Wow, apart from the TMP112 everything listed here is in QFN or BGA package. If its really a requirement for the challenge this will be fun for everybody. Or will there be evaluation boards available (the LMP910x0 ones look interesting)?
I think solder paste will be mandatory as one of the component purchases. Either that, or use evaluation boards. Will be interesting to see what methods people use to solder these. michaelkellett has mentioned using a hot plate to solder these. Another method I've heard is a hot air gun, but that sounds more difficult to get right without toasting the chip, but I could be wrong.
Flux would be another good thing to purchase (and tweezers, and a magnifier and very fine solder, and desoldering braid, and spare chips to experiment with) and then they may be hand-solderable with a fine tipped soldering iron (some youtube videos show a few procedures), if the pads are created with space to place the soldering iron tip right up against the edge of the ic and exposed copper on the pcb.
I've got my boards ( a bit later than I hoped) with the 0.5mm pitch BGAs. I've so far soldered 2 of the 64 pin parts down (STM32F100RBH6B). One I broke off to look at the quality of the joints - the other is working as far as I can test it without completing the board.
The tools required for this were: microscope, sticky flux, tweezers, thermocouple based thermometer, Kapton tape and the domestic hotplate I posted a picture of in the other thread.
If any one is interested in actually doing this I'll try to find time to post some pictures.
(BTW the biggest expense so far is the boards with 6 layers and 0.075mm track and gap rules - about £400 min order quantity. I have a feeling that it is possible to layout a board for the Lattice 132 pin 0.5mm BGA using OSH Park's 0.125mm process because Lattice leave some rows unpopulated which makes 'escape' tracks much easier.)
Here's a picture of the processor on the board, I won't have time to do some step by step video or pictures for a couple of weeks.
The chip is a bit hidden by the test connector. You can see the sparse pinning on the two Lattice 132 pin footprints. The pads are about 0.27mm diameter.
Thanks for finding this. I've just ordered one, it's a good price It might need some sort of jig made to position the PCB over the hot spot, but I can manage that.
It's so low cost it's worth the attempt anyway, even if I can't get used to this method. I I have some thermocouples so I can measure the temperature as Michael mentioned,
and also ordered a few of the the cheapest QFN parts with a center pad too, to practice with.
Now I'm really looking forward to trying this out.
It comes with two magnetic board holders, so save the looking.
I haven't used mine, but the person I suggested it to (he got tired of waiting for a cheap chinese version), said its brilliant for desoldering parts, since you preheat from the bottom, and a bit more on top, and its off.
Yes it's a very cheap tool to add to your collection at NZ $93.
I think this preheater can be used as hot plate soldering for smd parts as it support 350C temperature... With this preheater We need solder paste and stencils for different smd footprints....
I have one of these and it's quite useful but it heats up much too fast for reliable soldering of parts using solder paste. The temperature control is very rough and ready. If you are trying to use it for re-flowing paste then initially set the temperature quite low - (check up the paste spec) to drive off the volatile stuff. Then turn the temperature up to the reflow temperature. Using a thermocouple to monitor temperate is a really good idea if you can. I get better results with the big iron hotplates because they heat up more slowly. If you don't drive the volatile stuff out of the paste it can end up boiling away under the chip for ages and you get voids or even unsoldered pins.
Top Comments