Post by thread:[PIC]: Microchip's PIC18F87J10

At 03:48 PM 4/26/2005 -0400, you wrote: >Spehro Pefhany wrote: >>Same for those cool new PIC18F87J10 series micros. > >Check the endurance spec on the flash. Yes, I noticed that. Doesn't bother me much (who reprograms a production product more than once or twice?), but it might affect some applications. It seems to be coming along with other dual-voltage devices (ST-7), but the LPC210x ARM has 100,000 cycles min endurance. Best regards, Spehro Pefhany --"it's the network..." "The Journey is the reward" spam_OUTspeffTakeThisOuTinterlog.com Info for manufacturers: http://www.trexon.com Embedded software/hardware/analog Info for designers: http://www.speff.com

Spehro Pefhany wrote: >> Check the endurance spec on the flash. > > Yes, I noticed that. Doesn't bother me much (who reprograms a > production product more than once or twice?), but it might affect > some applications. I guess I said that wrong. I meant check the retention lifetime. ***************************************************************** Embed Inc, embedded system specialists in Littleton Massachusetts (978) 742-9014, http://www.embedinc.com

> I guess I said that wrong. I meant check the retention lifetime. For all we information sponges, if there's something interesting happening here it would be nice to hear the actual data. As you know the figure and consider it notable it would take no more effort to quote it (even approximately) than to just talk about it. I'm not complaining at all - just noting that one may not go and look up something like this for a chip one wasn't expecting to use any time soon (or any time) but having extra information adds to the mental information base that may be useful in future. I appreciate that nobody is obliged to provide chapter and verse on everything they post, but it's useful if it's done if it takes no more effort to do so. Russell McMahon

Russell McMahon wrote: > For all we information sponges, if there's something interesting > happening here it would be nice to hear the actual data. As you know > the figure and consider it notable it would take no more effort to > quote it (even approximately) than to just talk about it. If you really must know, I didn't give any numbers because this is something I'm not at liberty to divulge unless it's already been published by Microchip. I'm too lazy to check the latest docs to see what exactly Microchip is saying and what they're not saying. So pointing people to the docs was a safe way to point out something without violating any confidence. ***************************************************************** Embed Inc, embedded system specialists in Littleton Massachusetts (978) 742-9014, http://www.embedinc.com

At 06:38 PM 4/26/2005 -0400, you wrote: >Spehro Pefhany wrote: > >> Check the endurance spec on the flash. > > > > Yes, I noticed that. Doesn't bother me much (who reprograms a > > production product more than once or twice?), but it might affect > > some applications. > >I guess I said that wrong. I meant check the retention lifetime. Ah so.. 20 years typical at 25°C does seem inadequate. Thanks for the heads-up. Best regards, Spehro Pefhany --"it's the network..." "The Journey is the reward" .....speffKILLspam@spam@interlog.com Info for manufacturers: http://www.trexon.com Embedded software/hardware/analog Info for designers: http://www.speff.com

>> >> Check the endurance spec on the flash. >> I guess I said that wrong. I meant check the retention lifetime. > Ah so.. 20 years typical at 25°C does seem inadequate. Looking at the datasheet (39663A.pdf) for the PIC18F87J10 Family, there is a table: TABLE A-1: NOTABLE DIFFERENCES BETWEEN PIC18F8722 AND PIC18F87J10 FAMILIES wherein is listed Characteristic PIC18F87J10 Family PIC18F8722 Family Program Memory Retention 10 years (minimum) 40 years (minimum) Program Memory Endurance PIC18F87J10 Family PIC18F8722 Family 1,000 write/erase cycles (typical) 100,000 write/erase cycles (typical) I've seen the "J" in the partnumber in the recent past but do not know what the meaning of it is. Is this explained anywhere on the Microchip website? MD

