Three phase inveter ouput issuse with optocouplers islation

[JUN2]

Hi, i am using opto as an isloation between a three phase gate driver circuit and controller(signal generator), but i met a problem that is when i connect for example INH3 and INL3(also the ouputs from opto) the same time from opto board to the driver board,
and when i give INL (low side mosfet switching) signal, the Vgs of low side mosfet shows nothing, but once i unplugged the INH3 connection bwteen driver and opto, the there will be vgs waveform, the whole time i didn't give any siganl to INH3.
Also when i directly give INH or INL signal without the opto islation, the waveform is right, high side and low side mosfets are switching, the problem should not be driver circuit, i tried to use a seprate Vcc power supply, and also tried to seprate INH and INL opto on different pcb board, but the issue still there, i think it might becuase they shared the same ground, however they have to share the same ground of driver, so i just get stucked, anyone could help, thank you so much

 

[JUN2]

Hi,

well done so far. (regarding app note and calculations)
Be sure to use to use "worst case" input values..

I checked your 70nF and the given 7500 Ohms.
It gives a time constant of 0.5ms. (70nF x 7500 Ohms).
This means the resistor discharges the capacitor within 0.5ms from initial voltage (100%) down to 37%.
37% is way too low for a reliable operation ... this simply means your HIGH phase must be less than 0.5ms. I´d say max 0.1ms.
But according your information .. your application can satisfy this.

From the pictures I see the capacitor voltage. it is way below the expected 15V (minus some drop).
--> here you first need to investigate why this is the case.
Either there is a measurement error .. or a problem with your circuit.

One idea:
I showed you the charging path. Did you notice that the charging path involves the LOW side MOSFET?
Be sure the MOSFET is ON (during the expected charging time)! This is urgent .. other wise the bootstrap capacitor will not charge correctly.

Thus usually one first tests proper operation of LOW side ... then - after LOW side is OK - one can test HIGH side.

For better understanding what´s happening:
* please use the schematic to tell us where exactly you connected scope signals and scope reference.
* also use multiple channels of your scope to see other signals as (timing) reference. Like when testing HIGH side .. always give an additional signal of the LOW side.

Klaus

Hi, Klaus
Thank you for replying,

70nf is indeed what I calculated, here all the bootstrap caps I use is 10uf.


And yeah I noticed that I’d need to turn on low side first and turn on high side, there might be some operation issues, I will use this test strategy again, and see what the results are.

Also the signal connection, the ground of signal I always connect to the circuit ground(VSS pin, Vss pin is GND, so I connect here) no matter I test high or low side,

and the the signal connection just connect to HINx and LINx, one signal at a time, I will be using a controller to give multi channels signal
Thank you again for helping
Regards
Jun

--- Updated Dec 13, 2024 ---

Optocoupler drivers need to have <1kOhms between Pin 2 & Pin 3, Optocouplers can easily pick up noise and drive the output HIGH , specially during start-up when the controller signal is still high impedance, this will create false High trigger in your gate driver, and Short circuit in the Half Bridge Power Mosfet Supply and Ground. Connect resistor to that pins then continue debugging.

Hi,
Thank you for replying, so I need to add a res between input pin and input ground?
I will try to add the res, thank you

 

[DIY TechFusion]

Hi,
Thank you for replying, so I need to add a res between input pin and input ground?
I will try to add the res, thank you

Yes, Pin 2 to Pin3 GND

 

[JUN2]

Hi
Yeah, i calculated the bootstrap capcitance based on their design tip and also applicaiton note so i got value of 70nF.
So i should increase C3 and also R22 to lower the voltage drop?
regards
Jun

Hi, klaus
So like you suggested, i used a controller with all channel connected, the testing results are correct, high side mosfet switching also correct and consistent, phase to phase also switch correct,

so for next step i am thinking that i will connect load to it, and for this load, i am thinking i could three resistors, but i am not sure if i could be able to observe current waveform, or i should just connect a motor for testing.

and if i use resistors, how i can connect the resistors? i know one line of resistor conenct to three phase ouput, what about the other line of res? should i connect the other line to the circuit ground? or
Thank you
regards
Jun

 

[KlausST]

and if i use resistors, how i can connect the resistors? i know one line of resistor conenct to three phase ouput, what about the other line of res? should i connect the other line to the circuit ground? or

use three resistors and connect them the same was as you would wire the three coils of a motor.

Klaus