Soldering the LITEON SMD LED

I continue to not understand this, and I hope someone can sort me out. This is the SMD Red/Green LED used in the Amber board and a few others.

The Amber board follows the recommended pad layout shown in the datasheet. If you look at the dimensions, you can see that the C2 copper extends across the bottom to where the C1 copper is. If you place the component across the recommended pads, both C1 and C2 will sit on the “bottom” pad. I’ve been able to only solder C1 to the pad using lots of flux and very very careful soldering, but how is it done in an oven? Don’t both pads grab solder? What am I missing about this? Why would C2 extend down to C1, and or why wouldn’t the pad on the PCB be split?

I’m currently creating this device in Eagle and I would really rather create a package layout that made it much more easy to solder C1 by itself. I must be missing something. Are you supposed to place the device only half onto the pads?

Oh, and if anyone knows of this device in an existing Eagle library, please let me know.

I couldn’t easily make sense of the datasheet either. Perhaps you could google for an image of the part and link it in? What does the actual part look like?

Generally there is soldermask on the PCB between pads that inhibits the solder flowing from one place to another. For components where the gaps are too small another way is to leave the copper pads to simply stand as little islands off the board and surface tension of the solder tends to keep it on the top of the pad and out of the valley.

This is just general information; I know it doesn’t totally answer your question.

The package layout on page 7 of the linked datasheet sure seems reasonable. there are three individual square pads with ample space between them.

no, I wouldn’t guess that the part does anything except get placed down straight onto the pads.

Here is an image from Digikey. Their link to the datasheet is broken, so you have to refer to Mouser.

I’ve hand soldered using two methods: Either slide the device out so that C2 isn’t over the same pad as C1, or use flux paste so that it sits up off the pad, and then very carefully tap just the C1 edge of the device with your iron, after soldering the other side of the device.

I’m considering making the pads like this in Eagle. Someone stop me.

Hi Joel,

Man, that sure is one hecka hard part to solder, I completely agree. Most of the hand-built ambers we’ve got use the through-hole components for the opto, and I tend to use top-fire LEDs and phototransistors in my designs ever since soldering the first one of these.

It can certainly be done in an oven, though - I think your confusion is just over what parts of the device are meant to be the actual solder pad for each of the terminals. Here’s a couple of notes to help you along:

  1. The recommended footprint for the device is included in the datasheet you sent. It’s on page 7. This matches the footprint on the amber. This footprint keeps the pads as far away from each other as possible and provides the best self-orienting action in an SMT oven.

  2. The side-by-side footprint you’re cooking up there MIGHT work, if you keep the pads “far enough” apart and have solder mask between them, but you’re definitely in danger of just getting solder bridges across the two terminals.

  3. The datasheet for the Lite-on part does another evil thing: it labels the cathodes “C1” and “C2”. These should really be “K1” and “K2” to avoid confusion with capacitor designators in a design.

  4. If you’re refactoring this design for yourself in Eagle, I’d recommend switching to top-fire components. I’ve had good luck hand-soldering this part, and used it on a number of my designs:

Hope this helps!


Thanks Tom. I’ve had a lot of success with the side-fire LED, as well as the PT, they both seem to also work nicely from the top. I might go with the other LED, but I guess I still want to figure out what deal is the original one. It seems to me that a better pad layout would be like this:

Edit — I think it should be like this, rather…

On the bottom of the device, the C2 pad is an"L" shape-ish, running from the center of the bottom down the left side. (from the perspective of the datasheet). On the recommended footprint, copper on the device from both C1 and C2 would sit over the C1 pad, right? I guess I’m asking, but I know that they do. if you sit the component directly on the recommended footprint, C1 and C2 are making connections to the C1 pad.

Sorry for going over and over this, but it seems that there is either a mistake in the recommended footprint, or I am totally missing some key piece of pad design.

@jwehr, we used this part:

Easier footprint and I hand soldered it no problem.

Yeah, I created the part that Tom posted in Eagle, which is nearly identical to the one you posted, and that is what I’ll be using. I’m still a bit frustrated by the recommended footprint for the other one. If someone has a close-up pic of the cathode solder joints from a reflow oven, I’d like to see it. I don’t see how it is possible to not short the cathodes on one of the pads. I’ve tried to contact about a half dozen of the LITEON offices in the US, and never got a person, or a return call. Doesn’t really make me want to use their parts. Oh well. Thanks for the tip!