Air and Water pressure measurements

That's a bit puzzling. Do you want to measure the ambient pressure from +10 to -5 meters, or do you want to ignore the ambient pressure and measure the pressure at some sampling port? ? At the same time? To what accuracy? Does the air sensor have to be able to tolerate the

-5 meter under water environment? And why are they buried? We've got a lot of questons, and we're not even from New Jersey.

OK, the main goal is to map the contour and measure the water level in a moving hillside.

I need to measure air pressure for references at several locations tracked by gps (within 1 meter or better)

? At the same time?

Yes, several water pressure sensors at the same time.

Fraction of a meter if possible.

Yes, replacing them as needed. Perhaps inside a pipe to be lowered during samplings.

For underground water level.

No problem, thanks for asking anyway.

One foot of water is roughly half a psi or more exactly 0.434 psi

Easily derived by remembering that a cubic foot of water weighs 62.4 pounds (and holds 7.5 gallons).

Fire science 101.

--
Many thanks,

Don Lancaster                          voice phone: (928)428-4073
Synergetics   3860 West First Street   Box 809 Thatcher, AZ 85552
rss: http://www.tinaja.com/whtnu.xml   email: don@tinaja.com

Please visit my GURU\'s LAIR web site at http://www.tinaja.com

Okay, that's helpful. Those pressure sensors sound about right for the job.

Of course there's plenty of environmental and packaging issues, so you'd better build up one or two of these pipes and test them out in the real world. You only get a few millivolts per psi, so you'll need some pretty stable and low-offset op-amps to condition the signals.

The hardest problem might be to get the 1 meter vertical positioning accuracy for the pipe's position-- hope you have some laser surveying equipment. I don't think GPS is up to the task. How do they measure hillside slump nowadays anyway?

We might bury a microcontroller with the sensor, so we only need to deal with digital interfaces.

There will be a mobile unit powering them up and collecting data.

For horizontal positioning, we are using a WAAS enabled GPS and statistical samplings of 100 points per location. Vertical positioning

is still up for suggestions.

You go to the building superintendent, and offer to give him this nice barometer if he'll tell you the building's height.

Oh, wait - wrong problem! ;-P

But seriously, how much is your budget here? I've seen a laser measuring device in the shop (we hire the guy to bring it in) that apparently gets

0.001" precision at, say, 3 meters range. I'd think there's some kind of laser surveying system - what do they use to measure the plate drift by faults? I'm almost sure I've seen something like that on edjamacashunal teevee. Don't volcanologists do something like this too?

But I'd have serious qualms about calibrating an altimeter to get within a meter of the actual altitude - this is why I thought, "barometer"; but there has to be some kind of reference. Where is the ground? ;-)

Is there a benchmark nearby, and maybe a surveyor in town?

Hey - here's a brainstorm - just anchor your posts, and put a little corner reflector at the top of each one at a calibrated height above ground, and track them with raster-scan LIDAR. ;-)

Good Luck! Rich

I don't even know what a 'mmm' is. )-;

Thanks, Rich

Not too far off what a modern surveyor does with his total station (no raster scanning, but using corner reflectors and lasers). And that is what someone I know used for tracking glacier motion a couple of years ago. Transits are so passe that I've had one on loan for over 10 years at this point. I don't forget that it's on loan, and even remind the guy who loaned it to me that I have it from time to time, but it's old, slow, out-of-date technology.

--
Cats, coffee, chocolate...vices to live by

It's a moving (sliding) hillside of several acres, directly above a community of expensive houses.

Well, the community spent a million dollars previously to patch things up. They are about to spend another million again. I am trying to convince them to fix it right, but first I need some data.

The system has to be in place and working continuously. Data exchanges/downloads can be scheduled in bursts. GPS can get within a meter horizontally. We can probably help correct the vertical data with landmarks or features. Some part of the hill moved 10 to 20 meters over the past 10 years.

The ground is subject to movement. Computer readable barometer would be acceptable.

Yes, I think they charge $500 (or more) per visit. We can't ask them to come everyday. They surely wouldn't come after a heavy rain storm, when we need the measurements.

PS: In response to your other post, mmm is my short hand for m * m * m or cubic meters.

**Groper alert !

** You must be totally nuts !!

Variations in air pressure due to WEATHER will completely SWAMP the tiny varition due to altitude you mention.

PLUS - you cannot equate wet soil and water.

Wet soil is much denser than water.

....... Phil

Ah! Thanks. :-) In the US, I've seen m^3 or such.

But, if you find a benchmark, you'd only need the surveyor once, to set your reference point. (I hope the thing I'm thinking of is called a "benchmark" - it's a reference point of some kind that surveyors use so they know where stuff is at, and it's supposed to be accurate to some ridiculously tight tolerances. If there are expensive houses there, there is very likely one nearby.)

But, I'm also thinking; isn't there differential GPS, that can get within centimeters? How hard would it be to get Z axis information from their signals?

