• State Resolved
  • Locked Locked
  • Replies 13 replies
  • Subscribers 63 subscribers
  • Views 267 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.

LMG2100R044: Idle power consumption of LMG2100 vs LMG5200.

Leo A.
Intellectual 535 points
Part Number: LMG2100R044
Other Parts Discussed in Thread: LMG5200, , LMG2100EVM-078

Hello,

I have been using the LMG5200 for a few years and decided to move to the LMG2100 as it is newer.

I see that with the LMG2100 the idle current draw for the board is 0.12A at 56V.

With the LMG5200 devices, the idle current was 0.08A at 56V.

The power consumption is 2.25W higher with the LMG2100 device.

I would have thought power consumption would be lower as the LMG2100 should be more efficient. Am I looking at this wrong?

Thanks;

Leo

  • 0
    TI__Expert 5185 points

    Hey Leo,

    I see the LMG5200 figure in the datasheet. However for the LMG2100R044, the datasheet specifies 0.09A. Are these numbers you're referencing coming from the datasheet or are these measurements?

    Best,

    Kyle

  • 0 in reply to Kyle Wolf
    Intellectual 535 points

    These numbers are for the whole circuit board including the support circuitry. All the other circuitry on the board has remained the same except for the change from LMG5200 to LMG2100.

    The board is powered by a singe 56V power supply. The board has 2 of the GaN devices as it operates in full bridge mode. So I see it as each LMG2100 is drawing about 1.125W more then the LMG5200 did.

  • 0 in reply to Leo A.
    TI__Expert 5185 points

    Are the boards you are measuring the TI evaluation modules or boards that were designed separately? 

    Best,

    Kyle

  • 0 in reply to Kyle Wolf
    Intellectual 535 points

    Boards that were designed separately.

  • 0 in reply to Leo A.
    TI__Expert 5185 points

    Hi Leo,

    I have reached out to our mid-voltage team for any insight they have into this. I will reply back here with their feedback end of this week.

    Best,

    Kyle

  • 0 in reply to Kyle Wolf
    Intellectual 535 points

    Ok, thanks for the update. I will continue to monitor the thread.

  • 0 in reply to Leo A.
    TI__Expert 5185 points

    Hey Leo,

    I received an email for this from Sean, I will follow up there once I get a response from our mid-voltage team.

    Best,

    Kyle

  • 0 in reply to Kyle Wolf
    Intellectual 535 points

    Any news on this? Thanks.

  • 0 in reply to Leo A.
    TI__Expert 5185 points

    Hi Leo,

    I have tested LMG2100R044 using the LMG2100EVM-078. I biased the LMG2100R044 Vdd and turned off the high side FET, I applied 56V to VIN. (leaving the device idle, no switching) I removed the heat sink and with no airflow I measured the thermals of the device using thermal camera and this showed no heat originating from the device.

    This shows that the energy being consumed is not being consumed by the GaN device, if it were consumed by the GaN device there would be a temperature rise in the GaN device.

     

    Best,

    Kyle

  • 0 in reply to Kyle Wolf
    Intellectual 535 points

    Kyle,

    Thank you but I see I should have been more clear in my original description of the problem.

    Both of the boards that I have the one with the LMG5100 and LMG2100 when they are idling the devices are switching at 800kHz with a 50% duty cycle. So the net output between both halfbridges is 0.

    This is the situation where the board with the LMG2100 devices draws a lot of additional power.

    Regards;

    Leo

  • +1 in reply to Leo A.
    TI__Expert 5185 points

    Hi Leo,

    That makes sense, this will add the Coss losses for the devices. If the device is switching at 0 load current, there will still be Coss loss. From my calculation this should be just over 1 watt of loss from Coss loss from the LMG2100 when switching at 800kHz at Vds 56V.

    The LMG5200 should have lower Coss loss because it is higher Rdson which will lead to smaller die area and therefor lower Coss. It is also a lower Vds rated device (80V) so the Coss will also be slightly lower due to this as well. 

    Best,

    Kyle

  • 0 in reply to Kyle Wolf
    Intellectual 535 points

    Ok, I do see this now the COSS and CISS for the LMG2100 are significantly higher vs the LMG5200.

    Considering the LMG2100 is a newer and faster device one would think these numbers would be lower.

  • 0 in reply to Leo A.
    TI__Expert 5185 points

    Hi Leo,

    The LMG2100 is new generation device than the LMG5200 so the figure of merit have gotten better, but Coss is a factor of the die area and is inversely related to the Rdson. Since the Rdson is over 3 times smaller even with the process improvements the Coss will still be larger.

    Best,

    Kyle