[RFI] Noisy LED shop lights

[Dennis Monticelli]

Charles,

I think I said that:  "along with a series resistor to set the current."
The more resistor drop the more current regulation.  The current does not
need to be perfectly constant.  Having said that an LM317HV with a heat
sink (probably need the HV version) is a better solution.

Probably moot anyway.  If direct radiation is the issue I think most hams
would just get rid of the lamp and buy a recommended quiet one.

Dennis AE6C



On Thu, Jun 14, 2018 at 11:58 AM, Charles Gallo <charlie at thegallos.com>
wrote:

> On Thu, June 14, 2018 2:40 pm, Dennis Monticelli wrote:
>
> >
> > This situation cannot be easily corrected. Trying to add current
> > filtering is likely to mess up the feedback loop.  Basically, it's a
> > redesign.  If one goes to the trouble of a redesign then I encouraged
> > ripping out the SMPS board and using a linear wall adapter along with a
> > series resistor to set the current.  Yeah, the efficiency would drop but
> > the result would be absolutely zero noise of any kind because the LED's
> > themselves are dead quiet.
> >
> > Dennis AE6C
>
> The problem with this is that the internal resistance of an LED falls
> dramatically with temperature, particularly on high output LEDs.  You
> either have to design worst case, in which case you end up with a fairly
> dim LED, or risk thermal run away
>
> The proper way to drive an LED _IS_ with the constant current power
> supply, but it does not need to be switch mode.
>
> If you are going to rip out their SMPS, go grab yourself a LM317, and look
> at this article for building a higher current supply, and run it in
> constant current mode, and that will get you up to 1 amp
>
> See the data sheet or
> http://www.bristolwatch.com/ccs/LM317.htm
>
> For more current
>
> http://www.bristolwatch.com/ccs/LM317boost.htm
>
>
>
>
>
>