Post by thread:RESET Problems on PICs

Hi, pauls on the list, I just entountered a problem in my recent project. As I am a green hand in PIC application, I'd like to share this with you, and hope I can get some advice. My problem is about the identification of RESET STATUS of PICs at work. If the PIC(e.g. P16c54) has disabled the WDT and doesn't use SLEEP on whole process, after RESET on MCLR, how could I reliably determine whether this RESET is caused by POWER-ON or MCLR? In other words, is there any FLAG BIT that can change its value in different RESETs? Do you have any suggestion? Thanks in advance!

Starfire Zhu wrote: > My problem is about the identification of RESET STATUS of PICs at work. > If the PIC(e.g. P16c54) has disabled the WDT and doesn't use SLEEP on > whole process, after RESET on MCLR, how could I reliably determine > whether this RESET is caused by POWER-ON or MCLR? In other words, > is there any FLAG BIT that can change its value in different RESETs? I get the PICLIST in digest form and I kind of waited to see if anyone else might answer you but since I can't find a reply, here is my two pence. An idea that occurs to me (someone please correct me if I'm wrong) is to write value(s) from the EPROM into RAM variable(s). On a MCLR reset, the values should be unchanged. On a power-up, the RAM locations come up with random contents. Your initialization code can check for the correct values. If they are not present then you know that this is a power-up so you write the values into the RAM locations. Of course, if they are there, then you know you are resetting from a /MCLR. If you use one 8-bit variable, then the chances are 255 to 1 that power-up would produce the same value; two variables = 65535 to 1, etc. I have never tried it, but it makes sense. Michael REMOVE THE .NS (NO SPAM) FROM MY ADDRESS TO REPLY ***************************************** The strong do what they have to do and the weak accept what they have to accept. --Thucydides

> Van: Starfire Zhu <spam_OUTzhuxhTakeThisOuTcheerful.com> > Aan: .....PICLISTKILLspam@spam@MITVMA.MIT.EDU > Onderwerp: RESET Problems on PICs > Datum: woensdag 22 april 1998 19:24 > > Hi, pauls on the list, > > I just entountered a problem in my recent project. As I am a green hand > in PIC application, I'd like to share this with you, and hope I can get > some advice. > > My problem is about the identification of RESET STATUS of PICs at work. > If the PIC(e.g. P16c54) has disabled the WDT and doesn't use SLEEP on > whole process, after RESET on MCLR, how could I reliably determine > whether this RESET is caused by POWER-ON or MCLR? In other words, > is there any FLAG BIT that can change its value in different RESETs? > > Do you have any suggestion? Thanks in advance! Yes, I have. In the documents about the PIC16C5? it say's : Document : DS30015N Page : 4-31 Reset conditions for special registers: Condition Status-register ------------------------------ ---------------------- Power-On reset 0001 1xxx MCLR 000u uuuu MCLR from Sleep 0001 0uuu WDT reset 0000 1uuu WDT from Sleep 0000 0uuu x = Unknown u = Unchanged Hope this helps Greetz, Rudy Wieser P.s. This document is available at the Microchip-site

On Fri, 24 Apr 1998 11:31:18 +0200 Caisson <caissonKILLspamTELEBYTE.NL> writes: {Quote hidden}>> Van: Starfire Zhu <.....zhuxhKILLspam.....cheerful.com> >> Aan: EraseMEPICLISTspam_OUTTakeThisOuTMITVMA.MIT.EDU >> Onderwerp: RESET Problems on PICs >> Datum: woensdag 22 april 1998 19:24 >> >> Hi, pauls on the list, >> >> I just entountered a problem in my recent project. As I am a green >hand >> in PIC application, I'd like to share this with you, and hope I can >get >> some advice. >> >> My problem is about the identification of RESET STATUS of PICs at >work. >> If the PIC(e.g. P16c54) has disabled the WDT and doesn't use SLEEP >on >> whole process, after RESET on MCLR, how could I reliably determine >> whether this RESET is caused by POWER-ON or MCLR? In other words, >> is there any FLAG BIT that can change its value in different RESETs? >Reset conditions for special registers: >Condition Status-register >------------------------------ ---------------------- >Power-On reset 0001 1xxx >MCLR 000u uuuu >MCLR from Sleep 0001 0uuu >WDT reset 0000 1uuu >WDT from Sleep 0000 0uuu > >x = Unknown >u = Unchanged > This is worth studying. However the original poster has also studied it and realized that a MCLR reset leaves the bits unchanged. Unless a WDT timeout or SLEEP has occurred since power-on, the bits will read 11 after a MCLR reset, exactly what they would read after the power-on reset. _____________________________________________________________________ You don't need to buy Internet access to use free Internet e-mail. Get completely free e-mail from Juno at http://www.juno.com Or call Juno at (800) 654-JUNO [654-5866]

