Maker Pro
Maker Pro

BUCK switching waveform ... problem

x1215

Oct 30, 2016
9
Joined
Oct 30, 2016
Messages
9
I have a 5V to 1V bulk which is built in PMIC, the schematics diagram is as below picture
KVT7HSW.jpg

I measure at the point LX, the waveform as below is wierd, I think it should be square waveform, but it is a little distored.
XtG12Rm.jpg

Need help what is going on?
 

Bluejets

Oct 5, 2014
6,901
Joined
Oct 5, 2014
Messages
6,901
So am I reading this correctly?

You have a 26mv spike lasting 2.5usec?
What is the load?
 

x1215

Oct 30, 2016
9
Joined
Oct 30, 2016
Messages
9
So am I reading this correctly?

You have a 26mv spike lasting 2.5usec?
What is the load?

YES you are right.
Load is CPU's V-CORE, current consumption MAX to 2A.
Because I am not much familiar with DCDC buck, so I don't know if the LX signal is fine or not.
 
Last edited:

duke37

Jan 9, 2011
5,364
Joined
Jan 9, 2011
Messages
5,364
If Q1 turns on fully then the LX waveform should have a flat top at 5V.

I thought that Q2 should have a diode across it to take the current while Q2 is turning on.
 

WHONOES

May 20, 2017
1,217
Joined
May 20, 2017
Messages
1,217
If Q1 turns on fully then the LX waveform should have a flat top at 5V.

I thought that Q2 should have a diode across it to take the current while Q2 is turning on.
No, the waveforms shown are quite normal.
The circuit shown is a Synchronous Rectifier type where the normal diode is replaced with Q2. Q2 would normally be on when Q1 is off.

The output ripple can be caused by a number of things.
  • Parasitic capacitance of the inductor windings. This allows the fast edges of the switching waveform to partially bypass L.
  • Poor ESR of Cout. Large electrolytic types will exhibit this. Several smaller value types will help to mitigate it. A few 100n ceramic caps will also help.
  • Poor layout will also contribute. One of the main candidates here would be poor grounding. All the high current points should sum at one junction and not be daisy chained.
  • A lot of the rubbish on the output could be removed by the use of a high impedance Cylindrical Ferrite core on the output leads.
 

WHONOES

May 20, 2017
1,217
Joined
May 20, 2017
Messages
1,217
Loading the output will make the ripple worse.
 

BobK

Jan 5, 2010
7,682
Joined
Jan 5, 2010
Messages
7,682
Some switchers do not work well with no load.

Bob
 

WHONOES

May 20, 2017
1,217
Joined
May 20, 2017
Messages
1,217
I agree but, that generally affects their ability to regulate properly.
 

x1215

Oct 30, 2016
9
Joined
Oct 30, 2016
Messages
9
the output have a ARM V-CORE sink,I will try to put some 100nF first.
 
Top