Post by thread:TMR0 Questions

Hi all, I recently received an email from one of the spanish speaking members of the PICLIST who had a few questions about TMR0. He asked me if I would either answer them, or translate his email and send it to the list. I tried to answer them as best I could, but I figured I ought to send a copy to the list as well to see if there are any further suggestions. The original sender (Jose) and I would appreciate it if you would take a look at the questions. What follows is the original question and my response translated into english: ~~~~~ ORIGINAL QUESTION ~~~~~ >Date: Fri, 04 Sep 1998 10:21:37 +0200 >X-PH: spam_OUTV4.1TakeThisOuTcornell.edu (Cornell Modified) >From: Jose Antonio Gracia <.....jgraneKILLspam@spam@colon.net> >X-Mailer: Mozilla 4.03 [es] (Win95; I) >To: "Sean Breheny (Piclist EE.UU esp)" <shb7KILLspamCORNELL.EDU> >Subject: Peticion de ayuda > [SOME SNIPPED] I am now starting to program pics, and I have a series of doubts that I would like to resolve. If you know the answer, good. If you don't, fine, but if you would, please transfer this to the list translated to english. In Spain, I found two books about this topic, en them there appears a line which says: "when TMR0 is set by loading it with a value, two clock cycles are lost for synchronization." If I want to make a clock with a 1 MHz crystal, the frequency of instruction execution is 250,000 per second. To obtain a frequency of 1 tenth of a second (sic), I should adjust the prescaler to 5(count up to 250) a later add a counter to count up to 100. The question is: Does one need to load TMR0 with 4 every two times to compensate for the two lost cycles? If ones adds (maybe he meant changes) the prescaler too are two cycles lost? Does the prescaler get reset (when changing TMR0, I presume)? Can you suggest to me more resources (about PICs)? I wrote to the authors of those books and they told me that they don't know! [CLOSING SNIPPED] > Jose Antonio Gracia > >.....jgraneKILLspam.....colon.net >http://www.geocities.com/SiliconValley/Pines/7097 ~~~~~ MY RESPONSE ~~~~~ I don't think that the prescaler can be set to 5. The possible values are 1,2,4,8,16,32,64,128,256. These values come from the bits in the OPTION register. If you have a prescaler of 1, count up to 250 in TMR0, and later have a counter (software postscaler) to count up to 100, then you have a frequency of 10 Hz. For a frequency of less than 10 Hz, there are many posibilities. You can use a crystal of a frequency less than 1 MHz. Or, if you need the main chip clock to be 1 MHz, you can use a second crystal in an oscillator feeding TMR0 directly. There are 32kHz watch crystals made especially for slow clock rates. Now to the specific questions: Yes, you should load TMR0 with 2+value to compensate for the lost cycles(if they matter in your application). For example, to clear TMR0, load it with 2. I don't know if you loose any cycles if the prescaler is changed. I would guess not. I will ask the PICLIST. Yes, if you write to TMR0, the internal prescaler value gets reset, but not the division factor.In other words, you don't need to reload the OPTION register when you change TMR0, but the prescaler/TMR0 system gets reset. Well, you know better than I about books and web sites in spanish about PICs. You ought to make an effort to read the PIC microcontroller datasheets from Microchip. [ENDING SNIPPED] +--------------------------------+ | Sean Breheny | | Amateur Radio Callsign: KA3YXM | | Electrical Engineering Student | +--------------------------------+ Save lives, please look at http://www.all.org Personal page: http://www.people.cornell.edu/pages/shb7 EraseMEshb7spam_OUTTakeThisOuTcornell.edu Phone(USA): (607) 253-0315 ICQ #: 3329174

Sean Breheny wrote: {Quote hidden}> Hi all, > > [SOME SNIPPED] > > I am now starting to program pics, and I have a series of doubts that I > would like to resolve. If you know the answer, good. If you don't, fine, > but if you would, please transfer this to the list translated to english. > > In Spain, I found two books about this topic, en them there appears a line > which says: > "when TMR0 is set by loading it with a value, two clock cycles are lost for > synchronization." > > If I want to make a clock with a 1 MHz crystal, the frequency of > instruction execution is 250,000 per second. To obtain a frequency of 1 > tenth of a second (sic), I should adjust the prescaler to 5(count up to > 250) a later add a counter to count up to 100. > > The question is: > > Does one need to load TMR0 with 4 every two times to compensate for the two > lost cycles? > If ones adds (maybe he meant changes) the prescaler too are two cycles lost? > Does the prescaler get reset (when changing TMR0, I presume)? > Can you suggest to me more resources (about PICs)? > > I wrote to the authors of those books and they told me that they don't know! > > [CLOSING SNIPPED] > > > Jose Antonio Gracia > > > >jgranespam_OUTcolon.net > >http://www.geocities.com/SiliconValley/Pines/7097 > My stategy would be to use a /16 prescaler. You now have 15625 counts to handle. If you set the TMR0 to 132 on every overrun, you get another division of 125 (255-125+2). Count 125 of these overruns. Take care to set the TMR0 = 132 immediately after overrun. If you can't, add the extra program cycles to TMR0. An easier approach could be to let the TMR0 overrun 61 times per second. This gives you a 'second' of 0.999424s. A 'minute' becomes 59.96544s. By periodically increasing the overrun count, you can adjust by 1/(250000/(16*256)) = 0.016384s per extra overrun. -- Friendly Regards Tjaart van der Walt @spam@tjaartKILLspamwasp.co.za |--------------------------------------------------| | WASP International | |R&D Engineer : GSM peripheral services development| |--------------------------------------------------| |SMS KILLspamtjaartKILLspamsms.wasp.co.za (160 chars max)| | http://www.wasp.co.za/~tjaart/index.html | |Voice: +27-(0)11-622-8686 Fax: +27-(0)11-622-8973| | WGS-84 : 26¡10.52'S 28¡06.19'E | |--------------------------------------------------|