pwn signal getting spikes when connected to gate driver

Thread Starter

ak52

Joined Oct 15, 2014
230
Hello Everyone,

My pwm signal gets very spiky when i connect it my my gate driver,i'm not able to deduce why.
I have tried reducing the period of the pwm signal ,but no effect.
The gate driver circuit is as followes:
gatedriver.JPG

At the moment Jumpers J8,J9 and J11 are connected.

When i disconnected J9,the pwm signal from the controller is like this:
pwm_without_j9.jpg

When i re-connected J9,the pwm signal from the controller is like this:
pwm_with_j9.jpg

I do not know what is causing these sudden spikes ,it is the gate driver issue?
The same spikes are also carried to the output of the gate driver.

Thanks,
Arun
 
Last edited:

Papabravo

Joined Feb 24, 2006
21,159
There are a number of possibilities, but without seeing the layout it may be difficult to speculate. It looks like you are driving a digital input with an opamp. This may or may not be a good idea depending on the source and termination impedance of the connection. Are you sure the part is suitable for single supply operation, and will it swing rail-to-rail? The first thing to do is check your probe calibration/compensation. The next thing to do is check for unexpected inductance in your signal and ground paths. If the spikes are narrow enough it is possible they are artifacts of the measurement process.
 

Papabravo

Joined Feb 24, 2006
21,159
Should be the Gate Driver Input, according to the scope trace settings. Looks like one opamp is connected and the other is not.
 

Thread Starter

ak52

Joined Oct 15, 2014
230
I am measuring the pulse on the second pin of J9.
Also i am not using the opamp at the moment,Jumpers J7 and J10 are always open.
 

Thread Starter

ak52

Joined Oct 15, 2014
230
Why a pull-down resistor on pin 12 but not 10?

:confused: In post #1 you said J10 was connected.
Apologies i meant J11(edited the post).
I am designing a non-synchronous buck ,so i don't need the bottom pwm signal.,but in the future when i use the bottom FET for a synchronous buck then the bottom pwm signal will be used.(and R22 will be removed)
As of now i am pulling down the low side pwm signal so that it remains grounded permanently.
 

Papabravo

Joined Feb 24, 2006
21,159
So the opamp input impedance is in parallel with the driver impedance? That doesn't seem like a good thing to do. How long is the signal trace driving the driver chip?
 
Top