| We are looking to bring in a HP 5DX System. This is | the laminography equipment formerly known as Four Pi. | I have our sales rep getting us a few references, but I | wanted to get some opinions from the tech net. | I am looking for input from someone whom has worked with | a modern model (last 2 years), and most contacts have horror | stories of the Four Pi equipment from the early 1990s. | From the info available, I am working under these assumptions | for the 5300 series machine: | For complex assembly with double sided actives and 20K solder joints | a program run time of 5-7 minutes for effective coverage. | 10 day programming time for first 2 months, moving down to 6 days after | 6 months, and 4 days after a year. | 100% of Process Engineer and 100% of Programmer for 6 months, and | then 50% Process Engineer and 100% of Programmer continuing. | Obviously these variables are dependant on the number of products, | and depends on how thorough we plan to be, but I wanted to know if | my assumptions are way off! | Steve Abrahamson | Steve - sounds like you've got some humungous boards to test!! To sanity check, We (GPT in the UK) have been using 5DX for 2 years typical board sizes 2500-8000 joints. run times 90 secs to 4 mins for biggest. But this is Series 1 - expect SII 40% faster. We're about to upgrade - I'll tell end July! Programming time - 1.5 days for small board Your boards I would expect 3-4 days max, depending on complexity. But don't try to be too clever - stick to gross defects only - shorts, opens, missing, misplaced misaligned. Time consuming activity is bringing false call rate down - a good target is 500 dpmo for mature process. / program. It's easier to train a process engineer than a board test guy. If you want to discuss further drop me a line. Brian Pollitt Process Eng. Mgr. GPT.
reply »