home made wave height sensor

Way cool piece of equipment you've built!

Just taking a swag here, but I suspect if you want more differential movement at the indicating end, you need to allow more total volume movement. So increase the diameter all the way from the base of the pressure tap through the manometer.

But don't start drilling until one of our specialists comes along and confirm.

A lever would be easier to adjust, as cool as this rig is, how about a wave following float with an indicator at the other end?

Wow! Who'd a thunk it. I don't know where your research is going, but it looks like fun, Onliest thing I can think of is using a venturi in place of your carefully machined receptor, with seperate inputs above and below the choke point. Mebbe someone in EE&C can advise on better sensors. It is a speciality of boat designers to pay particular attention to waveform mechanics. I hope you have opened up a new way of looking at it.

Best regards, Carl

Clever, but it will not work. Water has mass and essentially you have a spring attached to a weight. The weight will bob up and down, but it will not be phased to the wavefront you wish to measure due to inertia. Your measurement will affect your experiment.

Additionally, your pressure transducer will not be able to resolve the delta in your waves for the very reason you cited. You have a problem with discrimination.

Ask yourself how might you take advantage of the refractive index of the boundary between air and water?

I agree with the AH. That's as in Hero.

Slow, fuzzy, useless response, at best!

I must be really thick. Are you trying to predict the amplitude to trigger the camera?

Just look at your picture...
If you want to see the wave height...then just look at it.
A glass/perspex/whatever panel in the side of the tank and an optical sensor.
Alternatively a vertial row of stainless steel pins and sense the conductivity between each pin and a common rod which runs up alongside them all.
Alternatively just have big bore tube hanging down into the water, closed at the top with a pressure sensor mounted in the top, ok the level in tube won't actually follow the level in the tank, so you will need to make calibration adjustments, these can be acheived by using a stick... the wet bit gives the max wave height (This is very much like what you have already done, but aviods the long thin pipe, mass of water etc.

Blimey, this is too easy, A float with a stick attached and an optical sensor, or a felt tipped pen and a roll of moving paper....can we all come play?

Or just simplify what you have done and calibrate it.

Del

I work in a hydraulics laboratory and for wave height measurement we would usually use a capacitive wave height sensor, which you could make yourself if you have some electronics knowledge.

Earlier commentors are correct in saying that it is an inertia problem and also the thin diameter of the tubing. Due to the thin tubing, the velocity in the tube will be high and will be subject to friction which will slow down the movement of the water. Try 12mm hose and see how it works. make sure the pressure tapping hole is big enough as well.

Frankston

I like Del's suggestions

For simple school demos a fish tank cleaner with rare earth magnets and a bigger float over an indicator/scale (red inside attached to float):

or more adjustable for finer measurements, a float with a pivot and lever arrangement:

And since there WILL be lag in either system, how do we quantify it in an analog world?

Or are we off on our own tear and missing the point of what you wanted it to do?

I wanna come play too, obviously!

Oooh, Yes, I like your idea too, prrrr prrrr... nice, a horizontal lever attached to a potentiometer with the other end of the lever bent down and a float attached..........arm must be long relative to vertical deflection, but it's all simple geometry.

Del

Between the pot's mechanical resistance and the mass of the buoy the resulting data will not be very accurate.

Additionally, pots generally have a rotation of about 300°, so small deflections will be very hard to resolve.

Of course the original poster did not discuss what the requirements are for the measurement. We don't know the range of period, the range of amplitude, nor the degree of accuracy that is required.

The best mechanical solution that incorporates that design concept I can see is as shown below. If the magnet is the pivot point and a small radial diameter is chosen for the magnet, then the amount of radial momentum will be minimized. The system still has mass, so it will be mechanically inductive.

The pivot point also shold be at the neutral position and teh distance between the bobbin float and the pivot point should be close to each other to minimize mass of the wire.

Using a precision instrument amp you can boost the small magnitude of change in the Hall sensor to a range that is sufficient for measurement discrimination.