Good Luck! Rich

Using a Garmin C320 and sampling hundred points (while eliminating some points) , I can position horizontal locations within a meter. So, given enough samplings, horizontal positioning is not a problem. In theory, vertical positioning is at best 10% of horizontal. I don't even bother to find out how to get that from my C320. By the way, you have to poll the GPS from a PC. Their great software is supposed to hide that detail from the user.

DGPS can get millimeter precision within small areas. Horizontal, at least. I know that GPS favors horizontal accuracy over vertical, but I don't really know how much it can be corrected. You could make use of it by using two GPS receivers and calibrating them at a reference location, then moving one around to make your measurements. That's how public DGPS works.

But if vertical DGPS is comparable to horizontal DGPS, you can get sub-meter precision at least. Maybe even millimeter precision. Does anyone here have enough knowledge to say?

By the way, you

So get a couple of handhelds that you can read directly :-).

John Perry

I believe that the signal precision is 10:1 (horizontal:vertical).

I can pull a 9 digits radians (in double precision floating point) out of the C320. But the number tends to drift and jump. It could be some government requirements to degrade the signal. I am still working on filtering the signals. Actually, I am looking for GPS modules with direct radian output, so I don't need to hog my wife's C320. I would need to bury several of them and expose the antenna only. The problem is that anything above ground is subject to lost or damage from man or nature.

It appears that the MPX2010 is available in many packages, including some SIP ones like the MPX2200. You might want to double-check and see if you can get those instead; some are in-stock at Digikey so I doubt they're especially rare.

Assuming for whatever reason you're stuck with the SO, and you just want wires going down to the sensor, your best option is probably to get a standard SO8 to DIP or SIP adapter for two or three bucks and ignore half of it. "Surf Board" is one brand; there are others.

I would be concerned about noise on the long wires if you're putting just the sensor underground and everything else up top; it is not necessarily unworkable, but if this were my project, I would try to confirm the viability of that approach early on.

Yes, I understand that. But is that inherent somehow that can't be overcome by differential GPS? My application didn't require vertical reading, so I'm not as well informed there.

They shut off Selective Availability in the mid-90's. To get better, you need special processing like they use in the (expensive) surveying receivers. Filtering won't help; the fluctuation that remains is due to atmospheric interference. That's why DGPS is still useful, even after SA.

The point of DGPS is that a reference receiver with exactly known location receives the satellite signals with roughly the same atmospheric interference experienced by all receivers within a few dozen to a few hundred miles, depending on conditions. The reference station then broadcasts corrections so that an appropriately equipped receiver can use the broadcast corrections to correct its readings.

But if you have decent receivers very close together, you can do the corrections based on your own reference receiver. Simply subtract the GPS errors measured by your stationary reference receiver from the remote receiver that you carry around, or the planted remotes that you want to monitor for movement, and you should be able to get very precise _relative_ locations.

That seems to make for a straightforward application of differential techniques as I described above. Several C320's should do a very good job for you, I would think.

My one and only GPS project ended just before SA was shut off, so I'm not current. Good luck!

jp

Very roughly, with full view of the sky you can expect vertical accuracy to be about 2-3 times worse than horizontal - given by the geometry.

If you have a DGPS system which can do sub-cm accuracy, you'll get sub-3-cm vertical accuracy.

To the OP, please don't expect 1m accuracy by "sampling 100 points and eliminating some". GPS is not that predictable. The ionospheric errors have a tendency to 'wander', slowly. In other words, you may have a whole hour where your handheld is consistently off by 5m or more.

Some handheld GPS devices are cheating - if they think you're not moving, they "nail down" whatever position they have, to look more accurate. If you can, switch off filtering. If you can't, power cycle the device for each measurement.

You can indeed get good positions out of many samples, but you need to sample over a few days.

Samples can be eliminated if you have access to the residuals from the position solution.

Kind regards,

Iwo

Is DGPS same as WAAS? If not, how would you compare the two? The Garmin C320 is WAAS enabled.

Yes, that's what I am seeing. I just need to space out the 100 points over days.

That's not a problem. We can sample for days or weeks.

I get radians in double precision floating points. I am currently looking at 9 digits precisions.

Thanks.

WAAS appears to be a rather coarse, distributed DGPS that negates you needing to have a stationary reference receiver, as they have some and broadcast the corrections. Presumably this means that you can get the corrections "in real time", as well. Per Garmin, it is apparently limited to the US, though other, similar systems are being implemented elsewhere.

DGPS is a pretty simple concept, and doing it locally to your project gets you the best possible resolution. You have a stationary reference receiver, which sits there and stores its location as reported over time into a computer. You park this somewhere that isn't thought to be moving, yet is near the area of interest that is moving.

You have other receivers, and you store where they say they are over time as you take measurements with them and move them around, or in this case, as the hillside they are fixed to moves with them.

Then, you take the one set of data and compare it with the other set of data to get a more precise idea of where the moving ones really were, by removing the relative recorded motions of the one that was not moving, at the time of each measurement. You don't really get a precise idea of what is happening until you do the data-crunching step, but when you do, it's quite precise.

--
Cats, coffee, chocolate...vices to live by