Results from my testing the LEDs in the kit before assembly

[D R]

I hate tracking down problems especially when they involve removing a soldered-in component.

I had a little time and got out my admittedly non-research grade component tester/identifier as one of the builders did here earlier.

I went through all the LEDs.

Almost all tested perfectly, with a correct brightness and perceived color temperature as best my memory from one to the next would recall.

Two had a sort of OK test.  The voltage was, for most, 1.92 to 1.96, but two of them tested at 2.09 or 2.07 v.  With extras in the kit, I set those aside even though they did emit light and looked OK otherwise.

Two of them did NOT light, and were identified as a resister, see the picture labeled LEDtest2.jpg

The most curious thing was the one other that while it lit, gave me this very odd identification as a dual diode pack.  That got set aside also; see picture LEDtest1.jpg

Just my two cents worth, but I'm glad to have done it and hopefully will more than make up for pulled hair and attempts to carefully unsolder.

Dale

[Monty Brandenberg]

On 6/1/2024 2:13 PM, D R wrote:

> The most curious thing was the one other that while it lit, gave me
> this very odd identification as a dual diode pack.  That got set aside
> also; see picture LEDtest1.jpg

Possibly a bi-color LED tossed into the bin or a manufacturing defect
causing low-voltage avalanching?

For *really* cheap test equipment, there is:
https://www.amazon.com/HiLetgo-Tester-2-150mA-Light-Emitting-Battery/dp/B07VHCBSY3/ref=sr_1_9
These are useful for vetting LEDs in inexpensive kits (learned from LEDs
with cathode/anode switched) but don't trust the current markings.  The
test points are simply resistor-limited and the markings based on an
assumed forward voltage (probably red).

[oscarv]

I've come to the conclusion that Chinese wholesalers have a real problem with the quality of yellow LEDs at the moment. 

For a long time, I was doubtful about Bad LED reports, but we tested 3,000 last week. In a few bags of 1,000 LEDs, we found up to almost 1% of bad LEDs. That's roughly one per kit!

So last week we went on a mad hunt for this problem, I think by now I can say a bad production run of yellow LEDs is rippling through the China wholesalers chain. It's not just one vendor, lots of them start sending bags of LEDs with (typically, 0.5%) bad uns in them. 

We now identified the good LED bags to use, but it seems that for the next few months, the headache of testing 10% of LEDs per bag will be upon us.

(And it is not that we use dodgy vendors either, it seems to hit the major China wholesalers)

In the previous batch of kits, there should be plenty of spare LEDs to insulate builders from this problem. But if not, then please do let us know!

Kind regards,

Oscar.