Searching \ for 'chronometer' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/index.htm?key=chronometer
Search entire site for: 'chronometer'.

Truncated match.
PICList Thread
'chronometer'
1998\10\30@114218 by Lucuix Christian

flavicon
face
Hi,
I've got to realize a chronometer with a PIC controller, with a precision
of the centi-second 0.01s .
So I'm looking for a chronometer circuit to interface with .
I'm opened for any idea!
In advance thanks
Christian

Christian Lucuix
Observatoire du Pic du Midi.
BP136 65201 Bagneres de Bigorre cedex.
spam_OUTlucuixTakeThisOuTspamobs-mip.fr
Tel ( 05 62 95 82 17 ).
FAX ( 05 62 95 10 70 ).

1998\10\30@121729 by Peter L. Peres

picon face
On Fri, 30 Oct 1998, Lucuix Christian wrote:

> Hi,
> I've got to realize a chronometer with a PIC controller, with a precision
> of the centi-second 0.01s .
> So I'm looking for a chronometer circuit to interface with .
> I'm opened for any idea!
> In advance thanks
> Christian

Are LEDs ok or do you want to use a LCD and the usual lights ? What is the
trigger ? Start/stop or just one button. Is there a need for a switched
output ?

Using an alphanumeric LCD and a pic and nothing else will get you there on
the fastest way imho, but you need to specify more.

Also, 0.01 sec must be held over how long a time ? A 100 ppm crystal will
drift so it will reach that error in 100 seconds. A 10 ppm crystal will
reach it in 1000 seconds (16 min 40 sec). I think that you want to do
something with a frequency standard, you must have one or two at the Pic
;)

Peter

1998\10\30@124340 by greg

flavicon
face
You might want to considering using  a Dallas Semiconductor TCXO and an I2C RTC
chip.
The TCXO  will provide +/-1Min/yr (about +/- 2 ppm) accuracy (over 0 to 40C)
but will only consume 150uAmps @ 5vdc. Of course it will still have to be
corrected via a sync with a time standard periodically to correct for drift. A
$79 OEM GPS receiver could be used for this purpose.


Peter L. Peres wrote:

{Quote hidden}

1998\10\30@151751 by Peter L. Peres

picon face
On Fri, 30 Oct 1998, Greg Fountain wrote:

> You might want to considering using a Dallas Semiconductor TCXO and an
> I2C RTC chip.  The TCXO will provide +/-1Min/yr (about +/- 2 ppm)
> accuracy (over 0 to 40C)  but will only consume 150uAmps @ 5vdc. Of
> course it will still have to be corrected via a sync with a time
> standard periodically to correct for drift. A $79 OEM GPS receiver could
> be used for this purpose.

Your 2 ppm solution will keep the time within his spec for just 1 1/4
hours at a cost that I don't want to think about. Also, the Pic du Midi is
a pretty tall mountain in the Alps if I remember my geography, and it has
altitude and temperature problems (astronomers are famous for freezing
their bums off in godforsaken places while gazing at stars). I believe
that the GPS solution will also have some (deliberate) problems.

I think that he wants to hook up to an atomic standard, probably over
computer network.

{Quote hidden}

1998\10\30@222215 by Gordon Couger

flavicon
face
Use a gps. A Rockwell Jupiter has a one second and 100 kHz signal that
is super accurate. It also gives the time in ASCII so you can set the clock
if you need to. They cost about $125 US. They will work for you right out
of the box with out any setup.

Gordon

Gordon Couger .....gcougerKILLspamspam@spam@couger.com
Owner PRAG-L PRactical AGriculture List  http://www.couger.com/prag-l
Stillwater, OK        405 624-2855   GMT -6:00
{Original Message removed}


'chronometer'
1998\11\01@113228 by Peter L. Peres
picon face
On Fri, 30 Oct 1998, Gordon Couger wrote:

> Use a gps. A Rockwell Jupiter has a one second and 100 kHz signal that
> is super accurate. It also gives the time in ASCII so you can set the clock
> if you need to. They cost about $125 US. They will work for you right out
> of the box with out any setup.

How accurate is a Jupiter GPS when used in France (temporal accuracy) ?

Peter

1998\11\02@125415 by Lucuix Christian

flavicon
face
At 22:11 30/10/98 +0000, you wrote:
{Quote hidden}

precision
>> > > of the centi-second 0.01s .
>> > > So I'm looking for a chronometer circuit to interface with .
>> > > I'm opened for any idea!
>> > > In advance thanks
>> > > Christian
>> >
>> > Are LEDs ok or do you want to use a LCD and the usual lights ? What is
the
>> > trigger ? Start/stop or just one button. Is there a need for a switched
>> > output ?
>> >
>> > Using an alphanumeric LCD and a pic and nothing else will get you
there on
{Quote hidden}

First of all, I want to thanks all of you for the replies.
I'm a bit confused, because my question was not so precise.
For the display I'm using a Futaba 2x20 characters fluorescent display
because of the working conditions (deep dark under the sky), the user have
a 12 keys keyboard to interact with the chronometer, and I'm using a PIC
16C84 to control it.
The idea is to measure the time between two events (CCD camera shutter
open/closure), so an absolute precision is not needed ( only differential
measure) . What is important is to have a drift as low as possible for a
duration of half an hour.
I first planed to use the PIC himself to implement the chronometer (all
software approach) but the user can interact with the chronometer during
the measurement to act whit the instrumentation, so an all software
approach is not usable. That's why I'm looking for a chip, which is able to
handle the time. If this pretty all done chip is not yet on the earth
(sic), I'm going to realize the timer with discrete circuits (re-sic)
For Peter L. Peres, the Pic du Midi observatory is in the Pyrenees, but
congratulation for your analyze (the top of the mountain is a very colder,
higher, and E.M.C disturbed place, ha ha ha).
Christian.

