Am I going insane ?

I have opened two packets of different coloured LEDs (Green, Yellow) bought from an RS Online reseller in Dubai and the legs are cut wrongly on both sets.

Very annoying to find out after having soldered 15 of them !

Comparing them with other LEDs and thier internals I can see that the long leg is on the wrong side in both packets of new LEDs.

Duane B

rcarduino.blogspot.com

No, you're not going mad. The leg length isn't a 100% infallible way of telling the polarity. The only sure-fire way is to find the flat edge of the body of the LED, which will be by the cathode.

Incidentally, I have had a batch of 0805 SMD red LEDs from RS which have been marked on the back with an arrow - in the wrong direction. Fortunately it was just one of a number of different makes / models of the same type of LED while I was trying to find the exact ones to use in my LED boards (see my eBay shop), so I just ignored that make and concentrated on the others.

Bummer about the rework. Did you get a good price on them at least?

This is where my perennial failure to remember which lead length corresponds to the cathode (so I always have to test) turns out to be a boon! It also stops one picking the wrong colour when the bodies are clear plastic.

Yep same here, I can never remember so test before using.

I usually just check the actual led junction, the side attached to the larger part of metal is always negative, and the part with the little gold wire is always the positive
works well for smd since I don't need to bother turn it over or test, 5mm leds are even easier because the internal negative metal is alot larger, 1w leds are a bit trickier to see but still possible, but I usually test those for color quality anyway

Very easy to check them with the multimeter continuity test, since they light up.

The only sure-fire way is to find the flat edge of the body of the LED, which will be by the cathode.

I have some LEDs with no flat edge at all...

the larger part of metal is always negative

This is ALMOST always true. IIRC, HP had some early high brightness LEDs that broke this rule because they shaved off most of the substrate in order to allow light to exit both sides, or something.

A cr2016 coin cell makes a good tester...