Ping Project
Tom Estes
Posts: 4
Ken Gracey,
Last winter I built a sonic anemometer using 4 ping modules mouted in quadrature spaced about a meter apart.· I didn't use a Stamp to control them, I used a pic device instead but a stamp, Propeller for sure, should be able to measure the microsecond resolution needed to determine wind speed and direction from the Pings.· It doesn't work on reflected waves but one Ping measures the time of flight of the wave sent by another Ping.· Inspiration (and many details) were provided in a January 2006 Circuit Cellar article "Internet-Connected Sonic Anemometer".
It works pretty well but I ran into a problem of the ping frequently "shifting" about· +/- 25usecs.· I think this is due to the rx detector slipping from one cycle to another while detecting the 40khz modulation but I'm not sure.· Not a big problem for "sonar"·accuracy but a significant one when 5usec represents 1mph over 1meter.· Unless a brilliant hardware solution comes to mind, I plan to find a software solution to negate the shifts.· Something along the lines of:· I know what time of flight should be with wind at zero at a certain temperature.· If I detect a 25usec shift, just discount it.· There are mitigating factors though, such as actual wind speed could cause a 25usec shift. This is a hobby and I had to shelve the project due to other pressing matters and probably won't be able to take it back up until winter returns.· There may also be other factors I have yet to discover, like weatherproofing.· I will have to find a sonic transparent waterproof cover.
Thought I would let you know about this·as your Ping guru's may want to look into a similar project as they investigate uses for the Ping modules.
Tom Estes
Last winter I built a sonic anemometer using 4 ping modules mouted in quadrature spaced about a meter apart.· I didn't use a Stamp to control them, I used a pic device instead but a stamp, Propeller for sure, should be able to measure the microsecond resolution needed to determine wind speed and direction from the Pings.· It doesn't work on reflected waves but one Ping measures the time of flight of the wave sent by another Ping.· Inspiration (and many details) were provided in a January 2006 Circuit Cellar article "Internet-Connected Sonic Anemometer".
It works pretty well but I ran into a problem of the ping frequently "shifting" about· +/- 25usecs.· I think this is due to the rx detector slipping from one cycle to another while detecting the 40khz modulation but I'm not sure.· Not a big problem for "sonar"·accuracy but a significant one when 5usec represents 1mph over 1meter.· Unless a brilliant hardware solution comes to mind, I plan to find a software solution to negate the shifts.· Something along the lines of:· I know what time of flight should be with wind at zero at a certain temperature.· If I detect a 25usec shift, just discount it.· There are mitigating factors though, such as actual wind speed could cause a 25usec shift. This is a hobby and I had to shelve the project due to other pressing matters and probably won't be able to take it back up until winter returns.· There may also be other factors I have yet to discover, like weatherproofing.· I will have to find a sonic transparent waterproof cover.
Thought I would let you know about this·as your Ping guru's may want to look into a similar project as they investigate uses for the Ping modules.
Tom Estes
640 x 480 - 77K
Comments
I'm doing no smoothing or averaging at all and I'll be investigating that. However it seems once the detector shifts it may stay there for many minutes (sometimes hours) before shifting back so averaging may not be the best way to combat the problem. I spent most of last winter trying to characterize the problem (many, many graphs), this winter I'll work on solutions.
Thanks for your suggestions.
1) Has anyone waterproofed the ping sensors ?
2) Has anyone tried to do this after Tom?
Cheers!
http://www.ebay.com/itm/360857671987