ringing at input current

[fifahaywire]

i'm analyzing a buck converter and as i see the input current waveform, i can see something which i believe is called ringing.

View attachment 64755

and as i increase the frequency, the waveform turns into this.
View attachment 64756
can somebody explain the reason?

 

[ccurtis]

As you increased the frequency, the pulse width is also decreased and as a result, in the bottom picture you are looking at a smaller portion of the green trace as you are in the top picture.

Edit: In other words, you are only seeing the first couple of cycles of the ringing in the botton trace of the bottom picture as compared to the bottom trace in the top picture because the pulse is terminated sooner in the bottom trace.

 

[fifahaywire]

As you increased the frequency, the pulse width is also decreased and as a result, in the bottom picture you are looking at a smaller portion of the green trace as you are in the top picture.

Edit: In other words, you are only seeing the first couple of cycles of the ringing in the botton trace of the bottom picture as compared to the bottom trace in the top picture because the pulse is terminated sooner in the bottom trace.

is it that when i increase the frequency and then the ringing sort of takes a longer time to go back and forth?

 

[ccurtis]

If you measure the time it takes for the first one and a half cycles of the ringing in the top picture, it is quite close to the amount of time it takes for the one and a half cycles of ring in the bottom picture. It looks more like the ringing is the same in both pictures except that is it cut off in the bottom picture because the pulse above it terminates sooner, cutting it off. The peak amplitude of the ringing in both pictures looks close to equal, also.

 

[fifahaywire]

If you measure the time it takes for the first one and a half cycles of the ringing in the top picture, it is quite close to the amount of time it takes for the one and a half cycles of ring in the bottom picture. It looks more like the ringing is the same in both pictures except that is it cut off in the bottom picture because the pulse above it terminates sooner, cutting it off. The peak amplitude of the ringing in both pictures looks close to equal, also.

i think i get it. i mean, i increased the frequency, of course the pulse was cut off quite earlier in the second picture. but, do you have any idea on why the ringing could have occur? is it the power supply with it's inner capacitance or is it something else?

 

[ccurtis]

Without a schematic showing the points you are monitoring it's not possible to say, but in general ringing occurs whenever there is inductance or capacitance or both with a current/voltage step (pulse) applied to them in parallel or in series (like you see in switching power supplies, by golly). It can be reduced by lowering the Q, such as by adding resistance, or reducing the rise time of the pulsed source that is hitting them. The frequency of the ringing can be changed by changing the amount of capacitance or inductance. It can also be caused by circuits or components (such as op amps) that react slowly to fast changes (under-damped). They overshoot their mark, come back and undershoot, then overshoot again, and so on, until they finally settle to a final value. It can also be caused by a mismatched impedance between a pulsed source and a load at the end the wires feeding the load. It can be difficult to quickly diagnose and eliminate. Sometimes it can be reduced by brute force filtering it out if doing so doesn't affect the desired parts of the signal adversely.

 

[fifahaywire]

Without a schematic showing the points you are monitoring it's not possible to say, but in general ringing occurs whenever there is inductance or capacitance or both with a current/voltage step (pulse) applied to them in parallel or in series (like you see in switching power supplies, by golly). It can be reduced by lowering the Q, such as by adding resistance, or reducing the rise time of the pulsed source that is hitting them. The frequency of the ringing can be changed by changing the amount of capacitance or inductance. It can also be caused by circuits or components (such as op amps) that react slowly to fast changes (under-damped). They overshoot their mark, come back and undershoot, then overshoot again, and so on, until they finally settle to a final value. It can also be caused by a mismatched impedance between a pulsed source and a load at the end the wires feeding the load. It can be difficult to quickly diagnose and eliminate. Sometimes it can be reduced by brute force filtering it out if doing so doesn't affect the desired parts of the signal adversely.

thanks for your reply, ccurtis. you have helped me tremendously
yeah there is an inductor and capacitor but no op amps. I'm having a buck converter to power my LEDs. since that is the input current pulse, is it really alright to have pulses that has ringing as such in the second picture? the first one looks fine as it finally did settle to a value before the switch turns off, but i have doubts on the second picture.

 

[ronv]

Without a schematic it is hard to tell, but one possibility is that you didn't use a schottky diode on the switch side of the inductor?

 

[ccurtis]

thanks for your reply, ccurtis. you have helped me tremendously
yeah there is an inductor and capacitor but no op amps. I'm having a buck converter to power my LEDs. since that is the input current pulse, is it really alright to have pulses that has ringing as such in the second picture? the first one looks fine as it finally did settle to a value before the switch turns off, but i have doubts on the second picture.

You're welcome. If the bottom trace is the actual LED current, I'm hard-pressed to understand how the reverse (negative) current peaks in the ringing are so large.

 

[johansen]

looks to me like the self capacitance of your inductor.