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.

LMG1210: Best low side GaN driver, LMG1210 or UCC27611 ?

Part Number: LMG1210
Other Parts Discussed in Thread: UCC27611, LMG1020

Hi

We have a synchronous rectifier setup using two EPC2045 (peak voltage < 30V) switching at 1MHz.

We can either use two UCC27611 or one LMG1210 (with both drivers configured as low side)

Which solution would be the better choice from a performance perspective?

Best regards,

Roger Kamben

  • Hi, Roger,

    Thanks for your question on the LMG1210 vs UCC27611.

    My colleague will get back with you on Monday. Have a nice weekend!

  • Hi Roger,

    Either of these devices would work, but I can list below a comparison of some key features so you can choose what is best for your application:

    • In terms of timing, the LMG1210 has slightly better propagation delay (LMG1210: 10ns typical, UCC27611: 14ns typical) and faster output rise/fall times (LMG1210: 3.5ns typicalUCC27611: 5ns typical both for 1nF load). 
    • In terms of peak source/sink currents, UCC27611 has better drive strength (LMG1210: 1.5A peak source, 3A peak sinkUCC27611: 4A peak source, 6A peak sink). 
    • With LMG1210 you would have a size advantage since you are using only 1 device instead of 2. 
    • Both devices have UVLO protection, and LMG1210 has overtemperature protection.
    • UCC27611 can handle -40C to 140C, while LMG1210 can handle -40C to 125C. 

    Best regards, 

    Leslie

  • Hi

    Yes this is all evident from the datasheets.  What I'm looking for is a deeper knowledge of how these drivers work in practice.

    It seems LMG1210 has much better grounding and is like 3rd generation GAN driver while the UCC27611 is quite old.

    We have a test board here with the UCC27611 on the secondary side (rectifier) which has severe ringing problems.

    The LMG1210 is used on the primary side however and is behaving very well. The difference could of course be due to layout and use-case.

    For our next revision we're gonna do everything we can to improve the layout, but it would be helpful to chose the best driver which will set the bottom line of what is possible to attain with regards to ringing and switching performance..  (package inductance etc)

    Best regards,

    Roger Kamben

  • Roger,

    As my colleague Leslie previously highlighted, the drivers do not have the same drive strength. With regards to ringing and switching performance in your application, there are several external factors that will obviously contribute this however please note that UCC27611 has much higher drive current which will generate much faster slew rates and di/dt which will obviously translate to higher overshoot/undershoot compared to LMG1210 which generates slower dv/dt (because of lower Ipk) when driving similar load. UCC27611 might be too strong for your application, you may need to adjust the gate resistance to slow the dv/dt at the gate and di/dt at the drain.

    This is not an apple-to-apple comparison as the drivers use different designs structures: LMG1210 is a half bridge driver while UCC27611 is low-side driver. LMG1210 uses delay matching and deadtime features to enhance the performance for its specific target application while UCC27611 has a simpler architecture and a different package. 

    I hope this helps.

    Regards,

    -Mamadou

  • Hi

    Yes of course, all things being equal a higher di/dt will cause more ringing, but that's not what I'm asking.

    My point is these drivers aren't "equal" so lets take something specific then:

    What is the package inductance from the Internal Driver GND (on Die)  to external GND pin/plane for the LMG1210 vs UCC27611 packages?

    If one driver has twice the current it can still have less ringing if the effective inductance is much less ( assuming ideal layout )

    Best regards,

    Roger Kamben

  • Hello,

    The package inductance will depend on internal bond-wires used which will obviously increase parasitic inductance and resistance between die and leadframe. For details on the bond wires used and package inductance, I suggest our customer support center which may better address the IC's quality process.

    Meanwhile the LMG1020 is also an alternative to the UCC27611 (more recent than the UCC27611) and using similar technology and process than LMG1210 though I will caution that you may encounter similar ringing in same test conditions than LMG1210 due to fast dv/dt.  

    Let us know if you have additional application related questions.

    -Mamadou