Author Topic: Reduce Noise ripple form Constant Current power supply. Maybe low pass filter?  (Read 866 times)

0 Members and 1 Guest are viewing this topic.

Offline wrightpt

  • Newbie
  • Posts: 1
  • Country: us

I have the following power supply: HLG-120H-350A

I took it to a local tv technician to measure the noise in the circuit. He said it was around 7.2Khz. Nothing in the circuit can create noise. Just LEDs in the circuit other than the power supply.

I was wondering if a low pass filter would work in dropping the noise coming from the power supply? I have not been able to find anything off the shelf that is really suited for this. I think i may be looking at it wrong.

But wouldn't a low pass filter work in calming the noise? At least in theory.

Offline evb149

  • Super Contributor
  • ***
  • Posts: 1666
  • Country: us
Why is the noise a problem?  What specifically are you using it for, how are you using it, and what is the noise affecting?

Well it is a bit odd that it is a constant current supply intended to power lighting LEDs 1V-10VDC, 350mA, with dimming capability.
Usually you would not "see" the noise visually in the lighting so it would not be a problem.

7kHz is also a somewhat low frequency for noise.  Usually that might be the bandwidth of the control loop of the SMPS but often the majority of SMPS noise is at the 70kHz to 10MHz region depending on the switching frequency ripple, switching spikes and so on but of course the better they eliminate those noise sources in the design the more likely you will be left with mostly lower frequency kHz or XX Hz range noise.

If I was building a circuit from scratch to suppress noise I'd consider things like:
* Low pass filter (as you said)
* Notch filter tuned to suppress the range of the peak noise if it is predictable and narrowband.
* Capacitance multiplier type filter if it would help make the low pass filter more practical / effective.
* Active ripple cancellation regulator, a different approach of noise filtration.
* Common mode choke if the problem noise is common mode and not differential mode.

But all of these are somwhat complex and invasive if your noise problem is truly such low frequency.  Filtering differential noise is harder the lower you go in frequency.  I'm afraid you're past the range of just clamping a split ferrite bead over the line / lines and calling it good if your kHz information is actually correct. 

And you may not want to limit the bandwidth too much below X kHz since the thing is actually supposed to actively respond to and regulate variations in the load like a LED shorting out or transients etc.  Though unless you add too much inductance I guess filtering shouldn't have that bad of an affect on your operation if you're using it for something like the intended lighting application...


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo