LNK306 non-isolated buck convertor dropping out on load changes

Posted by: ccarlsen on 02/17/2011

I have an existing LNK306 design using the non-isolated buck circuit in your app notes. I used the PIXIs Designer to calculate component values. The supply was working fine for it's original application, however I am trying to use this design in a system where we are switching relays on and off, so the load is inductive and changes suddenly. Every time the relays are switched, the DC output of the convertor drops to 0 for a fraction of a second. If I look at the DC output with a scope, I can see the voltage drop off, the convertor seems to be driving somewhat, as I can see brief pulses (10 or more pulses at a time), but none of them go to the full output voltage. I am using the 10 uF soft-start capacitor as recommended in app note 47 (47uF electrolytic). I need an immediate answer, as we will need to make the decision whether to use this product within about a day.

Thank you,

Curtis Carlsen

Comments

Without knowing the load characteristic exactly, it is not easy to tell what could happen.

But if the load is not resistive and not controlled, there might be an overload condition occurred.

If an overload is detected, the LNK306 might be in auto restart mode to protect the system and device itself.

The load includes 30 mA constant current running a microcontroller and peripherals. The relay load consists of either 30 or 60 mA inductive current switched in and out by integrated darlington drivers with flyback diodes.

Apparently the flyback diodes are not removing enough of the spiking from the inductive load. If I switch in a resistive load of 120 mA, I see no problem at all.

So the problem is apparently from the inductive nature of the load. The only recommendations I found in your app notes and data sheets was to add a soft-start capacitor. My current circuit uses the maximum recommended soft-start cap of 49 uF, and it has not corrected the problem. Are there any other circuit modifications I can carry out that well alleviate the problem?

Thanks for your help.

Ï'm about to build a circuit whith the same characteristics. did you have the problem on turn on or turn off or both?

The problem occurred at turn on. The only way I found to improve it was to put a series resistance in front of the inductive load. About 100 ohms did the trick.

Thanks for your reply.
I'll have to watch out for this. It is so strange that this occurs at turn on. The inductor load current increases at a rather slow rate. I would have thought that switch-off voltages got into the feedback path and did strange things.
I will be switching the relay at 40 kHz to reduce power consumption, so it's going to be interesting to see how that works out.

I was also expecting the problem to be on release. It took me a little by surprise. My best guess is that it is the contact bounce causing chatter on the line. The contacts only bounce when closing, not when releasing, so I only saw the problem on contact closure.

Curtis

i have use two relay at 5VDC, when first relay on 125mV drop across at VCC, when first and second relay on 250mV drop across at VCC.

please any one can say what going on.

Hi Piyush mathukiya,

Can you share your design so I can help with your problem? Your schematic and PIXls design can help me asses your problem. Also what is the model/manufacturing pn of your load relay. Is the drop you are monitoring is across the power supply output capacitor or at the VCC of your relay? Thanks.