How to force LNK454 to shut down?

5 posts / 0 new

I am designing an industrial LED light that is remotely controlled.  It has a small micro which has a separate bleeder power supply and needs to be able to shut down the LED output.  There's an additional need to keep the LNK454 from switching at all if possible, to reduce power line noise.

I  have used a MOSFET power switch in series with the LNK power, but that's brute force and adds heat in the switch control circuit.  I'd like to pull the FB voltage up above 2V instead.  So my questions are:

1)  What is the purpose of the 2uA bias at the FB pin?  Does this ensure the LNK stays off if FB becomes disconnected?

2)  What is the input impedance of FB?  The app notes show 1k series resistors and 10nF.  I'd need to raise that to at least 10k and 1nF in order to add an additional switchable pull-up (above 2V) to turn the LNK off.

I can easily disconnect the LED output string, but then the LNK's auto restart still makes noise on the AC line, and I'd like to avoid that.

Do you have other suggestions?

Is there a different part I should consider?  - I don't need dimming ability.



Force BP pin to Source pin to shutdown LNKPL. It can re-start once BP pin is released.

The feedback pull-up current of -1uA is the current used for detecting if the FB pin is disconnected.  Floating this pin will still enter to auto-restart mode.

FB input impedance is closed to 1Gohm. Yes you can use 10K as series impedance to FB pin to force it above 2V.


Thank you very much for your quick response PI-Jono,

Shorting BP to ground\ would cerctainly be easy and effective, however I want to understand the effects.

It looks like on the LNK454, the current could be up to 5.9mA, and on the LNK456 it could be up to 8.3mA

That would cause a power dissipation of roughly 240*0.0059 = 1.42 watts for the LNK454 and roughly 2 watts for the LNK456.

My system runs on 240VAC, and I haven't yet decided between the LNK454 and 456.

The ambient temperature in my fixture can be as high as 85C, so judging by the thermal resistances, this much power dissipation would cause damage,

Perhaps I should consider adding a heatsink to the package?


You weren't aware of my operating voltage or fixture temperture when you made your suggestion, sorry.


 I'm running at 60Hz, so I guess I should I expect a "hiccup" every second or so.  During the "hiccup", the PWM will be at 33%, until the output reaches regulation *normally*, but if I hold the FB pin above 2V all the time, how long should I expect the unit to be "on" during auto restart?  Would it even turn on it's PWM at all?





If you can guarantee to hold the FB pin continously above 2V then no switching will occur in the drain and source pin.  You can use this approach to turn-on or off the LED driver.

It will not even turn on the PWM at all but internally it waits for FB pin to go down.

Nice, you have a point in the power dissipation once BP is forcely shorted to ground. 



Excellent.  I was hoping that would be the answer :)

This will really improve my design.

Thanks again.