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.

LMG1210EVM-012: Converting the LMG1210 EVM to 200V

Part Number: LMG1210EVM-012
Other Parts Discussed in Thread: LMG1210, LMG1020

I need to generate pulses with 10ns pulse width at 200V .

Can I modify the EVM circuit by changing the GAN FETs and Bootstrap diode to 200V specs?

Will there be any issues that will crop up if I make these changes?

If so can you help me out on how to make these changes?

Thank you

  • Hi Sai,

    thanks for asking about LMG1210, and welcome to e2e!
    For operating the EVM above 100V we have already populated the EVM with 250V bus capacitors so the only components that needs to be replaced is the 100V bootstrap diode D1 and the GaN FET with a similar footprint as EPC2001 like EPC2010 which is a 200V FET. The diode should not be difficult to replace since they have leads to work with. An EPC GaN FET which should be the only package footprint able to similarly fit, is more difficult to solder. It can be done by hand for a couple of devices or reflowed with paste and an oven for mass production. You can follow the EPC's assembly recommendations of SAC solder paste with no clean flux and careful handling and reflow profile. An Xray can be used to make sure the solder process was properly reflowed. please let me know if you have any more questions with this 200V EVM mod or need help achieving 10ns pulses.

    thanks,

  • Hey Jeff,
    Thanks a lot for your reply ,it really helped me a lot.
    I have a few more questions regarding this circuit.
    1)The bootstrap diode presently being used is 100V 0.25A.
    Should I increase the current rating for the 200 V diode ,If so by how much?
    2) Coming to the output side,The 4.7uH inductor and all the caps at Vout ,is that necessary for my circuit or can I just remove it.
    3)I didn't understand the reason for using that inductor,Can you tell me what purpose does that inductor serve?
    4)In case I want an isolation between the pwn and enable buffers and the driver inputs,what are the specs i should be looking at and any suggestions for isolation ics that I can use?

    Thanks Jeff,
    Regards
    Sai
  • Hey jeff,
    I have another question,You have suggested EPC2010
    Would it be ok if I use EPC2034,Apart from the footprint would it affect the circuit in anyway?
  • Hi Sai,

    thanks for your question,
    EPC2034 has a different footprint like you mentioned. It will not fit on the EVM however you can modify the EVM and remake a prototype using EPC2034 assuming the layout has space. It should not affect the circuit other than allowing higher current pulses since the total gate charge and Rdson values are very similar. The biggest difference I saw was the Coss is double for EPC2034 and might add do the switching losses. Let me know if you have any more questions.

    thanks,
  • Hey Jeff,
    I had posted the same questions above,I guess you must have missed it so its a repost
    I have a few more questions regarding this circuit.
    1)The bootstrap diode presently being used is 100V 0.25A.
    Should I increase the current rating for the 200 V diode ,If so by how much?
    2) Coming to the output side,The 4.7uH inductor and all the caps at Vout ,is that necessary for my circuit or can I just remove it.
    3)I didn't understand the reason for using that inductor,Can you tell me what purpose does that inductor serve?
    4)In case I want an isolation between the pwn and enable buffers and the driver inputs,what are the specs i should be looking at and any suggestions for isolation ics that I can use?

    Thanks Jeff,
    Regards
    Sai
  • Hi Sai,

    Thanks for letting me know about your other questions.
    1) The VDD cap charges the bootstrap cap with a fast peak current through the bootstrap diode every switching cycle. The current, which is highest during start up can be limited by using a bootstrap resistor. BAT46 has 2.5A of peak current and should be enough for most applications since LMG1210 is 5V. The diode should be more than the 200V bus voltage with margin.
    2) the output inductor and capacitor can be removed if needed
    3) The inductor is needed for the power stage part or output filter of the half-bridge converter which can be for example a buck or boost topology
    4)A digital signal isolator like the reinforced ISO772x and ISO774x should be fast and high CMTI enough for your application. The prop delay should be similar to the delay of the driver or about 10ns and the output rise time which is 1ns should allow high bandwidth. The CMTI or the slew rate of 100V/ns should be able to meet your specs as well.

    let me know if you have any more questions,
    thanks,
  • Thanks a lot Jeff ,this really helps me a lot.
    Just one last question ,
    1)The feedthru cap C14 has both its ends shorted what is the reason for doing this?
    2)C14 cap has a voltage rating of 6.3V,for 200V operation will I need to increase the Voltage?
    Thanks Jeff
  • Hi Sai,

    thanks for your follow up question,

    1) C14 is a 3 terminal feed thru cap to lower the inductance in the bootstrap loop which is important to source the high peak current for the high-side FET gate. The first HB connection goes close to the HB pin and the second HB connection goes to the HB plane which part of the bootstrap diode sits on. The closest HS connection goes to the HS pin and rest of the HS pad connection goes to the HS plane on the second layer which the high-side portion of LMG1210 sits on. This can be seen by checking out the layout and schematic from the altium design files.

