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.

TIDA-01573: LMG1020

Part Number: TIDA-01573
Other Parts Discussed in Thread: LMG1020,

According to TIDA-01573 app note, in order to really benefit from the high speed capability of the LMG1020 driver, separation of signal ground and power ground is recommended. So the source Kelvin pad of the GaN FET is connected to the signal GND while the other source pads are connected to the power GND. This way the two grounds are only connected together inside the FET.

My question refers to how to assign grounds for the following laser driver daughter card that plugs onto a motherboard:

Input voltage is 12V coming via a connector from the motherboard.

On the laser daughter card there are:

LDO providing 5V for the LMG1020 and also for some digital circuitry that shape the pulse for LMG1020.

Boost converter providing high voltage for the capacitor bank.

GaN FET and adjacent circuitry (capacitor bank, laser diode, clamping diodes, current sense resistors).

Following the app note recommendation it seems the power GND needs to connect to the source of the GaN FET and the GND nodes of the boost converter while all other GND nodes connect to the signal GND.

The connector for the 12V from the mother board has only one pin for GND. Which GND should this pin be tied to: signal or power GND?

Please advise.

Thank you,

Michael

  • Hi Michael,

    thanks for your question on LMG1020 and TIDA-01573,
    You explained this very well and your correct, the signal GND is the same as the driver GND and the 12V GND would need to stay separated and put on the power GND. Driver (signal) GND needs to be separated from all the power traces so noise and GND bounce does not occur which might affect the input signal. Ideally the driver GND is connected to the GaN FET kelvin or source pin as close to the FET as possible. The source of the GaN FET and its trace has some parasitic inductance effects especially for high di/dt seen with high laser diode current. This di/dt creates creates a corresponding voltage drop from GND equal to L*di/dt. This voltage drop works against the drive voltage slowing down the rise and fall times of the GaN gate and therefore the peak laser current. To bypass this parasitic source inductance as much as possible the gate driver GND needs to see a very low impedance to the source in order to drive it efficiently and remove common source inductance. The goal is to separate the driver GND from power GND to reduce the common source inductance. The driver GND should contain the LDO 5V GND as well as the pulse shaping circuitry or signal GND and the bus voltage circuitry or 12V GND should contain the power GND.
    Does this make sense? Please let me know if you have any more questions,

    Thanks,
  • Hi Jeff,

    I understand all you were saying. My problem is the 12V connector between the laser daughter card and the mother board has only one pin for GND because the mother board uses a single GND. So the question is what should I tie this GND pin to: signal GND or power GND? Since the connector has few other pins carrying some digital signals it makes sense to have its GND pin connected to the signal GND. But in this case the input current to the boost converter would use as return a path that’s shared with the signal GND return path.

    My feeling is this is OK because the input current to the boost is pretty much free of high current spikes due to the electrolytic at the input of the boost converter. Therefore the additional current flowing through the signal GND return is basically a relatively small DC current that will not impact proper function of the GaN driver or the other digital circuits.

    Is my thinking correct?

    Thank you,

    Michael

  • HI Michael,

    Thanks for clairfying this post with a SBD!
    My thinking is similar to your feelings, either way the 12V return current would be circulating through both GNDs so there should be a star point ground or single point ground connection between signal and power ground for the circulating current that is separate from kelvin ground. As long as the 12V current stays on the laser daughter card circuit the ground for the 12V should be able to connect to either signal ground or power ground since the return current would run through both with either ground connection option. As long as the circulating current stays on the daughter card it should be ok. Since the 12V return current would see power ground you can put a small input capacitor on LMG1020 input to filter out any noise that might reach the input via parasitic capactiance. Does this help answer your question? please let me know if you have any more questions.

    All information in this correspondence and in any related correspondence is provided “AS IS” and “with all faults”, and is subject to TI’s Important Notice (www.ti.com/.../important-notice.shtml).

    thanks,