12v/24vdc 9-LED spotlight open/troubleshoot

Wurkkos

Dave_H

Enlightened
Joined
Nov 3, 2009
Messages
593
Location
Ottawa Ont. Canada
Interesting experience with a 12v/24v 9-LED spotlight picked up at HD for 75% off ($7.50).Last one
on the shelf, had been opened but decided to take it.

https://www.homedepot.ca/product/briteled-utilityled-9/1001491688

Powering up using variable supply, 3 of 9 LEDs did not light, and the voltage-current behaviour for
light of this type did not look right. Rather than return it, decision was to open and troubleshoot. As these
are well sealed and not designed to be be easily opened up, it likely would have been scrapped, so it
was saved from that.

On PCB inside was found a single IC driver (no markings) connected to 3 paralleled sets of 3 series
LEDs. No sign of current balancing so design must rely on close matching of the LEDs (batch, bin) and
that they all run at approximately same temperature to keep current through each string roughly
equal.

Problem with this is any LED/string which fails open, same current forced through remaining strings
will make LEDs run brighter/hotter for a short time anyway, but likely leads to early failure.

Using diode check function on DMM, each LED was probed with power off. DMM drove enough
voltage/current to dimly light each good LED. One bad LED in the dead string was found. I was hoping
it was just a bad solder joint, or LED soldered in backwards.

Here's a shot of the PCB with LEDs; one at "8-o'clock" position is bad:



Replacing the bad LED means finding exact match for part number, and binning; not feasible for
average person. Only hope would be a junker light of the same type for parts.

If one driver per string were used (some designs do), bad LED could be bypassed and would
have working light with slightly lower output i.e. 8 of 9 LEDs. No such luck in this case.

Getting two strings working properly (6 of 9 LEDs) required total current reduced to 2/3, which
meant raising sense resistor by 3/2 (50%). These resistors have low values, usually easy to locate on
the PCB. In this case it was 0.125 ohms consisting of 0.2 ohms and 0.33 ohms in parallel. By luck,
removing the latter, one SMT part, did the trick; probably no accident as the same circuit could be
used for a 6-LED design with no value changes, just component (de-)population.

Lamp is working at reduced current but also with lower brightness and irregular beam pattern.
I will probably replace the plastic focusing lens (slightly damaged when removed) with clear
plastic to make it more of a flood.

Dave
 
Last edited:

Dave_H

Enlightened
Joined
Nov 3, 2009
Messages
593
Location
Ottawa Ont. Canada
2Q==
 
Top