AVR-Chat

AVR-Chat@yahoogroups.com wrote:
> G'day all
> 
> I'm looking into a project for an outdoor time/temp/date
> giant display (20" high possibly).
> 
> As the unit will be mounted on the side a multistory building
> as I understand, I don't think the client will want to get a
> cherry picker or climb a window washing lift to reset the
> time twice a year to cater for the dreaded daylight time changes.
> 


This is not answering your time question at all, but a word of advice
about the display itself - make sure you pick the beam angles of the
LED's you use for the display very carefully.

I built something exactly like this for the side of a nightclub and we
used 20 degree LEDs. At a distance, 20 degrees is a lot narrower than
you think, and you could hardly see it unless you were dead in line.

We eventually had to rebuild it with 40 degree LEDs which worked much
better, but we still had to angle the sign downwards.

My sign has an RS485 link to the manager's PC inside the building where
he can correct the time. It uses a Philips PCF8583 real time clock and
hardly ever needs correcting. It does daylight saving adjustments
automatically (add 1 hour at 1am on last Sunday in March, back 1 hour at
2am on last Sunday in October). However if I was doing it now I would
include a GPS time module as others have suggested. But I'm not sure how
fixed your daylight saving times are - I was in Oz during April and they
changed end of daylight saving by a week to allow the Commonwealth Games
to occur in summer time. Not sure GPS would help with that.

-- 
Tim Mitchell
tim@sabretechnology.co.uk
http://www.sabretechnology.co.uk

> My sign has an RS485 link to the manager's PC inside the building where
> he can correct the time. It uses a Philips PCF8583 real time clock and
> hardly ever needs correcting. It does daylight saving adjustments
> automatically (add 1 hour at 1am on last Sunday in March, back 1 hour at
> 2am on last Sunday in October). However if I was doing it now I would
> include a GPS time module as others have suggested. But I'm not sure how
> fixed your daylight saving times are - I was in Oz during April and they
> changed end of daylight saving by a week to allow the Commonwealth Games
> to occur in summer time. Not sure GPS would help with that.

Apparently, GPS time also doesn't deal with leap seconds, and that GPS
time is currently 19 seconds off of UTC due to this.

-- 
Dave Hylands
Vancouver, BC, Canada
http://www.DaveHylands.com/

On Thu, Jun 15, 2006 at 07:35:59AM -0700, Dave Hylands wrote:
> 
> Apparently, GPS time also doesn't deal with leap seconds, and that GPS
> time is currently 19 seconds off of UTC due to this.

No, the GPS system does account for leap seconds and other adjustments.
But to get that one has to receive the long/large frame that takes about
15 minutes. I don't know what/where you are getting your data but in
general a new GPS has to see satellites for 15 minutes before making the
12 second (last time I checked) adjustment.

-- 
David Kelly N4HHE, dkelly@HiWAAY.net
========================================================================
Whom computers would destroy, they must first drive mad.

> No, the GPS system does account for leap seconds and other adjustments.
> But to get that one has to receive the long/large frame that takes about
> 15 minutes. I don't know what/where you are getting your data but in
> general a new GPS has to see satellites for 15 minutes before making the
> 12 second (last time I checked) adjustment.

Ahhh. I found a page that describes it fairly well:
http://gpsinformation.net/main/gpstime.htm

-- 
Dave Hylands
Vancouver, BC, Canada
http://www.DaveHylands.com/

At 03:35 AM 16/06/2006, you wrote:
>Ahhh. I found a page that describes it fairly well:
>http://gpsinformation.net/main/gpstime.htm

Interesting, for this time of application the error should not
matter. So WHAT is GPS time? (I know Google is my friend) It just
occurred to me that it may not be just a simple matter of reading
the time off the receiver and displaying it. Say if the clock was
located at the border of QLD and NSW (different time zones in Summer)
would it give me QLD or NSW time? Is it that clever to differentiate
the border line?

Regards

John Samperi

********************************************************
Ampertronics Pty. Ltd.
11 Brokenwood Place Baulkham Hills, NSW 2153 AUSTRALIA
Tel. (02) 9674-6495       Fax (02) 9674-8745
Email: john@ampertronics.com.au
Website  http://www.ampertronics.com.au
*Electronic Design * Custom Products * Contract Assembly
********************************************************

Hi John,

On 6/15/06, John Samperi <samperi@ampertronics.com.au> wrote:
> At 03:35 AM 16/06/2006, you wrote:
> >Ahhh. I found a page that describes it fairly well:
> >http://gpsinformation.net/main/gpstime.htm
>
> Interesting, for this time of application the error should not
> matter. So WHAT is GPS time? (I know Google is my friend) It just
> occurred to me that it may not be just a simple matter of reading
> the time off the receiver and displaying it. Say if the clock was
> located at the border of QLD and NSW (different time zones in Summer)
> would it give me QLD or NSW time? Is it that clever to differentiate
> the border line?

Umm - GPS time is a few seconds off UTC time. So it doesn't adjust the
time based on your location.

You probably want to have the ability to enter a timezone factor,
which adds or subtracts some number of minutes from GPS time to come
up with your local time.

-- 
Dave Hylands
Vancouver, BC, Canada
http://www.DaveHylands.com/

John Samperi wrote:
> At 03:35 AM 16/06/2006, you wrote:
> 
>>Ahhh. I found a page that describes it fairly well:
>>http://gpsinformation.net/main/gpstime.htm
> 
> 
> Interesting, for this time of application the error should not
> matter. So WHAT is GPS time? (I know Google is my friend) It just
> occurred to me that it may not be just a simple matter of reading
> the time off the receiver and displaying it. Say if the clock was
> located at the border of QLD and NSW (different time zones in Summer)
> would it give me QLD or NSW time? Is it that clever to differentiate
> the border line?

Try comparing the time on a mobile phone. They *should* be accurate.

On Jun 15, 2006, at 6:11 PM, John Samperi wrote:

> Interesting, for this time of application the error should not
> matter. So WHAT is GPS time? (I know Google is my friend) It just
> occurred to me that it may not be just a simple matter of reading
> the time off the receiver and displaying it. Say if the clock was
> located at the border of QLD and NSW (different time zones in Summer)
> would it give me QLD or NSW time? Is it that clever to differentiate
> the border line?

GPS keeps time in UTC. If the software in the GPS wishes to attempt  
adjustment for local time zones then that is up to the GPS's  
manufacturer.

The basic system sends pulses and markers always equally spaced. It  
happens due to leap seconds & the like there has been a slip between  
that and what humans consider the actual time. They don't slip the  
markers because the spacing between is critical for the navigation  
function. To resync the markers the entire GPS system would have to  
reboot.

There is a block of data being sent, slowly, that takes between 10  
and 15 minutes to complete. This block contains information such as  
updated satellite constellation as the GPS has to have accurate  
knowledge of where the satellites are located that it can compare the  
differences in arrival times of signals from each to know where  
itself is located. All kinds of things that are sure to change since  
the time of manufacture such as time slip adjustments.

--
David Kelly N4HHE, dkelly@HiWAAY.net
========================================================================
Whom computers would destroy, they must first drive mad.