• State TI Thinks Resolved
  • Locked Locked
  • Replies 3 replies
  • Answers 1 answer
  • Subscribers 63 subscribers
  • Views 676 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

LM3409HV: Shunt FET switching ringing.

Matthew Santoro
Prodigy 20 points
Part Number: LM3409HV
Other Parts Discussed in Thread: LM3409

I am using the LM3409HV as a modular panel driver. meaning you can add different LEDs to the output, with different specs (stack voltage from 11-22v and what not).
Firstly let me know if allowing different stack voltages with this chip will end up being a pain, because it is now. and im wondering if i should switch to another chip.

But the big questions is when i am using FET dimming, when the N-ch gate is going low, the Output voltage is ringing and spiking a lot.

Any help would be super, Do you guys have any suggestions on what are some common fixes, 

  • 0
    TI__Mastermind 25500 points

    Hello Matthew,

    The LM3409 was designed to adapt to different output voltages so it should not be a pain. Just design so that you will have the maximum switching frequency you will use at 50% duty cycle (or the closest load to 50% duty cycle).

    Noise and spiking with shunt FET dimming is a common issue. So my suggestions:

    First, make sure you are not using an output capacitor. It doesn't sound like you are because otherwise you would likely have damaged the shunt FET.

    Second, minimize parasitic inductance. Use as short of traces as possible between the LED load and the shunt FET. Ideally the FET is directly next to the LED(s) rather than directly next to the regulator (assuming the regulator and LEDs have some distance between them). If possible it is also good to minimize the inductance of the LED load itself. For example if you have a string of LEDs connect them together with the forward current trace and then route the return current trace directly next to the forward current trace to minimize the loop it creates. In the case of the LM3409 where the return is ground a good ground plane will also work.

    Lastly, if minimizing parasitics isn't enough (it often isn't) then place an RC snubber across the shunt FET. It will need to be tuned to the parasitics of your load.

    Regards,

    Clint

  • 0 in reply to Clinton Jensen
    Prodigy 20 points
    Hey Clinton!

    So to give you more info,

    Its a 12-33v input with LED stack voltages from 12v to 24v, with systems in place to not allow LEDS with stack voltages more than Vin.

    But for our use, the user has to be able swap the LED around live (with power off), so it must work with no changes to Roff, or the inductors. we are trying to use average values for the calculations, This is why i thought the chip might be a pain.

    We have no output cap, but we have damaged our shunt FETS before in some occasions the spikes are up to 80v and its a 60v FET, we are looking at a 100v or higher FET for the future.

    The user LED can have long cables between the LEDs and the driver so the distance will be variable from 1 foot to ~10 foot.

    I will try the snubber across the nFET again and see if it helps, We may just have to have the analog dimming and EN pin dimming for the lower end instead of analog and FET dimming, I wanted to try FET dimming because we need the LEDS to work at low light and with high speed cameras so I was using a FET dimming frequency of 100khz but i may have to do EN dimming at like 20khz, Any way to make the EN not make audible noise or run at higher frequencies if not any other sugestions .

    Matt
  • 0 in reply to Matthew Santoro
    TI__Mastermind 25500 points

    Hello Matt,

    It shouldn't be a pain for varying loads and input, this device does very well with that. Also a higher voltage FET can do the trick if the spikes aren't a problem otherwise. But I know your issue, I worked on some of the first shunt FET dimming regulators and the spikes can be huge. Think v=Ldi/dt and you can see that with such fast edges it doesn't take much parasitic inductance to induce huge voltage spikes. 1ft to 10ft will be very significant for sure. A snubber can make a huge difference and is worth a try if you need the high frequency.

    EN dimming is another option as well. In that case the parasitics aren't so important. You can either use an output capacitor to compensate for parasitics or you can use no output capacitance and then the parasitics just become an addition to the inductor you are already using. Combination analog and PWM dimming is also an option as you mention. EN dimming can also work at 100kHz but the linearity will not be as good as shunt FET dimming at such a high frequency, but that may not be an issue depending on your particular application. But even audible frequencies can be used with careful capacitor and inductor selection and you will not hear it (assuming it's still fast enough for your high speed camera).

    Regards,

    Clint