Long precise interval

[John Ashurst]

I wish to generate a delay of 29sec, pulse of 1 sec and repeat to quartz clock accuracy. 18x plus Dallas clock chip and separate crystal seems a bit over the top for this application. Can anyone suggest a simpler solution. I am new to PIC tecnology

 

[Dippy]

If you search back in the recent month or two you will find a thread by Hippy about using the PIC Timer.

If you are new to PICAXEs/PICs programming you may go a little cross-eyed at the code, but you'll get used to that after a while.

There will be a tiny time error but you will have to check that. I can't remember the thread name but I'm sure hippy can tell you when he emerges (he doesn't sleep).

 

[hippy]

I doubt you could get "quartz crystal accuracy" with a PICAXE, even one running from an external resonator/crystal or using a Dallas or other time-keeping chip.

If you want something near micro-second accurate you'll either need a much faster executing processor or will have to do it in hardware. Of course a quartz clock is not necessarily 'accurate' as its crystal, due to varying mechanical or processing delays.

How accurate do you actually want it ?

[added] Yes, I'm wake - It's actually daylight as well

Edited by - hippy on 13/02/2007 15:29:23

 

[Dippy]

Excuse the potential 'Red Herring' but whilst I was looking in Farnell for something else I came across this:
<A href='http://www.farnell.com/datasheets/85678.pdf' Target=_Blank>External Web Link</a>

It may no be good enough for your application (its just a posh 555), but it may be a useful chip in its various pinouts.

 

[KMoffett]

What part of the cycle has to be accurate?
The 30-seconds, the 29-seconds, the one second, the periods, the intervals, or seconds per day? Have you ever noticed how accurate those cheap, battery-operated, wall or desk clocks are? When was the last time you had to reset one that was off...except twice a year. They are a good, &quot;quartz crystal&quot;, 1-second clock source:

<A href='http://www.josepino.com/circuits/index?one_second_timebase.jpc' Target=_Blank>External Web Link</a>

You would need to add a voltage dropping circuit to replace the 1.5v battery, and maybe a drive transisor/FET to bring the clock to an input logic level. But, it should be as accurate as the time for the Picaxe to count input pulses and do what ever you need to output.

Ken

 

[hippy]

<i>When was the last time you had to reset one that was off...except twice a year </i>

Sometimes three, occasionally four, as there are leap seconds to be dealt with ;-)

Leap seconds are normally missed as 'expected drift', but it's an interesting issue relating to timing signals - Are they repetitively accurate. going out of synch with 'real world time', or should they synch up with it. Specifications are rarely clear on this.

 

[BeanieBots]

Hmm...
60kHz receiver, and sink with the Rugby time signal. Could be a nice little PICAXE project in it's own right.

Stan, (or any other RF guru), got any simple receiver circuits that will give a PICAXE compatible logic signal from a 60kHz carrier?
There's also a German one but not sure of the frequency. Anything like it for AUS/NZ/US?

 

[Dippy]

Maplin used to do the Galleon one didn't they? - though Maplin only do toys nowadays.

Do a search on MSF or DSF (Germany) - I'm sure That Ghastly Google will come up with something. Maybe a schematic?

I just knew this thread would meander off onto MSF.... I think the original author has run off.

 

[eclectic]

BB/D and others.
There&#8217;s lots of basic info. in this topic.

http://www.rev-ed.co.uk/picaxe/forum/Topic.asp?topic_id=2541&amp;forum_id=16&amp;Topic_Title=Radio%2BControlled%2BClock&amp;forum_title=No+new+posts+please%21+10

For a good start, at least in GB, buy a 7.99 Argos clock.

http://www.argos.co.uk/static/Product/partNumber/2551887.htm

Then, Softly-Softly, dismantle it.

Solder three terminals; Gnd PON RF

Ground PON, under software control, then take your signal from RF.
(Looks very pretty on the &#8216;scope.)

If you&#8217;re REALLY careful, you&#8217;ll have a working clock
AND a 1.5v signal source.
(I wasn&#8217;t. I broke my first one!)

The real hassle is replacing the clock hands correctly!

e.

 

[manuka]

UK mags. have had assorted 60kHz based clocks over the years, but since L.W. Rugby is rather weak here on the other side of the globe they've had little personal appeal. The leather version more than makes up for that of course for most Kiwis! My thoughts are also at the cheapo digital clock hack level. Stan

 

[Mycroft2152]

First, start with a rabbit......

 

[John Ashurst]

I wish to generate a delay of 29sec, pulse of 1 sec and repeat to quartz clock accuracy. 18x plus Dallas clock chip and separate crystal seems a bit over the top for this application. Can anyone suggest a simpler solution. I am new to PIC tecnology

 

[John Ashurst]

I'm sorry I didn't reply sooner. I thought I'd ticked a box which would email me if I got a response. Seems not. Duplicate of 1st post was because Roboform did a &quot;fill in and submit&quot; before I could stop it. I can get quartz movements from Rapid for 0.95 I need a few seconds a day so I don't think an RC circuit will hack it but I did like the gold plated 555.Haven't seen hippys hard code yet. Where is it please hippy?

Thanks all. I'll get the hang of it soon.

 

[pjrebordao]

The link proposed by KMoffett is probably the cheapest, reasonably accurate option.
However, the circuit as proposed outputs a 2Hz signal. To obtain a 1Hz just use only one of the outputs with a single diode.
With 4.5V power, the signal can drive a Picaxe input.