can recommend a few solutions to have a look at. The first would be a fully isolated amplifier like the new AMC3301 (you could also look at AMC1300/AMC1302 if the power components of AMC3301 are unappealing).

    Alternatively, you could look at one of our non-isolated solutions, such as the new INA239, or its predecessor, INA226, and isolate using a digital isolator such as ISO7842. We have an example of this using INA260 here. Unfortunately the INA260 is not a good solution here, as the package can only withstand up to 15A of continuous current. For this level of current, you're going to have to use an external shunt for thermal reasons. 

  • 0 in reply to Carolus Andrews
    Intellectual 410 points

    Hi Carolus,

    Thanks for the quick response. The shunt goes on the output of the secondary side in the PSFB converter. I haven't decided yet whether it will go on the high-side or low-side. Common mode will only be 12V.

    Why would a isolated solution be required? This is on the secondary side so its already galvanically isolated from the high voltage side.

    Best,

    J

  • 0 in reply to JSly
    TI__Mastermind 26415 points

    J,

    I was getting ahead of myself here. You're correct that you may not immediately need an isolated solution. The reason I reached for isolated solutions is that typically your feedback must also cross back over the isolation barrier as well. Which side of the isolation does your MCU sit on? If you're feeding back to the MCU and it sits on the secondary side, you should be able to use a non-isolated solution, but I would think that somewhere in the signal chain feeding your duty cycle back to Q1-Q4 you would need to move the signal back across the isolation barrier, unless I'm visualizing your setup incorrectly. 

    If you're measuring at the output of the secondary side, then you should have a relatively smooth current, although you will have the voltage ripple present on the line (although I'm assuming at 12V, ΔVo is relatively small). For a non-isolated approach, have a look at the INA240. It has a 400kHz BW, so its 4x your switching frequency, and includes our proprietary PWM rejection, which will help against any switching artefacts there might be present on the line.