Hi, I'm in search of some guidance and/or advice with a small arduino project to build a simple altimeter for measuring altitude differences within a local frame of reference. More specifically this will initially involve finding how high our house is above the stream which runs along the bottom of the garden. (Before joining the River Severn a few hundred metres further on.) I am estimating this vertical distance as somewhere between 3-4 metres and my intention is to simply deduct one pressure reading from the other and use this difference to calculate a height delta.
Rooting around in my parts box I have found a BMP085, labelled on the breakout board as a GY-65. Downloading the datasheet I see that the "Absolute accuracy pressure" min/typ/max at 0C is -4 / +/-1.5 / +4 hPa with the "Relative accuracy pressure" at 25 0C as +/-0.2 hPA.
I'm sure that buried somewhere in the garage I also have a BMP280 and I can fairly readily purchase a BMP388, both of which appear at first glance to have slightly better specs.
My question(s) is (are), will any of these devices be sufficiently sensitive (and accurate) to provide a reasonable measurement at this kind of scale. If so should I go ahead with the BMP085, or alternatively ferret though the garage until I get fed up with the search and order a BMP388?
Any and all assistance would be gratefully received.
To achieve the highest accuracy make hundreds of measurements and average them. For short term relative altitude measurements you might be able to get +/- 20 cm accuracy with those sensors.
At sea level a pressure change of 1 Pa corresponds to a change of about 8 cm in altitude
You might be able to do this with trigonometry - using the shadow of the house compared to that of a known height of a stick and it’s shadow - there are also other techniques , lengths of string , protractors etc .
Google may help you .
Or just use a level and a rod, like a surveyor would do.
If you don't need great precision, substitute a hand level for the instrument shown in the figure below. Or, better, buy or rent a laser level, or rent a "real" optical level+tripod.
And use a tape measure or "yard stick" (the longer the better) in lieu of the "graduated staff."
You may need to do this in several steps.
Sure, it would not be quite as much fun as using an Arduino with a spud (potato) gun: calibrate the propellant charge so that when fired horizontally, the "shell" lands consistently in the same general area near the stream. This may take a few tries. Then put a large contact switch in that area, wire it to an Arduino, and use the Arduino to fire the cannon and measure the time it takes for the shell to hit the switch. Estimate the drop from d=0.5gt^2
jremington:
To achieve the highest accuracy make hundreds of measurements and average them. For short term relative altitude measurements you might be able to get +/- 20 cm accuracy with those sensors.
Should initial readings seem credible, I was indeed thinking in terms of capturing bursts onto a SD card for subsequent post processing on the PC. That degree of accuracy would probably be sufficient for my needs.
Re traditional surveying and levelling methods and techniques: I did actually make a half hearted attempt using this approach when we first moved in. Unfortunately there are a number of intervening features/obstacles and the collateral damage to the hostas in particular has never been completely forgotten. Any attempt to repeat the exercise in similar fashion is likely to be meet with suspicion if not outright hostility.
DaveEvans:
...with a spud (potato) gun: calibrate the propellant charge so that when fired horizontally, the "shell" lands consistently in the same general area near the stream. This may take a few tries. Then put a large contact switch in that area, wire it to an Arduino, and use the Arduino to fire the cannon and measure the time it takes for the shell to hit the switch. Estimate the drop from d=0.5gt^2
Pure genius. I will undoubtedly be using this idea as a subsequent sanity check.
petethegeek:
More specifically this will initially involve finding how high our house is above the stream which runs along the bottom of the garden. (Before joining the River Severn a few hundred metres further on.)
Oh dear, I guess the River Severn is in you backgarden by now ...............
Anyway, if you were leveling a large area of ground you would use a long plastic tube or hose filled with water, think bathtub shape. The ends of the long tube are upright and clear so you can see the level.
Easy enough to arrange to use the same technique to measure the differences in height.
jremington:
To achieve the highest accuracy make hundreds of measurements and average them. For short term relative altitude measurements you might be able to get +/- 20 cm accuracy with those sensors.
At sea level a pressure change of 1 Pa corresponds to a change of about 8 cm in altitude
You want to log GPS and pressure and go round the region several times so that the general atmospheric pressure trend can calculated and subtracted off afterwards, then the measurements averaged.
Its pushing the limits of these sensors, but probably doable. The noise/fluctuations in such sensors over long
timescales are the unknown factor - hard to separate from atmospheric variation.
Don't do this on a windy day, winds will add pressure noise to the readings. High pressure stable conditions
should give the best baseline.
petethegeek:
...the collateral damage to the hostas in particular has never been completely forgotten. Any attempt to repeat the exercise in similar fashion is likely to be meet with suspicion if not outright hostility.
If you can carry a pressure sensor to the stream, couldn't you carry a 4m rod with a tape measure attached to it, with the same collateral damage to the hostas?
BTW, don't forget to embed a BMPxxx into the penultimate potato, so you can say you used a pressure sensor to measure the elevation difference...
Then put a large contact switch in that area, wire it to an Arduino, and use the Arduino to fire the cannon and measure the time it takes for the shell to hit the switch. Estimate the drop from d=0.5gt^2