suggestions for high-resolution tiltmeter (inclinometer) sensor?

[jmelson]

The thing I'm now curious about is that time interval for about a half-hour starting at t= 1.4 with increased noise, maybe 0.3 asec or so. This is not a one-time event; I have seen a similar period of such noise starting after the car leaves the garage, having been parked for some time. I wonder if it's related to the concrete, the soil underneath, or something else.

This is probably real.  The whole house creaks and groans when readjusting to the changes forced upon it.  I've been in the basement of our house, and been convinced somebody is walking upstairs, but it is just the whole house readjusting to temperature changes of the roof, or maybe air pressure changes.

Jon

[metrologist]

How did you correlate inclination to voltage? Do you have a long-term data-set?

[JBeale]

My "calibration" was simply observing the voltage change as I stood close, then far away on the garage slab relative to the sensor in realtime, as compared with the simple 4"/2mm manual-readout bubble level in the same place on the surface plate.  I found an indicated 2" of deflection from the bubble level, and about 50 uV change on the voltage readout.   So, about 25 uV per arcsecond. Roughly, it's by no means accurate, but I'm confident it's the right order of magnitude at least.

[rhb]

@JBeale

How did you solder to the foil tape?  On the vial or before? I can think of 3 options for the capacitor plates:

copper foil
silver paint
segments of 0.567 ID brass tubing mounted separately from the vial with springs to press into place.

I'm not sure how to ensure a reliable connection to silver paint and  I have no experience with the copper foil tape.  How good is the adhesive long term?  The brass tubing seems to me the most stable option over 20 years.

I'm just getting started with two 4"/2mm vials courtesy of @ez24 which are 0.560 diameter. The Chinese pivot using pointed screws requires a  thick wall tube, so I'm thinking gas pipe with the top and bottom cut away and annealed before setting the vial in plaster of paris.  Then mount with a differential adjustment screw in a piece annealed steel channel.

I have in mind bolting these oriented NS-EW under my deck to bedrock, so long term stability is desirable.  I'm about a mile from a large Corp of Engineers lake, so I should see lots of interesting effects from the water level changing in the lake.

[JBeale]

It was too scary to solder the foil afterwards, I was worried about the vial. So I soldered first, then peeled off the backing and stuck them on. It was quite fiddly to do this with the sticky tape and short fragile wires, but I managed it, with help of long-nose pliers and tweezers.  There is some question about whether the adhesive may allow some relative movement; I don't know. I was thinking plain Cu foil which is held in place with a layer of shellac or similar, might be more stable, but I don't know that either.

If you use a thicker electrode eg. brass tubing, it may not be tight against the glass vial, might cost some sensitivity. Might also cause temperature effects, if the geometry changes from expansion with temperature. If you use a spring against it, that would have to be some non-conductive type.  But most likely all of these possibilities work, and the difference between them may be small. All part of the fun of this game, I guess!

[rhb]

I *like* plain Cu foil held in place with nail polish.  Shellac is hygroscopic and water soluble.  Nail polish is polystyrene dissolved in toluene.  That should be stable for a very long time.  Actually, the self stick should be OK with an overcoat of nail polish and the adhesive simplifies placement.

Any thoughts yet about a long term data logger?  I assume you're using a 34401A or similar at present. I'd like to build a dedicated data collector using one of the 24 bit, 90 samples per second chips used for digital scales and something like an ESP32 to collect data and then upload it once a day via WiFi.

[JBeale]

Funny thing, it's easy to find adhesive-backed copper tape, but seems plain Cu foil is less common. This might work, https://www.amazon.com/Genuine-Copper-Leaf-Sheets-Blattgold/dp/B00RJNB992/ ...though it may be hard to solder to superthin Cu leaf without tearing it. They also offer a specialized "Metal Leaf Adhesive" whatever that is, but I'm sure nail polish works.

My meter is another antique, a Keithley 196 DMM from ebay. Only uh, 9 years out of cal...  I have played around with many different 24-bit ADCs.  For my needs the tilt meter doesn't need a fast readout, 1 Hz is more than enough and for that you can go with a simple one like LTC2400 talking with any sort of microcontroller- I usually use a Teensy or Arduino type for longer term datalogging. I have some of those that have been working for some years.

[ez24]

From the description it sounds as if you have several issues:

The soil is compacting under load.  This can be particularly severe if it has high organic content.

Yes - a 6 foot compost pile that last time I checked was 1/2 " (10' of concrete rubble on top of it) with water flowing through it

From the description it sounds as if you have several issues:

The soil is being pushed down slope by the load.

The soil is moving down slope under its own weight.

