??? 02/17/11 19:14 Modified: 02/17/11 19:17 Read: times |
#181159 - I'm not convinced that's real reason ... Responding to: ???'s previous message |
He wouldn't be going on about DIP-availability if he knew much about this industry at all. One of the "problems" that has, with the increase of speed, plagued DIP-packaged parts is the length of the leads, particularly power and gnd. Long wires inside the component package can't be improved by clever layout or thoughtful supply bypass. Signals adjacent to power and GND, at today's low power, couple pretty easily, and with the fast edges and large number of synchronous events, both the power and gnd, and the adjacent signals will "hear" the noise on the adjacent leads in the large lead frame. Aside from that, the package and lead frame cost and weight are factors to the large-volume producer, though not necessarily even to the large volume industrial consumer. The latter are merely concerned with the resulting cost.
Since the reduction in package size does benefit the end-product, it makes sense to use the small package in volume. Once you're prototyping, though, you should be using the product that goes in the target, and not something that just happens to fit your eval-kit socket. I got on the AVR bandwagon in the early '90's, and got off just as quickly when I found out about the abysmal ATMEL support. This was, BTW, a decade before the "other" much worse experience I had with their ARM team. I'd be surprised to learn that the O/P has tried some of the components he's "ruled-out" for non-DIL packaging. Every problem I recall he's mentioned is readily addressed by one 805x or another. I doubt it would hurt him to do a bit more reading and study the numbers more closely. RE |