Searching \ for 'RTC Clock' in subject line. ()
Help us get a faster server
FAQ page: www.piclist.com/techref/timers.htm?key=clock
Search entire site for: 'RTC Clock'.

Truncated match.
'RTC Clock'
1996\05\15@091445 by

I'm redesigning an Automotive 7-day Programmable Real Time Clock. The
current design uses a RTC chip with 32K768 crystal. I'd like to use the
PIC's clock as the time base and would need to use a variable cap to 'pull'
the crystal frequency to achieve the 2mins/year accuracy.

I figure that since the PICs use a parallel resonant crystal it should be
easy to overcome the crystal's initial tolerance of +/-30ppm. I'm planning
on using a 4Mhz crystal but may use a 4M194 unit if it proves too difficult
to divide down accurately.

As well, the PIC (a 16C622 or 16C84) will multiplex a 40 LED array (8X5) by
sourcing 8 segment lines to 5 displays using 1:5 timing. The source current
will be 15mA. I'd like to be sinking that kind of current (to reduce
temperature rise on the PIC), but a large number of displays need to be used
up. Ambient temperature may be as high as 105 Deg.C.

Is this paractical?  Any caveats?

Regards, Dana Frank Raymond
dfricom.ca
> I'm redesigning an Automotive 7-day Programmable Real Time Clock.
> I'd like to use the PIC's clock as the time base and would need to
> use a variable cap to 'pull' the crystal frequency to achieve the
> 2mins/year accuracy.
>
> I figure that since the PICs use a parallel resonant crystal it should
> be easy to overcome the crystal's initial tolerance of +/-30ppm.

Why pull the frequency at all?  If the crystal has the accuracy that
you have stated (+/- 30 ppm), then shouldn't it give you +/- 1.5 min
per year "out of the box"?  Computation as follows:

365 days/yr * 24 hr/day * 60 min/hr * 60 sec/min = 3,153,600 sec/yr

3153600 sec/yr *  +/- 30 ppm =  +/- 94 sec/year

Have I missed something?
Lee Jones

-------------------------------------------------------------------
Jones Computer Communications             leefrumble.claremont.edu
509 Black Hills Dr, Claremont, CA 91711         voice: 909-621-9008
-------------------------------------------------------------------
> > I figure that since the PICs use a parallel resonant crystal it should
> > be easy to overcome the crystal's initial tolerance of +/-30ppm.
>
> Why pull the frequency at all?  If the crystal has the accuracy that
> you have stated (+/- 30 ppm), then shouldn't it give you +/- 1.5 min
> per year "out of the box"?  Computation as follows:
>
>     365 days/yr * 24 hr/day * 60 min/hr * 60 sec/min = 3,153,600 sec/yr
>
>     3153600 sec/yr *  +/- 30 ppm =  +/- 94 sec/year
>
> Have I missed something?

Yup - you missed a zero - 31,153,600 sec/yr.
+/- 940 seconds a year

Also - add 2-4 ppm aging, and 5-50ppm TempCo and your error budget
starts growing rather large....

pesky decimal places
-mike
>Why pull the frequency at all?  If the crystal has the accuracy that
>you have stated (+/- 30 ppm), then shouldn't it give you +/- 1.5 min
>per year "out of the box"?  Computation as follows:
>
>    365 days/yr * 24 hr/day * 60 min/hr * 60 sec/min = 3,153,600 sec/yr
>
>    3153600 sec/yr *  +/- 30 ppm =  +/- 94 sec/year
>
>Have I missed something?

Yes, a trailing edge zero! :-)
The actual error would be 940 sec/year.
Regards, Dana Frank Raymond
dfricom.ca
To: mfahrionbb-elec.com
From: dfricom.ca (Dana Frank Raymond)
Subject: Re: RTC Clock

>Ouch! 2 minutes per year is pretty tight!  Gives you an error budget

Yes, that is tight. However, I have now been told that 6min/year is
acceptable. Thats still +/-11.4ppm however.

> If you can maintain a constant ambient temp this

No chance of that. Its automotive after all! sPEC IS -40 TO +80 Deg.C (which
is the industrial temp range).

>best crystal (coldweld package) you can get is going to be 5ppm over

Thanks for your suggestion... I'll look into cold welded crystals.
Regards, Dana Frank Raymond
dfricom.ca
> Yes, that is tight. However, I have now been told that 6min/year is
> acceptable. Thats still +/-11.4ppm however.
>
> > If you can maintain a constant ambient temp this
>
> No chance of that. Its automotive after all! sPEC IS -40 TO +80 Deg.C (which
> is the industrial temp range).
>
> >best crystal (coldweld package) you can get is going to be 5ppm over
>
> Thanks for your suggestion... I'll look into cold welded crystals.

You are likely going to have to use some type of temp stabiliser to
get the accuracy you require.  If you can spare the energy then you
could use a heater (classic crystal oven principal) to hold the
occilator at a temp of between 60 and 80 Deg.C with a simple
controller.  Insulation would then reduce fluctuations and if would
not spend a lot of time away from 60 C.

As mentioned you will have to get an aged crystal that has lost the
youthfull flutters and assuming this is a consumer applications
it might be better to reset the clock on every tune up from a radio
or GPS standard and use a regular crystal.

The accuracy of the time software can be made exact with some thought
and leaves you only with clock drift.

Cheers
--
Kalle Pihlajasaari     kalledata.co.za
Interface Products     Box 15775, Doornfontein, 2028, South Africa
+27 (11) 402-7750      Fax: +27 (11) 402-7751
> You are likely going to have to use some type of temp stabiliser to
> get the accuracy you require.  If you can spare the energy then you
> could use a heater (classic crystal oven principal) to hold the
> occilator at a temp of between 60 and 80 Deg.C with a simple
> controller.  Insulation would then reduce fluctuations and if would
> not spend a lot of time away from 60 C.

One approach I've hard of is to use a crystal together with an RC oscillator
and EEPROM table.  After the unit is assembled, a test fixture operates it
at different temperatures, measures the exact frequency of the crystal
oscillator, and measures the frequency ratio between the RC oscillator and
the crystal; these measurements are stored in EEPROM.

Later, in the field, the CPU can measure the RC oscillator versus the
crystal, ascertain the current temperature from this measurement, and then
use that to correct for any crystal frequency error.

For this approach to work correctly, it is imperative that the RC oscillator
and crystal oscillator be close enough together that temperature shifts will
affect them both similarly.  If not, a temperature shift that causes the RC
to run slow but not the crystal may have undue effects on the net frequency
output.

More... (looser matching)
- Last day of these posts
- In 1996 , 1997 only
- Today
- New search...