Yes to both - attached poor Google pic - some trees have been removed

From the description it sounds as if you have several issues:

The first two can be mitigated by increasing the surface area upon which the structure rests.  The last can only be resolved by piers drilled into bedrock of sufficient strength to resist the shearing force of the soil.  It also resolves the first two.

Surface area - I put 100's of yards of concrete rubble around it.  For years I checked Craigs List for free rubble

Piers drilled into bedrock is the really for sure answer - way beyond my ability to pay for them.  What I am going to do is print out your response and give a copy to the next owner

Of course, all this is speculation without pictures of the situation.

I went down there today and took pictures.   Good news and bad news.

Bad - the area looks like crap -  it looks like a junk yard -  very depressing - 2 years ago a storm snapped a 6x12 " beam in a structure next to the villa.  For 3 years I have tried to figure out how to fix it.  I get very depressed every time I see it 

Good -  it has not moved in 3 years      Three years ago (Mar 2, 2015)  I bought this :

https://smile.amazon.com/gp/product/B006JR8XBG/ref=oh_aui_search_detailpage?ie=UTF8&psc=1

And measured 0.00 on a door frame (two readings at right angles).   Today I measured again and it reads the same.
Another door frame reads 0.45 lean towards downhill, but I cannot remember if I checked it 3 years ago.

When I built it, I used cheap HF levels.

Actually I do not understand why it has not moved

I will post pictures in a few days, I need to recover from the shock of the mess and damage

[metrologist]

It was too scary to solder the foil afterwards, I was worried about the vial.

I just finished machining a nice slip-fit into an aluminum tube with press-fit endcaps. I was getting worried about thermal expansion and cracking.

The 4" vials are a little touchy, but really you see movement on your concrete slab? I had mine set up on a kitchen counter, substantial wood frame house (2x6 joists and 2x10 subfloor, 16" tile, granite counter tops) and I only see about 8" movement when I stand to either side.

rhb - a big drop of lipstick red nail polish on each end of the vial to hold it in place in my aluminum tube? I might be able to get my assembly put together tonight.

[rhb]

A suggestion:

Dig a post hole 2-3 ft deep in an area which is the least disturbed by human activity (fill, gardening, etc).   Make two piles, the upper part of the ole and the lower part of the hole.  Spread it out and let it dry. 

Then sift out all the material that will not pass through an 1/8" hardware cloth screen.  Set that aside.  Take the material form the top and bottom sections that passed through the screen, mix it thoroughly and measure off by division  enough to fill  a wide mouth gallon jar (available from any restaurant by asking) 4 deep with dirt. 

Divide the material that didn't pass through the screen in the same fraction as you put in the jars.  Rinse it and pour the water in the appropriate jar.

Fill with water, set somewhere where it can sit undisturbed for several days and stir like mad to get all the sand, clay and organic matter entrained in the water.  Then let it settle until the water is clear enough to see the sediment on the bottom looking through the mouth of the jar.   

 Do this for the upper and lower part of the hole and label the jars when you do it.

Once it settles, you'll have a good estimate of the soil composition by looking at the layering in the jars.  That will provide very important information for choosing a mitigation strategy.

In central Texas swelling clays are a huge problem.  But rather than drill some footings to bedrock, the recommended solution is to water your house with a soaker hose on a timer. Woe unto you if you ignore it as I did.  My house moved enough to break a window and open a 3/8" gap in the garage wall.  What happens if every time it rains, the clay expands and a little bit is forced out the side from under the house.  Then it dries out and the house settles a little bit.  Over the course of a few years it becomes a major problem.  The only solution is to dig holes 3-4 ft deep all around the perimeter of the house that extend under the slab.  A thick slab of concrete is poured in the bottom of each hole and after that has cured for a week, the house is leveled with jacks and more concrete is poured in the hole to hold it.

What this does is move the support of the house down below the layer of soil which is subject to drying and thus stable. Typical cost for this is $15-18k.  There are other methods using short driven piers, but there is no control over where they go,  One of my foundation engineering textbooks has a picture of one of these where they excavated after driving and the end of the series of piers threaded on the cable makes an arc and is headed for the surface.

[rhb]

@JBeale I have a rolling mill and plenty of copper wire. Time to put the rolling mill to serious work making plain copper foil.  Should I make a foot or so extra for you.

@metrologist Nail polish should do the trick.  Lots of electronics has that on the screws to keep them from moving.  i assume you made the walls of the Al tube thick enough that tightening the pivot screws won't deform the tube.  If there's any question, drill some shallow holes with a #60 drill for the pivots.

[branadic]

I'm still waiting for the vials to arrive. In the meantime I prepared a board from another project with PCap01-AD capacitance to digital converter and the firmware/config. First step will be to reproduce the results with the copper foil. Next step is to improve the way how the electrodes are attached to the glas without force or glue such as silver ink or similar ways. Step three will be, how to attach the vial to a base. One solution cold be an optical flat, but they are rather expensive.

