Dear Zainul,

Thanks for your kind explanation.
This makes me clear about why I couldn't achieve an accuracy better than 
100ms-order
with a conventional GPS receiver (e.g., Garmin **) alone.
(Actually I already tried to synchronize clocks using GPS receivers,
but was disappointed with the performance. :) )

Thanks again for you help.

Best regards,
Jinkyu






Zainul M Charbiwala wrote:
> Hi Jinkyu,
>
> If you want to go the GPS way, and have a scenario where a good GPS
> signal is available to you, you could use the PPS signal that is
> provided by most GPS receivers.
>
> This is a pulse-per-second signal, the edge for which is guaranteed to
> have a jitter no less than 1us (typically). To get the event time
> stamp in microsecond numbers however, you should wire the PPS signal
> to a hardware counter that runs off a decently calibrated high
> frequency crystal. You can then get event timestamps by reading the
> absolute time in seconds from the GPS and the fractional part in
> microseconds from the hardware counter.
>
> If you're adventurous, you could design (or use) a frequency
> multiplier that uses a PLL to lock a low-stability high-frequency 1MHz
> clock to the high-stability low-frequency 1Hz PPS signal. This will
> result in (ideally) a high-stability high-frequency clock that you can
> then wire to a counter.
>
>
> Hope that helps,
>
> Zainul.
>
> On Sat, Aug 15, 2009 at 7:39 AM, Jinkyu Koo<[email protected]> wrote:
>   
>> Thanks for the comments, Alan and Thomas.
>>
>> Wiring or powerful amplifier would be a good solution.
>> However, in my scenario, distance between two nodes sometimes reaches
>> several tens of miles,
>> which makes it difficult to use such an approach.
>>
>> So I am wondering if there is any equipment that can measure the time of
>> event in micro-second order precision,
>> and can be synchronized with some standard clock.
>> For example, I imagines a device made by mixing an oscilloscope and
>> GPS-receiver :)
>> If I can secure two such devices,  I will be free from physical distance.
>>
>> Thanks,
>> Jinkyu
>>
>> Alan Marchiori wrote:
>>     
>>> On Fri, Aug 14, 2009 at 11:19 AM, Thomas Schmid<[email protected]> 
>>> wrote:
>>>
>>>       
>>>> No, what Jinkyu tires to do is to evaluate the precision of FTSP over
>>>> a whole network. One possibility would be to have your beacon node use
>>>> a power amplifier like this one:
>>>>
>>>> http://focus.ti.com/docs/prod/folders/print/cc2591.html
>>>>
>>>> This would extend the range of your beacon compared to other nodes,
>>>> increasing the probability that more nodes will hear it. However,
>>>> depending on your distances and the accuracies that you look at, you
>>>> might want to incorporate time of flight into your precision
>>>> measurements.
>>>>
>>>> Cheers,
>>>>
>>>> Thomas
>>>>
>>>>         
>>> I see; you could also use some form of backchannel communication for
>>> syncornization.  Possibly even just a wire supplying a 1 pulse per
>>> second tied to an interrupt input.
>>> _______________________________________________
>>> Tinyos-help mailing list
>>> [email protected]
>>> https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help
>>>
>>>       
>> --
>> Ph.D. Student
>> School of Electrical and Computer Engineering
>> Purdue University, West Lafayette
>>
>> Office: EE338 & MSEE280
>> Cell: 1-765-337-1704
>> E-mail: [email protected]
>>
>>
>> _______________________________________________
>> Tinyos-help mailing list
>> [email protected]
>> https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help
>>
>>     

-- 
Ph.D. Student
School of Electrical and Computer Engineering
Purdue University, West Lafayette

Office: EE338 & MSEE280
Cell: 1-765-337-1704
E-mail: [email protected]


_______________________________________________
Tinyos-help mailing list
[email protected]
https://www.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help

Reply via email to