Christian Lucuix
Observatoire du Pic du Midi.
BP136 65201 Bagneres de Bigorre cedex.
lucuixspamKILLspamobs-mip.fr
Tel ( 05 62 95 82 17 ).
FAX ( 05 62 95 10 70 ).

1998\11\02@153128 by Andy Kunz

flavicon
face
>software approach) but the user can interact with the chronometer during
>the measurement to act whit the instrumentation, so an all software
>approach is not usable. That's why I'm looking for a chip, which is able to

Use interrupts.  Works fine.  I keep time very accurately this way without
a problem, even when I have to disable interrupts for some period (which,
fortunately, is much less than the rollover time for TMR0).

Andy
==================================================================
Andy Kunz - Statistical Research, Inc. - Westfield, New Jersey USA
==================================================================

1998\11\02@170633 by paulb

flavicon
face
Lucuix Christian wrote:

> I first planed to use the PIC himself to implement the chronometer
> (all software approach) but the user can interact with the chronometer
> during the measurement to act with the instrumentation, so an all
> software approach is not usable.

 Ah, you seem to have a very basic misunderstanding of embedded systems
programming here!  It's not just about turning outputs on and off in
sequence, it's *all about* doing two or more tasks simultaneously, such
as keeping time and receiving user input.

Andy Kunz wrote:

> Use interrupts.  Works fine.  I keep time very accurately this way
> without a problem, even when I have to disable interrupts for some
> period (which, fortunately, is much less than the rollover time for
> TMR0).

 .. and this is very true, but even then, it is by no means the *only*
way to do it.  The PIC16C54 and its cohorts didn't (don't) have
interrupts, but people still write beautiful multi-task timekeeping code
on them, using a technique called "isosynchronous" code, and/ or with
the help of the timer/ counter register.

 Isosynchronous code is written so that the decision tree which
implements each function takes *exactly* the same number of instruction
cycles to execute no matter what branches are taken.  Commonly, each
branch returns to the appropriate point on a "tail" which is a chain of
NOPs to delay the necessary amount to compensate for that branch's
execution time.

 Since each routine takes a precise time to execute, a chain of the
various routines to do all the necessary tasks (the "main loop") also
takes a precise time to execute and can be padded to represent the
desired timing unit ("clock tick") such as 1ms, 10ms etc.

 Of course the code which counts these timing units must also take a
precise,consistent time to execute and be included in the loop.

 There are variations for shorter or longer timing resolutions, but we
can explain these as you require.

 And the other way I mentioned (and elaborated only recently!), is to
use the timer/ counter (and prescaler) to generate the timing reference.
The timing loop consists of executing a number of basic functions whose
duration is inconstant, but guaranteed to be much less than the counter
rollover, then waiting (by polling) for the rollover, counting it, and
going back and executing the "foreground" function again etc.

 If the consequence of the rollover ("clock tick") is that a lengthy
procedure is performed, then this is done instead of the "foreground"
for that cycle, and the next rollover waited for and processed.

 As I noted, this is what "life" (embedded programming) is all about!
--
 Cheers,
       Paul B.

1998\11\03@125414 by Peter L. Peres

picon face
On Mon, 2 Nov 1998, Lucuix Christian wrote:

> First of all, I want to thanks all of you for the replies.
> I'm a bit confused, because my question was not so precise.
> For the display I'm using a Futaba 2x20 characters fluorescent display
> because of the working conditions (deep dark under the sky), the user have
> a 12 keys keyboard to interact with the chronometer, and I'm using a PIC
> 16C84 to control it.

A futaba display needs a high voltage driver. Maybe using a backlit LCD or
LEDs is better ?

> The idea is to measure the time between two events (CCD camera shutter
> open/closure), so an absolute precision is not needed ( only differential
> measure) . What is important is to have a drift as low as possible for a
> duration of half an hour.

So you can live with a 5 ppm timebase ;) If you stick a normal crystal in
a small oven and calibrate it it will do that. This means that you can use
a PIC ;)

> I first planed to use the PIC himself to implement the chronometer (all
> software approach) but the user can interact with the chronometer during
> the measurement to act whit the instrumentation, so an all software
> approach is not usable. That's why I'm looking for a chip, which is able to

Why not ? There are interrupts, constant run-time segments and other ways
to do it all in software.

> handle the time. If this pretty all done chip is not yet on the earth
> (sic), I'm going to realize the timer with discrete circuits (re-sic)

Aiee ;) Actually, for a 5 ppm timebase you should concentrate on as FEW
components as possible, at least in the oscillator. Otherwise behavior
over temperature will require a doctorate thesis to be compensated
somewhat imho.

> For Peter L. Peres, the Pic du Midi observatory is in the Pyrenees, but
> congratulation for your analyze (the top of the mountain is a very colder,
> higher, and E.M.C disturbed place, ha ha ha).
> Christian.

Ok I did not get my geo.

Peter

More... (looser matching)
- Last day of these posts
- In 1998 , 1999 only
- Today
- New search...