|
|
View previous topic :: View next topic |
Author |
Message |
ckielstra
Joined: 18 Mar 2004 Posts: 3680 Location: The Netherlands
|
|
Posted: Wed May 21, 2008 9:59 am |
|
|
arrow wrote: | My specs require that after 4 hours I have the two PICS different by no more than 0.5ms. |
arrow wrote: | ... do you think that the drift can be as large as 0.1ms over 4hours |
arrow wrote: | ... have them run at the exact same speed over say 8 hours. | It really helps if you get your requirements sorted out before you start designing. Within this thread your accuracy requirement changes a factor 10, that's a lot and will have a large impact on your design!
What accuracy do you want to decide on?
- 0.5ms in 4 hours (0.0000035% accuracy)
- 0.1ms in 8 hours (0.00000035% accuracy)
- Something in between?
Note that a 50ppm accuracy == 0,005%
You want to be 1428 to 14280 times more accurate. Not easy...
Quote: | Unfortunately I do not have the board space to use any external signal. | Sorry, I don't accept this. In a first prototype there are always methods for reducing size. You are limiting your design choices a lot by rejecting external components.
Another question: The accuracy requirement is it relative or absolute? I mean, does it matter if both units stay synchronized but the time shifts compared to an external time clock? |
|
|
arrow
Joined: 17 May 2005 Posts: 213
|
|
Posted: Wed May 21, 2008 10:51 pm |
|
|
Hi Ckielstra, SherpaDoug, and RLScott
Thank you all for replying.
Yes, my specifications are not set (yet).
The accuracy is defiantly relative! All I care about is that the the two PICS are synchronized to each other; I do not care if they drift with respect to an external time clock!
One thing I was thinking of was to reconnect the two PICS after 4 (or 8 hours) and determine how far out they are from each other. Then use linear interpolation. But I would prefer to have the two PICS synced in hardware if possible.
All the best
a. |
|
|
Ttelmah Guest
|
|
Posted: Thu May 22, 2008 2:41 am |
|
|
Seriously, if the PICs are in different locations, linear interpolation, isn't going to be that good. It'll correct for the basic differences between the crystals, but temperature, and supply changes, are still going to 'bite'. The best, and simplest solution, is going to be a PLL locked onto a common radio station, or (more complex, but 'better'), the colour sub-carrier on a TV station (better, since the frequency will then be constant, while the radio one - assuming the station is FM, will vary with the signal being transmitted). There are several FM radio receiver IC's, that have the PLL signal available externally, and with only a couple of components, a simple receiver can be built. Even 'better' do the same wth an AM LW station, such as the ones used to send the MSF signals, who often provide 'guaranteed' accuracies on their transmission frequencies. However since it is only relative accuracy that matters, this is not needed. Others have suggested GPS, and this too is a great solution, but the receiver is more complex to make.
Best Wishes |
|
|
ckielstra
Joined: 18 Mar 2004 Posts: 3680 Location: The Netherlands
|
|
Posted: Thu May 22, 2008 7:05 am |
|
|
If power supply in both units is from the mains and not from a battery you could use the power line zero crossing as a timing signal. Microchip has several application notes showing how to do this with just a single resistor. |
|
|
horkesley
Joined: 20 Feb 2007 Posts: 48 Location: Essex UK
|
Sync |
Posted: Thu May 22, 2008 9:11 am |
|
|
Lets go back to the start.
Why do you need to do this, what is the application?
I am troubled by this thead.
Regards _________________ Horkesley Electronics Limited |
|
|
SherpaDoug
Joined: 07 Sep 2003 Posts: 1640 Location: Cape Cod Mass USA
|
|
Posted: Thu May 22, 2008 9:47 am |
|
|
If you are trying to have 2 PICs "simultaneously" measure something, you could have each take a set of measurements. For example take 5 measurements, 2 early, one on time, and 2 late. Then at the end of the experiment look at the drift of the clocks and decide which measurements were actually "simultaneous". _________________ The search for better is endless. Instead simply find very good and get the job done. |
|
|
Humberto
Joined: 08 Sep 2003 Posts: 1215 Location: Buenos Aires, La Reina del Plata
|
|
Posted: Thu May 22, 2008 10:13 am |
|
|
This thread is plenty of good ideas but I´m still do not understand exactly the whole project,
I would like to know:
1) Does the boards share something? power supply, common bus, etc.
2) What's the distance between them?
3) What do you need to actuate in such synchronized way? (a valve, a relay, a bomb,
a motor, an actuator, a car engine/security stuff, an intercontinental missile...)
To know what/how do you need to do it and the involved hardware is a must to get better help.
Humberto |
|
|
|
|
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum
|
Powered by phpBB © 2001, 2005 phpBB Group
|