Post by thread:Overclocking a PIC16F84

Hi I need to run a PIC16F84 on 4.194304 MHz (2^22 hz) for some precision timing, and since this clock speed is a multiple of 1024, it works extremely well with the Timer0 module on the PIC. Question is, would it hurt to run a 4Mhz PIC at this speed, or should I rather opt for a 10Mhz PIC? Regards, Werner Soekoe Information Systems Manager Free State Legislature spam_OUTwernersTakeThisOuTfsl.gov.za Tel. (051) 407-1109 Fax. (051) 407-1137 Cell. 082 376 8383 Country Code. +27

> > I need to run a PIC16F84 on 4.194304 MHz (2^22 hz) for some precision > timing, and since this clock speed is a multiple of 1024, it works > extremely well with the Timer0 module on the PIC. Question is, would it > hurt to run a 4Mhz PIC at this speed, or should I rather opt for a 10Mhz > PIC? All F84's come off the same wafer and are tested for speed performance. An -04 is one that doesn't meet -10 specs, it may even run at 9.5MHz, but not 10MHz. I've an -04 that's quite happy at 7.3MHz. You can't GUARANTEE that an -04 will run at 4.2MHz but it probably will. If this is a one-off I'd say try it but for a production run you'd be safer using -10 devices, unless you want to test each -04 as it goes into service. Most of them would pass I think, the 4.2MHz being so close to the nominal 4MHz grading. In a sense you aren't overclocking, the -04's just have a possible operating range of 4 to 10 MHz.

Werner Soekoe <.....PICLISTKILLspam@spam@MITVMA.MIT.EDU> wrote: > I need to run a PIC16F84 on 4.194304 MHz .... would it hurt to run > a 4Mhz PIC at this speed, or should I rather opt for a 10Mhz PIC? Werner: All 16F84s are manufactured using the same die and the same process; the only thing that makes a -10 part different from a -4 part is that the -10 part has been TESTED (and passed) at 10 MHz. Microchip doesn't publicize any data on the number of chips that fail at 10 MHz but pass at 4, but the general feeling out here in the world is that the number must be very small. Many of us have played with PICs at high speeds; at room temperature, I got a PIC16C54-XT (the old name for the PIC16C54-04) to run at 45 MHz. It's EXTREMELY unlikely that a -4 part will fail at 4.194304 MHz; in fact, it's unlikely that it'll even exhibit ANY behavior outside of Microchip's published specs for the part. However... If you're going into large-scale production of your product, or if it's important that the PIC meet ALL of its published specs (current draw and temperature-range, particularly), then you should use a -10 part; if you use a -4, you won't have any recourse with Microchip should it fail. -Andy === Andrew Warren - fastfwdKILLspamix.netcom.com === Fast Forward Engineering - San Diego, California === http://www.geocities.com/SiliconValley/2499

Joe Colquitt <.....PICLISTKILLspam.....MITVMA.MIT.EDU> wrote: > All F84's come off the same wafer and are tested for speed > performance. An -04 is one that doesn't meet -10 specs, it may even > run at 9.5MHz, but not 10MHz. Um, no. It's a small disctinction, but an important one: A -10 part DOES meet the -10 specs, but a -04 doesn't necessarily FAIL to meet them. -Andy === Andrew Warren - EraseMEfastfwdspam_OUTTakeThisOuTix.netcom.com === Fast Forward Engineering - San Diego, California === http://www.geocities.com/SiliconValley/2499

Sorry to be pedantic now that the original question has been answered, but having trouble understanding "doesn't necessarily fail". "doesn't necessarily" implies that there is some flexibility with regard to the consigning of parts to the -04 grade. Is there not an absolute limit to exclude parts ? I realise that guaranteed perfomance is important (perhaps my example of 9.5/10 was not a good one) but a part that meets -10 specs should be a -10 part. What reason would there be to reject it ? From: Andrew Warren <fastfwdspam_OUTIX.NETCOM.COM> {Quote hidden}> > > All F84's come off the same wafer and are tested for speed > > performance. An -04 is one that doesn't meet -10 specs, it may even > > run at 9.5MHz, but not 10MHz. > > Um, no. It's a small disctinction, but an important one: A -10 > part DOES meet the -10 specs, but a -04 doesn't necessarily FAIL > to meet them. > > -Andy

On Tue, 2 May 2000, Jinx wrote: > Sorry to be pedantic now that the original question has been answered, but > having trouble understanding "doesn't necessarily fail". > > "doesn't necessarily" implies that there is some flexibility with regard > to the consigning of parts to the -04 grade. > > Is there not an absolute limit to exclude parts ? I realise that guaranteed > perfomance is important (perhaps my example of 9.5/10 was not a good > one) but a part that meets -10 specs should be a -10 part. What reason > would there be to reject it ? Since there's a price difference, there will be a difference in demand. This means you, as a manufacturer, only need X number of 10MHz parts. Once you've tested and verified X (or even 2X) you just don't need any more. If 100% of your stock meets 10MHz specs, and you make them all -10 parts, then there's a shortage of -04 parts. Dale --- The most exciting phrase to hear in science, the one that heralds new discoveries, is not "Eureka!" (I found it!) but "That's funny ..." -- Isaac Asimov

