Post by thread:[PIC:] Cycle time on a 12F675 with 4Mhz internal R

I have been playing with Microchip's PICkit 1 and the 12F675 8-Pin PIC. I have been using the internal RC oscillator, as it keeps the number of external parts down, which keeps the playtime with the PIC up. I have downloaded and read the documentation on the device and am confused by trying to figure out what the clock cycle time is. What I think I know; -The documentation says that the minimum cycle time is 200nS which is a 5MHz Clock -a 20MHz crystal used an external oscillator, would be divided by 4, in fact any external oscillator would be divided by 4 -the internal RC oscillator runs at ~4MHz, within 1% accuracy when calibrated. -the internal oscillator is not divided by 4, thus instruction cycle time is 250nS Is this correct? I understand some instructions may take 2 cycles to run. Loops I have made appear to support that the internal RC oscillator is not divided by 4, but the documentation is not clear on this. Thanks Andrew -- http://www.piclist.com hint: To leave the PICList spam_OUTpiclist-unsubscribe-requestTakeThisOuTmitvma.mit.edu

On Thu, 27 Nov 2003 10:56:33 -0500, you wrote: >I have been playing with Microchip's PICkit 1 and the 12F675 8-Pin PIC. I have been using the internal RC oscillator, as it keeps the number of external parts down, which keeps the playtime with the PIC up. I have downloaded and read the documentation on the device and am confused by trying to figure out what the clock cycle time is. > >What I think I know; > >-The documentation says that the minimum cycle time is 200nS which is a 5MHz Clock > -a 20MHz crystal used an external oscillator, would be divided by 4, in fact any external oscillator would be divided by 4 >-the internal RC oscillator runs at ~4MHz, within 1% accuracy when calibrated. > -the internal oscillator is not divided by 4, thus instruction cycle time is 250nS > >Is this correct? No, all oscillators are divided by 4. Internal 4MHz RC will give 1uS per instruction. -- http://www.piclist.com hint: To leave the PICList .....piclist-unsubscribe-requestKILLspam@spam@mitvma.mit.edu

> Is this correct? I understand some instructions may take 2 > cycles to run. Loops I have made appear to support that the > internal RC oscillator is not divided by 4, but the > documentation is not clear on this. The internal clock is handled the same way as an external clock, so the PIC will run at 1 MIPS (except for PC-changing intsructions). Wouter van Ooijen -- ------------------------------------------- Van Ooijen Technische Informatica: http://www.voti.nl consultancy, development, PICmicro products -- http://www.piclist.com hint: To leave the PICList piclist-unsubscribe-requestKILLspammitvma.mit.edu

At 10:56 AM 11/27/2003 -0500, you wrote: >I have been playing with Microchip's PICkit 1 and the 12F675 8-Pin PIC. I >have been using the internal RC oscillator, as it keeps the number of >external parts down, which keeps the playtime with the PIC up. I have >downloaded and read the documentation on the device and am confused by >trying to figure out what the clock cycle time is. > >What I think I know; > >-The documentation says that the minimum cycle time is 200nS which is a >5MHz Clock > -a 20MHz crystal used an external oscillator, would be divided by 4, > in fact any external oscillator would be divided by 4 >-the internal RC oscillator runs at ~4MHz, within 1% accuracy when calibrated. > -the internal oscillator is not divided by 4, thus instruction cycle > time is 250nS > >Is this correct? I understand some instructions may take 2 cycles to >run. Loops I have made appear to support that the internal RC oscillator >is not divided by 4, but the documentation is not clear on this. It's divided by 4, so 1us Tcy using the internal RC clock on the aforementioned part. Where the data sheet seems to offer insufficient information, you should be checking the midrange reference manual. Usually it's there. Best regards, Spehro Pefhany --"it's the network..." "The Journey is the reward" .....speffKILLspam.....interlog.com Info for manufacturers: http://www.trexon.com Embedded software/hardware/analog Info for designers: http://www.speff.com -- http://www.piclist.com hint: To leave the PICList EraseMEpiclist-unsubscribe-requestspam_OUTTakeThisOuTmitvma.mit.edu