So, how do we eliminate the mass problem. One way we can virtually eliminate it is by the following method:

The trick here is to size the volume of air in the narrow column and the wide column at the top so that the maximum displacement by the wave will not cause the microphone element to reach maximum or minimum excursion.

The downside with this approach is that the wavefront can not be stationary (i.e., a node or an antinode) since the mic has a lower frequency limit. If the test frequency range is above 5 to 10 Hz, this system will work quite well and is cheap. The mic must be sealed into the column, too.

If the system must function down to DC, then a very sensitive pressure sensor in place of the mic element would be the next choice. In fact it may the best choice, but I just don't have experience with sensors that would be extremely sensitive to very small pressure changes.

Between the pot's mechanical resistance and the mass of the buoy the resulting data will not be very accurate.

I disagree, it's a simple matter of designing the geometry to suit the application, one could always add in a gear if necessary and a multiturn pot. The mass of the float would by it's very definition be very similar to that of the water, and thus would be likely to have a similar frequency response to the mass of water supporting it.
Anyhow, whatever...this isn't my field of expertise and I'm quite willing to be wrong...it has happened before...although I could be mistaken.

It was an idea not a full blown design.
I thought we were all throwing ideas in to the ring, I have no wish to argue the point anyway, I'm sure Ram44 will sort the wheat from the chaff.

Del _{<I shall go and sulk in my secretest cat nest now>}

I work in electrical systems daily and pots do have a pretty significant rotary resistance. Gearing only exasberates the problem because you are amplifying the torque at the input end. If you think about it you are doing more mechanical work when you take 5° rotation and step it up to 300° rotation. Not only do you do 300° of work, but you add the mechanical friction of the gears and bearings into the system.

Yes indeed...
But water has big coefficient of floatiness.
This can support vast heavy loads which have even more inertia than a pot, things called 'ships' rely on this property I believe.
Anyhow I do indeed take your point as there would be backlash too, and to maintain linearity you'd want to use a small angular deflection, which would reduce resolution...and you still need an AtoD, unless you used a rotary encoder in the first place...I can think of a million reasons to not do lots of stuff.
This is why I believe in WAQAP^TM (Wrong As Quick As Possible) as a design and development tool.....
Maybe I won't bother going to shoot my longbow today...I'm just too depressed...I've got this terrible pain in the diodes all down my left side.

Del

Poor Marvin.

Actually, any system measuring the pressure generated by the rise and fall of the water will effect the rise and fall noticeably due to the damping effect of the trapped air.
An alternative approach may be to have a vertical pipe (say 1" diameter), the lower end being always submerged, a flowmeter of the type used for measuring respiratory function could be placed in the top of the pipe.
Such a flow meter has very little resistance to flow and as such would have very little effect on the waves being measured.
These flowmeters typically have either a rotating element with an optical or magnetic sensor, or are a series of parallel tubes (think bundle of drinking straws about 2-3" long) to laminarise the flow with a differential pressure sensor measuring the pressure drop either side of the tube bundle (this type is known as a Fleisch Pneumotachograph).

There are some relativley cheap instruments of the rotating vane type on the market with computer interfaces, I would think that the frequency respons used in human respiratory function would be eminently suitable for this task. I was for many years employed designing such instruments.

This is becoming a very interesting thread, be sure to let us know how get on with your project.
Del

A blood pressure and pulse meter used by thousands of peopel pretty much fit the criteria for period interval and the variantions of pressure can't be much different .

I would think such an instrument can be cannibalized for the necessary transducer bits and pieces. Even new it only cost around $89.95 at the pharmacy.

If you are going to push for measuring wave action with pressure readings, I am going to make a statement for chilling the water and exposing the interior wall to a heat lamp. Then it becomes a simple matter of recording the changes in temperature of the wall of the container. No friction involved. No back-lash in the linkage. And no wear concerns.

