Buck converter ringing waveforms with small loads / General Science and Electronics / Forums

Forums

4hv.org :: Forums :: General Science and Electronics

« Previous topic | Next topic »

Buck converter ringing waveforms with small loads

Move Thread

LAN_403

Alex M

Sat Mar 02 2013, 12:49AM

Registered Member #3943 Joined: Sun Jun 12 2011, 05:24PM
Location: The Shire, UK
Posts: 552

Hi,

I just put together a buck converter using an LM2576-ADJ and just have a question regarding the waveforms when only light (sub 250mA) loads are connected to it. 10v/div.

Below is the inductor voltage waveform with a 450mA load on it (9v out). As you can see it looks fine and its pretty much like this for heavier load currents too.

But with a much lighter load of 230mA (9v out) I get this. Just what is causing this ringing?

The lighter the load, the worse this ringing gets. To me it looks like its the effect of the IC increasing the dead time, or the inductor is too small for lower current loads (I am using a 100uH inductor). Supply voltage for the LM2576 is about 30v.

I looked in the datasheet for the LM2576 for anything regarding minimum load current/power but could not find anything.

Any pointers would be appreciated,
Thanks.

Dr. Slack

Sat Mar 02 2013, 08:16AM

Registered Member #72 Joined: Thu Feb 09 2006, 08:29AM
Location: UK St. Albans
Posts: 1659

At 450mA, the inductor is in continuous current mode, but is discontinuous at the lower current. This gives you pointers for where to look for its description in data sheet. 'The inductor is too small' is the reason the waveform has gained a third level at 0v, but is not the reason for the ringing.

At the lower current, the current stops each cycle. For a fixed frequency converter, there will always be some mean current below which this happens, so it is not a design error per se. The ringing on that transition means that although you apparently have tidy, tight paths for the current and its switching for input and output (no ringing there), the residual current in the inductor when either the IC decides to switch it off, or the freewheel diode gives up conducting at 0.7 or 0.4v or whatever, is not being handled so well. Have a look in those areas to fix your ringing.

It may OTOH just be where you have your scope probe grounded. You might accidently be including some current paths and not others to its 'ground'.

It may be that transition actually has more wallop than the other two. If your freewheel diode is a slow type, then charge storage will mean that when it finally does turn off, it will do so with a helluva twang, whereas the main two transitions happen through FET switching, which doesn't generate such fast edges.

Steve Conner

Sat Mar 02 2013, 02:41PM

Registered Member #30 Joined: Fri Feb 03 2006, 10:52AM
Location: Glasgow, Scotland
Posts: 6706

It is normal for SMPS to go into discontinuous current mode at light load. The control loop is supposed to handle it gracefully.

The ringing is also normal, it is caused by the inductor's self-capacitance, junction capacitance of the switching devices and so on, resonating with the inductance. The resonance is excited by the sudden collapse of the inductor voltage when the current in it reaches zero. It's only a problem if it makes you fail EMC.

Alex M

Mon Mar 04 2013, 04:26PM

Registered Member #3943 Joined: Sun Jun 12 2011, 05:24PM
Location: The Shire, UK
Posts: 552

Thanks guys those replies really helped! Turns out I had actually downloaded a very watered down datasheet for the IC and found the one on on-semiconductors website to be much more detailed.

Guess that is what is causing it then, at lower currents the IC is going into discontinues mode and causing the ringing with the non-ideal aspects of the inductor.

From what I can tell its a case of either increase the inductor size for lighter loads, or just put up with it (unless a snubber was used). The datasheet from on-semi shows larger inductors for lower current loads too.

The diode is a 1N5822 Schottky diode, which is actually suggested in the datasheet for the LM2576. The inductor is a 100uH powered iron toroidal one with a saturation current of 5a. Capacitor is a 1000uF low ESR type, I have bypassed this with a 100nF film too though out of habit. But I don't see this causing any problems.

But anyway this is the reason I wanted to make this circuit, to try and learn abit about them before I start putting them directly into other projects.

Some pics of it.

A mess I know but its the only way I can lower track inductance without a PCB.

Orange wire goes to the feedback pin.

Sulaiman

Tue Mar 05 2013, 01:32PM

Registered Member #162 Joined: Mon Feb 13 2006, 10:25AM
Location: United Kingdom
Posts: 3140

Get LM2596 modules from eBay
.... a buy-it-now 2A (or 3A) buck converter for GBP1.15 including p&p

(a few colleagues and I have independently bought a similar kit
.. all run at <<150 kHz, mine at 55.2 kHz
so these kits are cheap because the ic's are out-of-spec.)

Exactly the same performance/symptoms,
... high current = continuous mode, low current = discontinuous mode,

because I want c150mA to 300mA at 5V from a 10V to 15V dc supply with low emi
to use in a home-brew rf transceiver
I replaced the on-board 3A 33uH inductor with a low-leakage 1mH 650mA inductor ... cost almost as much as the entire module !
Replaced the capacitors too as they are not best quality,
still worth GBP1.15 !

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob