[TriEmbed] A little help
mcckevin at mcclaning.com
Wed Jul 29 11:03:28 CDT 2020
I use Brian's method all the time when driving LEDs or electromagnets.
The current limit feature on most power supplies is often overlooked,
but very useful.
On 7/29/20 10:05 AM, Brian via TriEmbed wrote:
> Some tips and educated guesses:
> This would be a great job for an adjustable benchtop power supply.
> Set the current limit to something relatively small, say, 100 mA, and
> start bringing up the voltage until the current limit is hit. Many
> high-power LEDs like to run at 750 mA, so 100 mA won't be nearly
> enough power dissipation to worry about overheating them just in the
> process of finding out the voltage and polarity. You should find a
> point where current starts increasing rapidly with small increases in
> voltage, and that'll get you in the ballpark. Then you can set your
> current limit to 750 mA and turn the voltage all the way up; whatever
> voltage ends up across the LED at 750 mA is the rated voltage. As
> others have said, do NOT run these at full power for more than an
> instant without proper heat sinking.
> Note that LEDs, being diodes, won't conduct at all until the potential
> across them exceeds the junction voltage, and that these modules are
> often individual chips wired in series. If you see 9 chips in there,
> my guess is that the operational voltage is probably somewhere around
> 12 V (1.something volts for each junction, which is very reasonable).
> If you wanted to take a 12-V power supply and just quickly tap the
> leads one way and then the other, you probably wouldn't toast the
> module (but you do have more than one, right? ;-) ). You could still
> put a few hundred ohms in series as ballast if you wanted to be careful.
> Finally, the "tray" that the silicon sits in is always the cathode,
> but it looks like you might not be able to see through the cover. It
> might be a reasonable guess to assume the chassis is the cathode (one
> lead would be clearly isolated from the rest of the chassis, while the
> other is not).
> I take it there are no identifying markings on the device at all? No
> numbers, codes, anything?
> Triangle, NC Embedded Computing mailing list
> To post message: TriEmbed at triembed.org
> List info:
> TriEmbed web site: http://TriEmbed.org
> To unsubscribe, click link and send a blank message:
> mailto:unsubscribe-TriEmbed at bitser.net?subject=unsubscribe
More information about the TriEmbed