At 07:39 PM 5/2/00 +1200, you wrote: > I need to run a PIC16F84 on 4.194304 MHz (2^22 hz) for some precision > timing, and since this clock speed is a multiple of 1024, it works > extremely well with the Timer0 module on the PIC. Question is, would it > hurt to run a 4Mhz PIC at this speed, or should I rather opt for a 10Mhz > PIC? Hi!, I have a F84 runing at 10 Mhz, it's a -04 part because here in my country is almost impossible to find a -10 part, I've been tested it between 5 - 60 C temperature and it was Ok. It's working actually as a packet transmitter only (it broadcast telemetry, a cheaper version of a pager jeje), and still working 2 months ago, so... don't worry, a couple of mhz more than the specified will not harm it at all!. But keep in mind that what I am doing could be unstable, taking it to the limit, It works and pass my tests but in a serious commercial aplication would be better use -10 part. so IMHO for 2 o 3 Mhz more, -10 part will not be necessary. see you. Leandro J. Laporta (LU2AOQ) mail: @spam@lu2aoqKILLspamyahoo.com wrk: Arg. Assoc. for Space Tech. ham: TCP/IP high speed group HSG __________________________________________________ Do You Yahoo!? Talk to your friends online with Yahoo! Messenger. http://im.yahoo.com

Jinx <KILLspamPICLISTKILLspamMITVMA.MIT.EDU> wrote: > > > All F84's come off the same wafer and are tested for speed > > > performance. An -04 is one that doesn't meet -10 specs, it may > > > even run at 9.5MHz, but not 10MHz. > > > > Um, no. It's a small disctinction, but an important one: A -10 > > part DOES meet the -10 specs, but a -04 doesn't necessarily FAIL > > to meet them. > > a part that meets -10 specs should be a -10 part. What reason would > there be to reject it ? Jinx: The reason's simple: If Microchip tested all parts at 4 and 10 MHz, then labeled each according to the highest-speed spec it met, they'd have warehouses full of unwanted -10s while -04s would be back- ordered for years. -Andy === Andrew Warren - RemoveMEfastfwdTakeThisOuTix.netcom.com === Fast Forward Engineering - San Diego, California === http://www.geocities.com/SiliconValley/2499

Werner Soekoe wrote: > > Hi > > I need to run a PIC16F84 on 4.194304 MHz (2^22 hz) for some precision > timing, and since this clock speed is a multiple of 1024, it works extremely > well with the Timer0 module on the PIC. Question is, would it hurt to run a > 4Mhz PIC at this speed, or should I rather opt for a 10Mhz PIC? > I wouldn't be too fussed about using a 4MHz part. -- Best regards Tony http://www.picnpoke.com TakeThisOuTsalesEraseMEspam_OUTpicnpoke.com

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 At 09:54 PM 5/2/00 +0200, Henrik Nielsen wrote: >Hi >I have 400 pic 16F84 04 running at 8 mhz, non of them has fail >in 6 monts running 24 hour a day >they are running a simple code with text on a video signal no problem When you can get a written variance from microchip to use them at 10, then call me. - -- Are you an ISP? Tired of spam? http://www.spamwhack.com A pre-emptive strike against spam! Where's Dave? http://www.findu.com/cgi-bin/find.cgi?kc6ete-9 -----BEGIN PGP SIGNATURE----- Version: PGPfreeware 6.5.2 for non-commercial use <http://www.pgp.com> iQA/AwUBOQ9+n4FlGDz1l6VWEQKaOACg8e7F7fD5SD8+/XEuYItBaoL26sUAn1hz 9HvefOs35FB6lJFbR/RXHzY/ =mRcs -----END PGP SIGNATURE-----

> Is there not an absolute limit to exclude parts ? I realise that guaranteed > perfomance is important (perhaps my example of 9.5/10 was not a good > one) but a part that meets -10 specs should be a -10 part. What reason > would there be to reject it ? Test cost. A distinction between 10MHz, 4MHz, and "bad chip" requires 2 tests. A distinction between either one of 10 or 4 MHz, versus "bad chip" can be done with only 1 test. A 4MHz part is seldomly a part that has failed a 10MHz test earlier, but most probably is a part that has been tested at 4MHz only - because the 4MHz stock was low and needed refilling. Whether the part runs OK at 10MHz or not is not known. You can screen the parts yourself if you want to select "good" from "bad" chips.

I had a nice discussion about clocks and other specs with a Microchip FAE, and what he said seemed really reasonable. Basically, the specs state what Microchip guarantees and is willing to stand behind. Once you buy the chips, you're free to do whatever you want with them, but if what you're doing is outside of the specs, don't expect Microchip to back you up. Any guarantees you offer will be strictly your own. The kicker with most of the specs is "over the specified temperature range". At room temperature, just about any chip will perform far beyond its specifications. If you are certain your chips will always be at exactly 25C, then you can run them way beyond the specs. This same FAE said he had a customer who was having good luck running some at 10 times the spec'ed clock rate in a certain application. The customer had to screen the parts themselves, and their temp spec was very tight, but it worked. Another customer was dropping PIC's down a borehole where they ran for a few minutes at some ungodly temperature like 300C. Again, they had to screen the parts themselves, but they got a tough job done. > {Original Message removed}