-branadic-

[metrologist]

@metrologist Nail polish should do the trick.  Lots of electronics has that on the screws to keep them from moving.  i assume you made the walls of the Al tube thick enough that tightening the pivot screws won't deform the tube.  If there's any question, drill some shallow holes with a #60 drill for the pivots.

What makes you think I'd have such foresight? The tube started as thin-walled tube and had maybe 2mm skimmed out of it. I didn't even want to use aluminum and hoped to find an old piece of thick-walled iron pipe - but used what I had. I'm still kind of working out the mounts, they may just end up being vertical studs and nuts.

[JBeale]

The 4" vials are a little touchy, but really you see movement on your concrete slab? I had mine set up on a kitchen counter, substantial wood frame house (2x6 joists and 2x10 subfloor, 16" tile, granite counter tops) and I only see about 8" movement when I stand to either side.

You don't have to take my word for it, this link contains the video.
https://www.eevblog.com/forum/projects/suggestions-for-high-resolution-tiltmeter-(inclinometer)-sensor/msg1566820/#msg1566820
In the video background at top, you see the concrete surface of the slab it is resting on. The bubble shows a small, but noticeable movement, I estimate 1/4 division which is 1 arc-second.  That was when I moved to within a few feet. If I stand immediately next to it, I get at least 2 arc-sec of movement.  Other people have reported a similar amount of motion on a slab floor.

Also, as you see in the other video here, my kitchen countertop responds in a similar way to yours.
https://www.eevblog.com/forum/projects/suggestions-for-high-resolution-tiltmeter-(inclinometer)-sensor/msg1567927/#msg1567927

[rhb]

You need a 3 point mount with a cross vial in the direction of two balls.  If you want a flat base buy a machinist's level for $50-80.  If you're setting up a reference surface for long term monitoring get a small piece of stone counter top.

My offer of plain copper foil is open to any who are interested.  Just send me a PM with a mailing address and I'll send 4 inches per vial you want to prepare.

An R-10 O ring is a good fit for the 4"/2mm vials I got from @ez24.  That's what I plan to use to force the foil against the vial while I wait for the adhesive to set.

I'm glad to hear I'm not the only one who paints himself into a corner.

But I hear my lawn mower calling....

[JBeale]

The countertop fragment is a great idea.  I see Amazon also offers "Grizzly G9647 6-Inch by 8-Inch by 2-Inch Granite Surface Plate, No Ledge" for $34.
https://www.amazon.com/Grizzly-G9647-6-Inch-Granite-Surface/dp/B0000DD0KC/
They claim a +/-0.0001" surface (tenth mil), which is probably not a specified value for a countertop.

I have access to a mill so I could make my sensor bottom surface reasonably flat.
For purposes of fixed long-term monitoring, I'm not sure if that's preferred over a 3-point mounting.
I want to be able to reverse it 180 degrees at any time to check zero drift, but so long as the supporting surface is flat, that should work either way.

[rhb]

If you want to check for drift by reversal use a surface plate and 3 point mounting.  Or hand lap a scrap of counter top at the mount contact points using the surface plate to set the level and then the level to monitor the lapping.

Drill two  holes in the counter top and epoxy  threaded inserts in place to hold the vial assembly in place.  Then set the counter top in concrete in your observatory.  If you really want to refine it, set press fit jig bushings in the mounting plate, use those to set hardened dowels with epoxy or molten sulfur.

But if you're doing that you need to set the levels perpendicular to a few arc seconds to justify the labor.  So still more work.

Then you'll need a Lacoste-Romberg gravimeter station to monitor lunar changes in the gravity vector.

If there is significant ground water withdrawal where you live, you'll be able to observe that extremely well even without the gravimeter.

Time to start making copper foil.

[rhb]

I'd never tried this before.  Not surprisingly it's not as easy as I'd hoped.  I can make foil 0.0035 thick but I'm having a lot of problems with the edges crinkling.   Copper work hardens very quickly, so I can only make a couple of passes before annealing.

I *think* that I need to anneal, clamp the ends and stretch the center so the crinkles go away or are at least diminished.

It may be that I'm trying to get too much a reduction in a single pass and the rolls are deflecting.  It's 80 F in my shop, so I think I'll sit in the A/C and read up on rolling.

Rolling wire is not that hard.  I got good results rolling  12 AWG, but it's only 0.15" wide.  But I made over 2 ft of that before I decided to try rolling a slit and flattened inch long section of copper pipe.

I'll probably dig around for some 8 AWG ground wire and try that in the morning when it's cooler.  For the vials I have foil 0.3-0.6"  wide seems to me about optimal.