    2)The voltage rating of the bootstrap cap is to be as high as 2xVDD since the high-side bootstrap voltage should not exceed VDD rating. When HS goes high the bootstrap cap will remain the same voltage since its tied to HS. To account for a high voltage ringing from reverse recovery from a higher bus voltage this cap can be rated at 2xVDD to be safe. 

    let me know if you have any more questions!

    thanks,

  • Hey Jeff,
    Thanks for the reply,I have a few more questions.
    1)C14 is the bootstrap capacitor and according to all the application diagrams one end is connected to bootstrap diode/HB and HS/between the two fets which is the output.
    In the schematic I dont see the connection between HS and output pin,Is this done through the HS2 plane?
    2)Since one end of the bootstrap will be connected to output which sees 200V,shouldn't the bootstrap capacitor C14 have a voltage rating of 200V?
    3)I am trying to give a 10ns pulse at pwm,I am trying to generate this using one shot but there are not many devices for this.
    Do you have any idea how I can go about this?
    Thanks,
    Sai
  • Hey Sai,

    Thanks for asking these clarifying questions,
    1)If you are referring to the output pin as the switch-node you are correct, HS2 connects HS with a plane on the adjacent layers. HS also connects to HS2 on the top layer with a trace. The middle pad on C14 is the HS reference the two end pads are connect HB and the diode anode with low inductance
    2) with a 200V bus HB pin will see ~205V when the high-side is on however the bootstrap cap floats on HS so that the cap never sees more than VDD - Vf - Vhs where Vf is the forward voltage drop on the diode and Vhs is the voltage drop from ground due to third quadrant conduction or negative current during dead-time. for more details on this check out section 7.3.1 from the lmg1210 datasheet. let me know if this does not make sense.
    3)any function generator that is capable of 100Mhz and above should allow a 10ns pulse. The on board buffer, U2A, will shape the input signal for cleaner edges. There is a external circuit apart from UA2 that can be added that shortens the pulse from a longer 100ns pulse to as low as about 1-2ns using an AND gate and low pass filter. for more details, check out section 7.2 from the lmg1020 user guide. (www.ti.com/.../snou150a.pdf)

    let me know if you have any other questions!
    thanks,
  • Hey Jeff,
    Thanks for the reply
    1)I went through the formula for calculating the capacitance of the bootstrap capacitor but I couldn't find Qrr for EPC2010C.Should I increase the capacitance when I am changing for 200V.
    2)Also should I change the capacitance values of the capacitors between VDD and VSS
    Thanks,
    Sai
  • Hey Sai,

    thanks for your follow up!

    1) GaN FETs including EPC2010 have zero Qrr, this spec can be found at the very bottom of the spec table as well as in the abstract on the first page of the EPC2010 datasheet. 

    2) the VDD cap voltage rating is similar to the bootstrap cap voltage rating, they are both the same caps on the EVM. For 200V operation 2XVDD rating is recommended if possible. Also, the capacitance of the bus caps may need to be increased when increasing the bus voltage due to the dc bias characteristic. 

    let me know if you have any other questions!

    thanks,

  • Hey Jeff,
    Thanks for the reply
    1)I wanted to know if I will be able to use -200V on the bus
    Will I be able to force a negative voltage of 200V?
    Will this be limited by the bootstrap diode?
    If it is possible what are the things I should take care of?
    What values of bootstrap diode and capacitor will I have to use?

    2)The dead time pins are sampled on start up to set either PWM or IIM mode.
    What mode will it be set in if I leave both the pins floating or if I give 0V on both the pins?

    Thanks,
    Sai
  • Hi Sai,

    Thanks for the update, good questions,

    1) If VSS=0V and HS=-200V then you are like you said using a negative voltage on the bus. If this is not the case, do you have a diagram for review to confirm this? what application is this for, I have never seen this type of application and I am interested!? You can use a -200V bus voltage with LMG1210 however this is without the bootstrap as spec'd in the recommended operating conditions table (Vhs - Vss). For using it with a bootstrap diode, check out section 7.3.6 equation 6 in the LMG1210 datasheet to understand the largest negative voltage (about -6V DC) that can be applied to HS before large currents flow out the BST pin and ESD diode. When using a bootstrap diode the limitation for 1210 will be the BST voltage absolute minimum rating created from the negative HS voltage. However the true limitation will be the 195V of forward bias voltage that would explode the bootstrap diode. If using a -200V bus voltage leave the BST pin floating and generate the HB-HS drive voltage with an external isolated supply.

    2) The dead-time is actually sampled through out run time and can be adjusted with out power cycling the part. If the dead-time pins are grounded then the dead-time circuitry will detect this and internally fault the part so the output wont switch. If the dead-time pins are floating the device defaults to independent input mode (IIM) or 0ns dead-time in which LI and HI are used and the controller needs to create the dead-time. However it's recommended in section 7.4 of the LMG1210 datasheet that when using IIM the DLH should be tied to VDD and the DHL can be left floating to be certain of IIM. 

    let me know if you have any further questions!

    thanks,

  • Hey Jeff,
    Thanks for the reply

    When I say -200V ,I will be connecting +200V terminal to PGND and GND terminal of my supply to Vin.

    Will this operation be possible since all grounds will see 200V?

    I would like to generate +200V and -200V pulses by exchanging 200V the supply terminals between Vin of bus and pgnd

  • Hi Sai,

    This seems like it is not possible and looks like when VIN which is 0V is applied to the switch node during high-side on time the bootstrap diode initially sees 205V and would become damaged since VDD would be 205V. What application is this for? Do you have a schematic or diagram of the circuit so I can review if your circuit is feasible?

    thanks,

  • Hi Sai, 

    We have not heard from you, we must assume that you have resolved the issue. I will therefore mark this thread as resolved again and close it. 

    Please let us know if you have further questions. 

    Thank you again for your interest in our drivers.

    Regards,

    -Mamadou