Paul Koning wrote: I just want PLCC packaging and 5 volt logic. I keep looking at all the new FPGA logic thinking that would be a nice chip and I don't like the packaging or voltage.Also with all the new chips you get more pins, but more of them are power. Also all my designs are nice and slow LS speeds.That is not to say I may not go faster but right now slow I/O devices are the factor for me. PS. BG. Micro also has some nice heat sinks just in case you need them for a linear power supply,

Paul Koning wrote: I am liking the small CPLD's more and more mainly that you don't have to configure them on power up. I use Atmel since I can get the devopment kit and CPLD's from Digi-Key. It is about 120 clb's each for the 2 chips making up the data path and 90? for the unfinished control unit. I allways wonder what the real clock speed is, after marketing is done with it. :) Ok , you can snicker at me ... I am using lowest speed err lowest cost chips, but then I just plan to run at 6800 speeds. PS. If somebody could sell me a low cost high speed paper tape punch/reader I might even do a PDP ?? clone to stay on topic.

At 11:43 AM 10/26/05 -0400, Paul wrote: I keep telling him that but he's stubborn! Actually that's BS! The military and NASA both REQUIRE them. They're a lot more resistant to vibration for one thing. Besides I didn't say MANUFACTURERS, I said "individuals". MOST manufacturers are only concerned with how cheap they can make the stuff, they don't give a hoot about durability, reliability, repairability or anything else. They are electrically I seriously doubt that. Joe

Joe> At 11:43 AM 10/26/05 -0400, Paul wrote: >> >> >> Of course manufacturers don't want to use DIPs. Joe> Actually that's BS! The military and NASA both REQUIRE Joe> them. They're a lot more resistant to vibration for one thing. Joe> Besides I didn't say MANUFACTURERS, I said "individuals". MOST Joe> manufacturers are only concerned with how cheap they can make Joe> the stuff, they don't give a hoot about durability, reliability, Joe> repairability or anything else. None of that makes any sense. Almost all manufacturers care about reliability, because if they don't, they go out of business, as well they should. A possible exception is cheap junky consumer stuff. Certainly IC manufacturers care about reliability, because much of what they sell goes into applications where the finished product is required to be reliable. As for NASA and the military requiring DIPs, I would find that very surprising. A lot of COTS parts don't exist in DIP form at all, since DIPs don't go much over 40 pins. Also, microwave stuff is universally surface mount; it has to be. Joe> They are electrically Joe> I seriously doubt that. Doubt away, but if you examine lead inductance and capacitance issues for high clock rate circuits, you will find that it makes a very significant difference. Complex ASICs often have 50-100 power and ground wires. Those are extra package I/Os that cost money -- they aren't there for fun, but because they are necessary to meet the performance goals of the device. Power lead inductance is the reason why. paul

der> Why? Is it just the bulk lead capacitance, or is there some der> other way in which a DIP package is electrically deficient? Lead capacitance, lead inductance. And you rarely can use more than one or two power/ground leads. Modern high speed devices often have dozens of power and ground connections to keep the power clean. I suspect SMTs may be thermally better, too -- same reason, shorter lead length from die to PCB. der> Bigger, yes. Slower, I can't see why, unless it's bulk der> capacitance. RoHS compliant, that's a historical accident der> rather than anything inherent in DIPs, isn't it? I can't see der> any reason why a DIP couldn't be made as cleanly and compliantly der> as the corresponding surface-mount part. Agreed. But I don't expect any lead-free DIPs. RoHS is forcing manufacturers to look at every part they sell, and decide either to go through the expense of updating it, or to drop it as obsolete or too low volume to be worth the trouble. paul

Paul Koning wrote: I get a kick out of the back-to-the-future aspect of SMT packages, considering the lineage/similarity of SMT to the flatpak IC cases (TO-85,91,etc.) of the early '60s, which predated DIPs. Across the history of ICs, DIP packaging begins to look like a 25-to-30-year sidebar.