--- In [email protected], "apluscw" <[EMAIL PROTECTED]> wrote: > > > At the end of the day both KiCAD and PCB pool came up trumps. Look in > > the photos section for a shot of some detail I took while > constructing > > the prototype. > > Ah, an LPC man.
Not really, more of a "used more micros than I care to remember man". Actually, thinking back to my first uP experiences, perhaps I should slip into my tie die T-shirt and say "used more micros than I care to remember, man". For the record, the choice of the LPC was heavily influenced by the need for 2 UARTS and an unusually large amount of RAM. There was a time when I used to enjoy stringing out busses with RAMS, EPROMS, 373's for the IO and making complex GALS with bank switching logic. But frankly nowdays if there is a micro that will stick everything in one chip I will jump at the chance, even if it means doing the IO on SPI as it is in this design. Nice micros. You probably knew this, but there is a > yahoo forum for that family. Did you put a JTAG header on the board? > The LPC2000 group? Yes, I used it a lot a couple of years ago when I did my forst LPC project. Then Yahoo started giving me pains (as has happened with me on this group) and I stopped frequenting it. > Did you solder the micro by hand? I am looking to do some SMT boards, > but I want to stick with packages that can be hand-soldered > reasonably. The lead pitch on that package is a bit tight. > It was hand soldered. It takes me 2 to 3 minutes to solder a component like that, so even for small to medium production (given that the board is essentially PTH) it is not worth having it done outside. BTW, SMD production equipment will generally get worse quality than that. SMD process seems to need a bit of a ramp up for each new board before normal quality is reached. By contrast soldering by hand needs an operator ramp up, that is you need to practice **a lot** before you perfect your techniques, but once you have mastered the art you can get it right first time on any board. > If I may ask, why did you mix the SMT ARM with through-hole > components? Ease of solderability? > The board is a PTH board, it is full of connectors, fuseholders, power devices etc. It would not make sense to do it in SMD. The LPC is SMD because there is no PTH version, but there was no reason to put SMD caps. Also, decoupling supply lines with PTH caps is also an excellent way of passing current from one side of the board to the other ;-) > Our board design here at work is very mixed. We are leaning towards > surface mount when possible, Why? (There are a lot of good reasons for using SMD, but they are not always valid). > > Glad that PCB Pool had no trouble with the angle of the micro. Since > the mask process is pretty much automated, it was really a question of > whether or not there process would hande it, I think. It is not automated, a CAM operator uses sophisticated photoplot manipulation tools to set up the panels, and they will get back to you if they see problems. There will also help you resolve problems at the photoplot level. They also have a gerber manipulation service where they will carry out small mods at a gerber level, but this type of service is normally used by producers who need to e.g. change a footprint on an established board. If you are the original designer and have the project you are much better off modding at the EDA level. Why did you > rotate the LPC? Would that be an issue with pick and place machines > were you to go to production? It is a traditional 2 layer board with a classic vertical/horizontal routing mesh. Quad packages integrate into such a mesh much more neatly if they are at 45 degrees. Cheers, Roger.