At 07:53 PM 4/26/2005 -0400, you wrote: >Yech: an endurance spec of 100 is not very good for a soldered in >prototype under development. > >Lets see..... At an average of 3 burn cycles per hour, 8 hours per day, >you could be dead meat in about four days. > >Bob Ammerman >RAm Systems Not ideal, but it's a minimum figure. Typical (probably with compromised reliability, so I wouldn't ship such a unit out the door) is 1,000 cycles. Which ought to be enough for even the most avid burn-and-crash programmer. Best regards, Spehro Pefhany --"it's the network..." "The Journey is the reward" speffKILLspaminterlog.com Info for manufacturers: http://www.trexon.com Embedded software/hardware/analog Info for designers: http://www.speff.com

Marcel Duchamp wrote: > Looking at the datasheet (39663A.pdf) for the PIC18F87J10 Family, there > is a table: > > Characteristic PIC18F87J10 Family PIC18F8722 Family > Program Memory Retention 10 years (minimum) 40 years (minimum) This is what I was trying to point out, but wasn't sure what Microchip had released about it yet. This is a major difference from previous parts, and I think 10 year lifetime is a serious limitation. That would have been unsuitable for the vast majority of my projects. > I've seen the "J" in the partnumber in the recent past but do not know > what the meaning of it is. I believe that indicates a particular fab process (1/4 micron?). These chips all seem to have the characteristic of low flash retention and that they only run at 3.3V. ***************************************************************** Embed Inc, embedded system specialists in Littleton Massachusetts (978) 742-9014, http://www.embedinc.com

On 4/27/05, Olin Lathrop <.....olin_piclistKILLspam.....embedinc.com> wrote: > Marcel Duchamp wrote: > > Looking at the datasheet (39663A.pdf) for the PIC18F87J10 Family, there > > is a table: > > > > Characteristic PIC18F87J10 Family PIC18F8722 Family > > Program Memory Retention 10 years (minimum) 40 years (minimum) > > This is what I was trying to point out, but wasn't sure what Microchip had > released about it yet. This is a major difference from previous parts, and > I think 10 year lifetime is a serious limitation. That would have been > unsuitable for the vast majority of my projects. Is this really even the case or is it simply that the datasheet is preliminary and they do not know the answer yet? A lot of the stuff in there is marked TBD. I would expect that they will find through their accelerated life tests that the flash retention is in-line with other parts. Does anyone (does Microchip?) really make an flash with only 10 year retention? We hear stories on the list about people using "10 year" EEPROMs for 25 years now. > > I've seen the "J" in the partnumber in the recent past but do not know > > what the meaning of it is. > > I believe that indicates a particular fab process (1/4 micron?). These > chips all seem to have the characteristic of low flash retention and that > they only run at 3.3V. Is this an issue with the process or is it just a new process to Microchip and they don't want to make claims that don't stand the test of time? Bradley

Bradley Ferguson wrote: > I would expect that they will find through > their accelerated life tests that the flash retention is in-line with > other parts. But they are differnt processes and are made in different fabs. ***************************************************************** Embed Inc, embedded system specialists in Littleton Massachusetts (978) 742-9014, http://www.embedinc.com

At 06:58 PM 4/27/2005 -0400, you wrote: >Bradley Ferguson wrote: > > I would expect that they will find through > > their accelerated life tests that the flash retention is in-line with > > other parts. > >But they are differnt processes and are made in different fabs. Olin-- any idea what the *typical* supply current is in the ENC28J60 units you've seen? Best regards, Spehro Pefhany --"it's the network..." "The Journey is the reward" EraseMEspeffspam_OUTTakeThisOuTinterlog.com Info for manufacturers: http://www.trexon.com Embedded software/hardware/analog Info for designers: http://www.speff.com

Spehro Pefhany wrote: > Olin-- any idea what the *typical* supply current is in the ENC28J60 > units you've seen? The first ones took nearly 1W. The latest batch seem to be a little better, but I haven't actually measured them yet. I've got a board with a ENC28J60 and a 30F4012 running at max speed. Both are in the same 28 pin SOIC package, and both seem to get about as warm. ***************************************************************** Embed Inc, embedded system specialists in Littleton Massachusetts (978) 742-9014, http://www.embedinc.com