On Thu, 23 Apr 1998 21:58:58 -0700 Michael Ghormley <mikegspam_OUTBLACKSAND.NS.COM> writes: >Starfire Zhu wrote: > >> My problem is about the identification of RESET STATUS of PICs at >work. >> If the PIC(e.g. P16c54) has disabled the WDT and doesn't use SLEEP >on >> whole process, after RESET on MCLR, how could I reliably determine >> whether this RESET is caused by POWER-ON or MCLR? In other words, >> is there any FLAG BIT that can change its value in different RESETs? >An idea that occurs to me (someone please correct me if I'm wrong) is >to >write value(s) from the EPROM into RAM variable(s). On a MCLR reset, >the >values should be unchanged. On a power-up, the RAM locations come up >with random contents. This works, but in some cases the "random" RAM data may accidentally match the EPROM data. Using as many bytes as possible for this function makes the possibility more remote. On a particular chip, the RAM will probably power-up with the same value almost every time. Another idea: put a resistor to Vdd and a capacitor to ground on an input pin. If the power has just been turned on, the pin will read "low" until the capacitor charges up. If the pin reads "high" after a reset, it's probably a MCLR reset and the power has been on all along. A diode in parallel with the resistor to quickly discharge the capacitor (and save some stress on the PIC's protection diode) may be a good idea if the power supply voltage can fall quickly. I've never tried it, but the warning in the PIC data sheet about changing the prescaler assignment suggests there is a "loophole" in the PIC logic that allows forcing a WDT reset. It appears that if the prescaler is changed to WDT while it has a 1 in the bit that is driving the WDT logic, the PIC resets. This happens *even if the WDT is disabled in the configuration bits*. I think the process would be to set the prescaler to TMR0, timer mode, clear it, wait until a bit is 1, then set it to WDT. Forcing a WDT reset would clear the TO bit in the STATUS register. Further MCLR resets would not change it, but another power-up will set it back to 1. Microchip may change the silicon so the force WDT reset trick doesn't work (if it even works now) and then your program wouldn't work either. _____________________________________________________________________ You don't need to buy Internet access to use free Internet e-mail. Get completely free e-mail from Juno at http://www.juno.com Or call Juno at (800) 654-JUNO [654-5866]

At 12:16 24/04/98 -0400, you wrote: >Another idea: put a resistor to Vdd and a capacitor to ground on an input >pin. If the power has just been turned on, the pin will read "low" until >the capacitor charges up. If the pin reads "high" after a reset, it's >probably a MCLR reset and the power has been on all along. To change "probably" into definitely use a RESET\ line to discharge the capacitor. Alternatively, (or partly anyway) I use a MAX691A which has a pin to indicate if the external watchdog bit - if it didn't it would have to be a power on reset (unless other resets could be generated - maybe Brown-out resets?). Regards Adrian WWW WWW Adrian Gothard WWW WW WWW White Horse Design WWWWWWWWWW +44-385-970009 (Mobile/SMS), +44-118-962-8913/4 (voice/fax) WWWW WWWW @spam@whdKILLspamzetnet.co.uk, http://www.users.zetnet.co.uk/whd --- Developers of GPS satellite-based tracking systems

> Van: Mike Keitz <KILLspammkeitzKILLspamJUNO.COM> > Aan: RemoveMEPICLISTTakeThisOuTMITVMA.MIT.EDU > Onderwerp: Re: RESET Problems on PICs > Datum: vrijdag 24 april 1998 18:28 > > On Fri, 24 Apr 1998 11:31:18 +0200 Caisson <spamBeGonecaissonspamBeGoneTELEBYTE.NL> writes: [Cut] {Quote hidden}> >Reset conditions for special registers: > >Condition Status-register > >------------------------------ ---------------------- > >Power-On reset 0001 1xxx > >MCLR 000u uuuu > >MCLR from Sleep 0001 0uuu > >WDT reset 0000 1uuu > >WDT from Sleep 0000 0uuu > > > >x = Unknown > >u = Unchanged > > > This is worth studying. However the original poster has also studied it > and realized that a MCLR reset leaves the bits unchanged. Unless a WDT > timeout or SLEEP has occurred since power-on, the bits will read 11 after > a MCLR reset, exactly what they would read after the power-on reset. _Now_ he tell's me ... ;-) You are right ofcourse. It didn't register when I copied it, but now it does. 2 bits that tell you what kind of reset you did have (4 possibilities !) and 5 types of reset .... I'll have to remember that. Greetz, Rudy Wieser