The one I remembered clean and clear is the 18F6680/8680, the new CAN module have two CAN status bit swapted, none of my friends who is working on that project worked it out until the FEA from Boston called us and told us the silicon errata thing. They really shouldnot have released the silicon until at least some basic test are done by their own team. I was bit by cannot enter CAN config mode after sleep(more than 95% of random system crash). Finally, I implemented a sw-reset to overcome this silicon issue on PIC18F458. I am not sure the problem got fixed or not. At least I didn't see it on the Silicon erratas. Funny N. New Bedford, MA http://www.AuElectronics.selfip.com ----- Original Message ---- From: Robin Abbott To: Microcontroller discussion list - Public. Sent: Tuesday, November 6, 2007 2:35:54 PM Subject: [PIC] Outrageous erratas - was PIC32MX Whilst we are at it I would like to highlight the two worst errata from Microchip one of which cost us a significant amount of money. Th 18F1x20 18 pin devices originally would skip random instructions if used with a clock greater than 4MHz. Not every device, sometimes temperature dependent. Our system designed for a 20MHz clock worked fine in prototype (we had some good devices) but in production we got horrendous reject rates. Worst you couldn't even select on test as some devices would work fine and then fail after a week. I always thought Mchip should have ceased production until they had sorted that issue. We had to redesign for a 4MHz clock which in our system had hardware and software implications. I am never sure as to the legal status of Microchip over that one either, although I expect the purchasing conditions give them a let out. The other was the 18F4550 family USB SIE inserting 4 corrupt bytes in certain timing conditions. I think one user who was flag bearing that one had huge trouble convincing the company to look at it. It caused me huge trouble creating a USB library for our compiler. Now I remember back to the 90's the 16C74 USART originally wouldn't work with BRGH set. I am sure there are others - those 3 bit me :( Robin Abbott Forest Electronics - Home of WIZ-C ANSI C Compiler for PIC's with RAD Front end robin.abbott@fored.co.uk www.fored.co.uk -----Original Message----- From: piclist-bounces@mit.edu [mailto:piclist-bounces@mit.edu] On Behalf Of Morgan Olsson Sent: 06 November 2007 13:33 To: Microcontroller discussion list - Public. Subject: Re: [PIC]: PIC32MX > and they have had almost no errata When there is NO official errata it always scare me. I have had stupid system problems because of simple SPI state errors in converters from both Linear Tech and Maxim, neither had any errata. MAxim told me after two nonth they have some undocumented test state tha ti happened to trigger, and Linear tech later came out with a new fixed revision though they said there was no problem in the first. Also Microchip does this: early PIC14000 had a A/D clock select swithc that was broken never expressed ina an errata, na dnow we bumped inte PIC18FxxJ borking the stack sometimes if an asynchronous hight priority interrupt happens the cycle after a low pri trigged (not expressed in any errata) I have also been workign wiht PLC systems bumping inte bugs. And compilers of course. Even th elatest QuickBasic bit me bad once jumping bad in some circumstances. I like the way ST once sent me a written letter explaining the nes ST6 was delayed (i had been on an seminar but not ordered anyting). But some years later i got a couple hundred mosfets from them, that did not stand the power ( I had recieved several hundreds before that worked, but suddenly a batch was bad) And in that case also the dealer (Farnell) seem not to have gone out to withdras the rest of the batch, resulting in many thousands products failing prematurely in the market. I have no problem that things are not perfectly made. The really bad screw-ups happen in next steps WHENEVER MANUFACTURERS DO NOT TELL ABOUT ANY PROBLEMS. So for a while we buy no more power semiconductors from ST, buy from Farnell when only really needed, try to avoid Maxim (at least al their digital stuff), etc. We aso had the failing capacitor thread here recently... There really is great need for a independant organisation to keep track of things in electronics, and we would have things working much more reliable. Then companies could compete much more fair, and not by hiding problams and making nice-looking presentations of their dreams... -- Morgan Olsson -- http://www.piclist.com PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist -- http://www.piclist.com PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com -- http://www.piclist.com PIC/SX FAQ & list archive View/change your membership options at http://mailman.mit.edu/mailman/listinfo/piclist