On 11/19/2011 12:36 PM, gene heskett wrote: > On Saturday, November 19, 2011 12:32:48 PM Dave did opine: > > >> On 11/18/2011 11:46 AM, andy pugh wrote: >> >>> On 18 November 2011 16:31, gene heskett<[email protected]> wrote: >>> >>>> An interesting question, and not that I know of. The problem is >>>> likely one of overflow in the emc internal counter, >>>> >>> The accumulator is 64 bits. At 3000rpm and 16x microstepping I make >>> that 3 million years before counter roll-over. >>> >>> I am unconcerned, that's not my bug to fix :-) >>> >> It is interesting how things like this, when translated into minutes, >> hours and then years oftentimes quickly exceeds our lifespan. >> >> I ran into the same thing last year on a one way - indexed chain drive >> on a machine. You might remember that. I first thought I had to deal >> with a 32 bit counter that would roll over every hour or so.. then I >> was told it is a 64 bit counter and the time required to roll over >> immediately exceeded the 1 year warranty period on the machine, buy a >> few hundred years! :-) >> >> The machine has been running 16 hours per day since June of 2010, and so >> far no rollover issues. :-) They have had to change the index drive >> gearboxes on the machine twice since then.... >> > Would that not be prima faci evidence that a redesign is in order, or do > they just consider it a CODB? How about chain life? At a reposition every > 3 seconds, there has to be accumulative shock affecting it. > > >> but that cost is >> incidental as they push the heck out of the machine to increase >> throughput and make more $. They are doing 24-26 indexes per minute.. >> which is about twice the original machine design rate. They love that >> machine. >> >> Dave >> >> >> ------------------------------------------------------------------------ >> ------ All the data continuously generated in your IT infrastructure >> contains a definitive record of customers, application performance, >> security threats, fraudulent activity, and more. Splunk takes this >> data and makes sense of it. IT sense. And common sense. >> http://p.sf.net/sfu/splunk-novd2d >> _______________________________________________ >> Emc-users mailing list >> [email protected] >> https://lists.sourceforge.net/lists/listinfo/emc-users >> > > Cheers, Gene >
>>Would that not be prima faci evidence that a redesign is in order Yes, but they are too busy making product to bother with a redesign and resulting modifications. They consider it a CODB. And it is clearly their choice. They have another machine about 100 feet away that is newer but it chugs along at only about 15 transfers per minute. The gear boxes last quite a bit longer but the machine puts out only 62 % of the output. There are a number of machines that feed this machine and the staffing is the same whether they run 15 or 24 transfers per minute. The only significant incremental cost is wear and tear on the machine. A gear box is about $1200 plus 1-2 hours to install it. So they just keep a few on the shelf. The last time I was at that plant back in July, I modified the software and servo setup so when they jam up the machine, they can recover in a shorter period of time - they might save 30 seconds or so to clear a jam. If things are running well something jams up about every 20 minutes or so due to product variations which they can't control. The production people were thrilled! The chain is very sturdy so it seems to last for years. Dave ------------------------------------------------------------------------------ All the data continuously generated in your IT infrastructure contains a definitive record of customers, application performance, security threats, fraudulent activity, and more. Splunk takes this data and makes sense of it. IT sense. And common sense. http://p.sf.net/sfu/splunk-novd2d _______________________________________________ Emc-users mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/emc-users
