Not doing well! I have no trouble normally with SMD, either with solder paste or with solder wire. I use a heat gun or the solder tip as appropriate. However, these LEDs have me foxed.
I started with solder paste. A smear on each pad and application of the hot air. Before I saw the solder start to flow, the edges of the LEDs had gone brown and crispy.
Next experiment was with hand soldering but I couldn't get the heat in where it was needed and again the amount of heat needed burnt the cases again. They were not quite as bad as the picture but they got burnt again when I removed them because I had the idents on the board wrong!
A proper hot-air rework station has precisely controlled air temperature so you can set it between the
melting point of solder and of the plastic. Still such LEDs are about the worst case for SMT soldering.
You don't have much hope without accurate temperature control, but that's always advised for any
soldering, be it iron, oven or hot air.
The datasheet for the LEDs will give a thermal profile for soldering.
The hand soldering managed to get some LEDs fitted for testing and my design failed miserably, so a re-design is in order. I failed to calculate the heat dissipation and have something like about 7W going on from the resistors alone. I deliberately left it for a few minutes and it got so hot that all the resistors started dropping off the board with molten solder!
I have a new 858d clone hot air station with the worst instructions I have ever seen. I wonder if someone would be good enough to write a short text on selecting temperature, airflow rate and general technique etc please?
2200watt seems like a cooking hotplate, which is useless if you ask me.
The one I linked to is 800watt. More than enough power, flat top, and it has precise temp control.
I only wish it could go lower (temp), for some drying jobs.
What is there to know about a hot air wand.
Temp setting, and airflow. And different nozzle sizes for large or small parts/areas to solder.
The trick is to pre-heat the board from underneath slowly and long enough. As said, I use about 160C.
Then use the hot air to flow the solder paste.
Experiment on some old board with smd parts.
Try to get them off cleanly, without burning anything.
Hot air set to 220C won't burn quickly. 240C does.
Leo..
The hand soldering managed to get some LEDs fitted for testing and my design failed miserably, so a re-design is in order. I failed to calculate the heat dissipation and have something like about 7W going on from the resistors alone.
What value resistors and how much current?
What is the supply voltage?
I deliberately left it for a few minutes and it got so hot that all the resistors started dropping off the board with molten solder!
That sounds more like a short than a miscalculation.
How much current was the PCB consuming?
The board was designed for 5050 LEDs, these are bigger and I imagine that the segments shorted together and of course pulled three times the current through a single resistor.
TomGeorge:
Hi,
What value resistors and how much current?
The three sections of the LEDs were running, all independently fed from 42 off 470R 1206 size (1/4W) calculated to give around 25mA per segment.
TomGeorge:
What is the supply voltage?
13.8V
TomGeorge:
That sounds more like a short than a miscalculation.
I think that you must be right, but what threw me is that I had full illumination of all the segments and no glowing PCB tracks.
TomGeorge:
How much current was the PCB consuming?
I didn't measure it but the design is marginal to say the least. The resistors are all running excess dissipation regardless. It's a shame about the wasted boards but I have revised and simplified the design now, just designing the new PCBs. I am down to 8 resistors running just under 100mW each. Far better design but it is going to be tedious reworking the PCB layout and waiting for them to be made.
AJLElectronics:
The board was designed for 5050 LEDs, these are bigger and I imagine that the segments shorted together and of course pulled three times the current through a single resistor.
Problem with SMD pads like that, you can't see the important bit that needs the solder because it under the component.
BUGGER!!!!!!!!!!!!
Tom.....