Andrew Herdman wrote: > -The documentation says that the minimum cycle time is 200nS which is > a 5MHz Clock Yes, if you mean 5MHz instruction clock. > -a 20MHz crystal used an external oscillator, would be divided by > 4, in fact any external oscillator would be divided by 4 Yes. > -the internal RC oscillator runs at ~4MHz, within 1% accuracy when > calibrated. Yes. > -the internal oscillator is not divided by 4, thus instruction > cycle time is 250nS No. All oscillators are divided by 4 because 4 clock phases are required by the hardware to execute one instruction. The internal 4MHz oscillator causes 1MHz instruction rate, which is 1uS per instruction. ***************************************************************** Embed Inc, embedded system specialists in Littleton Massachusetts (978) 742-9014, http://www.embedinc.com -- http://www.piclist.com hint: PICList Posts must start with ONE topic: [PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads

On 28 Nov 2003 at 8:08, Olin Lathrop wrote: <snip> > > -the internal RC oscillator runs at ~4MHz, within 1% accuracy when > > calibrated. > > Yes. > That's interesting. I saw that 1% spec and later couldn't find it. I tried running internal Osc and measured 1.5MHz CLK out which is 6MHz OSC in. Now either my DVM freq counter is way off (kinda doubt it, but I stuck a XTAL in b4 breaking out the real freq counter) or the internal Osc is off. What is this calibration? I couldn't find anything on that, just curves over temp and Vdd, IIRC. BRs, Mike <snip> -- http://www.piclist.com hint: PICList Posts must start with ONE topic: [PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads

> That's interesting. I saw that 1% spec and later couldn't find it. I > tried running internal Osc and measured 1.5MHz CLK out which is 6MHz > OSC in. > > Now either my DVM freq counter is way off (kinda doubt it, but I > stuck a XTAL in b4 breaking out the real freq counter) or the > internal Osc is off. What is this calibration? I couldn't find > anything on that, just curves over temp and Vdd, IIRC. Some of the newer PICs do have about 1% or 2% accuracy in their internal oscillator. I don't remember for sure if the 12F629 is one of them, but I thought so. I was confirming the general way the internal oscillator worked, and didn't really pay much attention to the accuracy spec. ***************************************************************** Embed Inc, embedded system specialists in Littleton Massachusetts (978) 742-9014, http://www.embedinc.com -- http://www.piclist.com hint: PICList Posts must start with ONE topic: [PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads

Olin Lathrop wrote : > Some of the newer PICs do have about 1% or 2% accuracy in their internal > oscillator. I don't remember for sure if the 12F629 is one of them, but I > thought so. MC claim 1% *if* your programmer setup takes care of saving and restoring the RETLW XX instruction in the last location of the program memory. *And*, of course, that you include a CALL to this instruction and saves the returned value in OSCCAL in your startup part of your application. Here is a major difference between this kind of "calibration" and the method the newer parts (as the 18F1220) uses. There you have the specific accuracy with a POR value (h'00') in OSCCAL (or OSCTUNE as it's called there). So no need of any RETLW or anyhing else... Jan-Erik. -- http://www.piclist.com hint: PICList Posts must start with ONE topic: [PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads

The 12F675 uses a calibration byte in the PIC to adjust the oscillator. The calibration is done at the factory. When you (re-)program the chip, you always have to read the calibration byte first (because it will be erased), then program the chip and then write back the calibration value. See for more information: DS41173B A good programmer will do that automatic, and you won't be aware of. If you loose the calibration value, Microchip has somewhere a program to recalibrate the chip. The 1% accuracy is only at a Vdd=3.5 V and a stable temperature of 25 Celcius. A more realistic value spec is +/- 2% at 2.5V ... 5.5V and 0..85 Celcius ! See for more information p95 of DS41190C. Stef Mientki Andrew Herdman wrote: {Quote hidden}>I have been playing with Microchip's PICkit 1 and the 12F675 8-Pin PIC. I have been using the internal RC oscillator, as it keeps the number of external parts down, which keeps the playtime with the PIC up. I have downloaded and read the documentation on the device and am confused by trying to figure out what the clock cycle time is. > >What I think I know; > >-The documentation says that the minimum cycle time is 200nS which is a 5MHz Clock > -a 20MHz crystal used an external oscillator, would be divided by 4, in fact any external oscillator would be divided by 4 >-the internal RC oscillator runs at ~4MHz, within 1% accuracy when calibrated. > -the internal oscillator is not divided by 4, thus instruction cycle time is 250nS > >Is this correct? I understand some instructions may take 2 cycles to run. Loops I have made appear to support that the internal RC oscillator is not divided by 4, but the documentation is not clear on this. > >Thanks > Andrew > >-- >http://www.piclist.com hint: To leave the PICList >piclist-unsubscribe-requestspam_OUTmitvma.mit.edu > > > -- http://www.piclist.com hint: PICList Posts must start with ONE topic: [PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads

Jan-Erik Soderholm XA (TN/PAC) wrote: >Stef Mientki wrote: > > > >>The 12F675 uses a calibration byte in the PIC to adjust the oscillator. >> >> > >It's actualy a 14-bit program memory *word*, not a byte... > >No big deal, maybe :-) > Yes, you are completely right ! Bad ..., this is my third error today in a newsgroup :-( So let me try to improve myself: Only the 6 most significant bits of the 14 bit program memory word, contribute to the oscillator calibration information ;-) Stef Mientki > >Jan-Erik. > >-- >http://www.piclist.com hint: PICList Posts must start with ONE topic: >[PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads > > > -- http://www.piclist.com hint: PICList Posts must start with ONE topic: [PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads

At 11:38 AM 11/28/2003 -0500, you wrote: >Now either my DVM freq counter is way off (kinda doubt it, but I stuck a >XTAL in b4 >breaking out the real freq counter) or the internal Osc is off. What is >this calibration? I >couldn't find anything on that, just curves over temp and Vdd, IIRC. You have to load the OSCCAL register. It's done in those parts by calling the last location in memory, IIRC (which had better not have been erased). Before you do that, the oscillator runs uncalibrated. Best regards, Spehro Pefhany --"it's the network..." "The Journey is the reward" @spam@speffKILLspaminterlog.com Info for manufacturers: http://www.trexon.com Embedded software/hardware/analog Info for designers: http://www.speff.com -- http://www.piclist.com hint: PICList Posts must start with ONE topic: [PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads

Spehro Pefhany wrote: > You have to load the OSCCAL register. It's done in those parts by calling > the last location in memory, IIRC... Correct, but not enough, you must also do a MOVWF OSCCAL after the CALL. The CALL only will not do any difference by itself. > (which had better not have been erased). It is erased in each programming cycle, but it has to be pre-read and re-written by the programming tools. Jan-Erik. -- http://www.piclist.com hint: PICList Posts must start with ONE topic: [PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads

Ahem... > Only the 6 most significant bits of the 14 bit program memory word, > contribute to the oscillator calibration information ;-) Only the 6 most significant bits of the *8 least significant bits of the* 14 bit program memory word, :-) Jan-Erik. -- http://www.piclist.com hint: PICList Posts must start with ONE topic: [PIC]:,[SX]:,[AVR]: ->uP ONLY! [EE]:,[OT]: ->Other [BUY]:,[AD]: ->Ads

> > The 12F675 uses a calibration byte in the PIC to adjust the > oscillator. > > It's actualy a 14-bit program memory *word*, not a byte... But that 14-bit word contains only 8 bits of data, and IIRC only a few bits out of that are actually used for calibration. Wouter van Ooijen -- ------------------------------------------- Van Ooijen Technische Informatica: http://www.voti.nl consultancy, development, PICmicro products -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics

Jan-Erik Soderholm XA (TN/PAC) wrote: > Stef Mientki wrote: > >> The 12F675 uses a calibration byte in the PIC to adjust the >> oscillator. > > It's actualy a 14-bit program memory *word*, not a byte... You are both right. The last word of program memory contains a RETLW instruction. This takes up a whole program memory word, but the data is one 8 bit byte. Of those 8 bits, only the high 6 are actually used in the current implementation. The low 2 bits written to the OSCCAL register are ignored. ***************************************************************** Embed Inc, embedded system specialists in Littleton Massachusetts (978) 742-9014, http://www.embedinc.com -- http://www.piclist.com hint: The list server can filter out subtopics (like ads or off topics) for you. See http://www.piclist.com/#topics