Edit:   I got 2 ft of 4 AWG, (0.20 diameter) and rolled a 2" piece to 0.0045" thickness and 0.25" wide.  The 2" piece turned into 3' of foil tape. I'm going to try a piece of copper pipe again, but this time annealing, stretching straight and annealing again between passes through the rolling mill.  It gets rather tricky once you get below about 0.020"

Edit 2:  The bubble in my vials is approximately 0.25" x 1.4" .  That suggests that the optimal capacitor plates are 0.25" x 1.5" on the top and 0.25" x 4.5" on the bottom.  That should maximize the effect of the bubble on the capacitance.  I made 33" of tape, so I have enough for 2 more vials besides what I need for myself.  I broke the piece while working it, so 2" will actually make enough for 5 vials.  So I'll be happy to practice using my rolling mill fi anyone wants sensor tape.. 

BTW I got the mill via an Harbor Freight Insiders club circular for $165 which is about 1/3 to 1/2 what Rio Grande charges for the same or similar.

[branadic]

I'm the opinion that plaster of paris is a very bad idea. Even though CTE is in the order of 12 - 20ppm/K it absorbs water and swells. This is what you don't want.

-branadic-

[JBeale]

Is there any material you could recommend for the purpose?  Would grout or portland cement be better?

[branadic]

If you want an idea on how to mount the level vial take a look at how they managed it in precision levels like shown here:



Source

It's a V groove with the mentioned differential screw on the right and a pivot on the left. What is missing here is the base the cross beam is mounted to.
What you can do is to mount the vial in a fixed V groove e.g. made out of Macor and zero the level by software instead mechanical.

-branadic-

[rhb]

Thin set tile mortar should work just fine if you're concerned about plaster of paris. 

I had a piece of 1/2" American standard gas pipe.  It's 0.622 ID vs 0.560 OD for the vial, so I'll just use epoxy at the ends of the vial.

As my mill is not set up to run at the moment I chain drilled the pipe and filed it.  I then cleaned the paint off the pipe using plumbers cloth backed sanding strip.  I'll probably roll some shim stock to center the vial in the pipe.

The total range of this vial is 2 minutes, so I don't see a viable way to avoid the differential screw for a 5" long mount.  Full range is 0.0015"

I'm copying the Chinese mount as it's simple to do with minimal tools.  I'll mount the PCB on the side of the pipe.  It's welded pipe, so I plan to anneal it before assembly.

To make the differential screw I plan to chuck a piece of 5/16" rod in the drill press and drill it out to take a #12-28 screw.  After the hole is drilled I'll thread the rod  27 tpi with a die, cut off a short piece,  tap the inside #12-28, insert the #12-28 screw and secure it in place with a lock nut.

[branadic]

I'm not sure I understand what you are planing to do, but I fear this kind of mounting will result in a lot of parasitic capacitance, if copper electrodes have been attached to the vial. This is just an experience of 12 years dealing with capacitve sensors.
Thus my suggestion was to use a base made out of Macor (machinable ceramic) with a V groove and to contact the electrodes by some spring alike flat pieces of nickel sheet.

-branadic-

[rhb]

If it doesn't work for a capacitance sensor, I can always use it as a normal vial mount.

However, in this case, I don't think the stray capacitance matters.  I gave a lot of thought to the stray capacitance problem. So long as the only thing moving is the bubble the strays should null out during the electronic calibration.  Mechanical leveling is still required because of the very small dynamic range of the vial.

My chief concern right now is the 10 degree tilt of the assembly pin holes. They don't really matter as the pointed screws are the functional pivots in the Chinese design.  But it annoys me that I made the mistake and I'm trying to figure out how to correct it without starting over.  Except for a bandsaw and drill press, this is all bench work.

[rhb]

I got the mounting frame mostly done today, but I need to get some stainless steel foil wrap to anneal the frame before going further.  Otherwise it will warp over time.

The background noise is probably surface waves.  You could make an interesting seismograph using orthogonal pairs of vials.  With 3 stations you could measure direction and velocity.  By making up the vials and covering them with heatshrink, they could be set in thin set mortar with something like an ESP32 or a LoRa node to perform the data transfer.  That would be fairly easy and cheap to do.  The electronics would need to be able to detect that the bubble was centered since you would not be able to see the bubble when placing them.  I found that I could hold the bubble on a 4"/2mm vial by pushing up or down on the table it was resting on.

Wherever you live you should be able to observe periods of increased road traffic.  You'd need to sample faster, but I don't recall what the typical spectral range for surface waves    So I'll look that up tomorrow.

I suspect that multiple stations would make a dandy intrusion detector which could plot every footstep.