Several rough concepts popped into my head, don't know how valid they could be, so I'll wait for and learn from the responses.

First, I thought of lasers used by police to listen in on suspects' conversations. The police point the laser at a window. The window acts like a microphone diaphragm and presents varying resistance to the laser. Electronics measure the resistance, and convert the pattern back into sound. At least, that's how they explained it on TV show a long time ago.

Because we are dealing with homemade system, a "laser tape measure" might be used instead. If that's not feasible, then a photoelectric system might work, but then I can see problems with resolution.

I also wondered about the relationship between the surface waves and sound waves. If they propagate through the water the same, then an underwater mic feeding into an oscilloscope might work.

My final thought is to use a timer. The float triggers the timer to start at the first crest, and stops at the next crest. Then a little math should get you the wave height.

There are cheap acoustic reflected Doppler devices used to automatically open doors for people walking towards the door. Pointing such a device vertically downwards and reading the analog output should do the trick. The signal will be proportional to velocity, easy to integrate or differentiate. Other cheap acoustic distance meters are available at hobby suppliers. Water reflects near ultrasound well enough.

dovy is correct; I read down through the list to see if anyone had thought of this. Here's an example: http://www.hobbyengineering.com/H2951.html. Some of these grew from the camera industry, where automatic focus required knowing the distance; they've become so common that costs dropped. Here's one offering software to deal with the outputs, and capable of readings at 5ms intervals: http://www.senix.com/family_ts.htm. There's a nice comparison chart set at http://zuff.info/RangeFindersComp_E.html that is worth some time: it gives price approximations, method of operation, links to web sites, etc.

Thanks for kick starting my memory...

There are several practical examples, but I'm referring to projects carried out a long time ago so that I don't remember the Doppler maker's name. The device was actually used to measure sea wave height in a harbour where calibrating a sunken pressure gauge was a requisite.

The distance measuring acoustic device was made by Proxxon in either Germany or Austria. It worked very well but it had no direct electrical analog output signal. When this kind of a device is pointed directly down it will receive the reflected sound wave only from the top of a wave, top being defined by the beam width. A nice exercise in Geometry.

Piezo sound generators and receivers are cheap and can even be used to measure wind speed by comparing peak delays upwind and downwind.

Long ago we had to measure the height of waves in a model of a channel in real time.

We ended up with a double sided (gold) pc board as the probe. The one side had a pair of parallel electrodes while the other side the pair was flaring out at the bottom.

We recorded on a multi pen XY plotter or a 6 Chanel ultraviolet galvanometer device.

As far as I remember it was rather accurate and neat.

One problem was that we had to control the PH of the water otherwise the readings would be way out. That was hard because we used a lot of cement in the models.

Clever!

Why did you mark yourself OT????

Your idea is clever because its also simple and probably easy to implement.

I cannot fathom out why nobody else appears to understand JUST how simple....

My only suggestion is to not measure the waves in the same axis as the wave movement, but at 90° to them.

Then you only get two reflections, one at the top and one at the bottom and the sender and receiver can be right next to each other aimed vertically down...tell me if you need a diagram....

If the sender and receiver sensors are next to each other, you only need to angle them a tiny amount to allow the Rx to receive the signal when a "flat" part of the wave goes by. This will only be at the top and the bottom of the wave. The rest of the wave will either reflect the laser beam either towards the oncoming waves, or to the retreating wave.....

Having the angle in your diagram means that you need to have the receiver at different distances away, depending upon the height of the waves.....you show several Rx positions in your diagram for that reason:-

My version, you only need on TX and one RX, basically bolted to each other, but its still basically your idea....I have shown 4 of them for clarity, but in fact only a single one (Tx/Rx pair) is needed......

Here is a simplified diagram of your/our idea:- (my waves are crude, but I hope you get the idea!)

Only Tx/Rx 1 and 3 will get a signal return, on the front or back of the wave, the signal will be reflected away from the Rx/Tx pair.....

When viewed from the direction of the blue arrow, it should look something like this:-

The distance between Tx/Rx and the tops of the waves, should be as great a distance as possible, to allow the focus on the bottom of the wave to still be achieved easily (angle to each other).

Unless I'm missing something, your setup with the RX & TX adjacent will tell you that there is an extremum - but not how far it is from the RX/TX (unless you use pulse timing, and at 300mm/ns you need some fancy stuff).

I actually wasn't quite that far along, but you are perfectly correct. I suspect that pulse timing may be easier.....hopefully Kris will tell us exactly how we need to do that part....

Don't know whether Kris wants to go that way - but if he does, I don't think he'll be able to do it with a 555 timer .

Also: 2 electrodes stuck to the inside of a plastic tube (bottom plugged) and the tube perpendicularly immersed. The electrodes connected so that capacity can be measured.

...as I suggested in Post #29? With highly skilled drawing!

Right!

Dovy - point taken: if this can be built into the wall, as the immersion probes can, then yes, it's still in contention. If it presents even a relatively small obstacle to the waves, I'd drop it a notch in the rankings. Personal opinion, true, but anything that potentially can disturb the system reduces accuracy and confidence in results, to my way of thinking.

Following that line of thought: ideally, initial testing would find information on wave patterns for various input conditions, and then the tests would be repeated with the proposed means of extracting energy installed, and comparisons made. If the amount of mechanical energy obtained is also recorded, even better.

At least some of the disagreement among proponents of various methods appears to be caused by the imagined sizes of the system. My personal bias comes from having done that in the wild with a real sea (also from being too lazy to look at all posts and see that capacity HAS been proposed!)

Good observation. The OP (Ram44) merely says that his (?) test trough is "long". Given the cost issues, I'm guessing that this means no more than a few tens of feet (single-digit numbers of meters) rather than a model basin sort of setup like a major university or other large entity might own and operate. Being able to work in a "wild sea" doesn't square with the mention of a paddle to drive the waves. This person needs to find a way to gain access during off-hours at a large wave pool [http://en.wikipedia.org/wiki/Wave_pool] like some amusement parks have!

Hi, your idea is great but needs some modification.

If wave is changing its height there is a moving water mass through a tube.

The tube acts as viscous damper. The mass-damper has a time constant τ = m/d

Damping constant to be measured in Force/velocity in N/(m/s) or N.s/m

So the time constant has the dimension <Kg.m/(N.s)> or s as necessary.

1/τ is ω_c or cut-off-frequency. At ω_c the amplitude response is 0.7 (first order system) and phase is 45°.

Above ω_c the amplitude response is linear (approximately) with frequency, so at 10ω_c you will have only an amplitude response of 0.07 and near 90° phase.

So shorten your tube and allow for a bigger diameter if shortening is not sufficient.

Bigger diameter is effective as long as there is no vorticity in the tube. Re near ? 40?

If laminar flow in the tube then mass is ~diameter² and viscous friction = damping ~diameter⁴.

Remove any restrictor if you have some in connection of tube to U-tube.

RHABE

hi all;

thank you so much for your ideas.

i am trying to determine the wave height and period of the waves initially. my idea of setting up the instrument is to test energy losses of waves on inclined planes. my aim is basic, to find a relation between the wave parameters and energy transfered up a inclined duct. You could say that i am testing an inclined Osscilating water column to minimize energy loss by reflection. since in shallow water wave particles have velocity almost parralel to the bulk motion of waves towards the beach, an inclined chamber will cause waves to rush up,compress and push a bulk of air to a bi-directional turbine. i also have a new design in mind which requires calculation of the best incline angle for energy transfer from waves.

I cannot afford expensive transducers or PIV technology, so i started of making a series of intruments that will record the wave parameters. also at a few locations i need the transducer to be well below water. but for other locations i think the ideas you have given are better. I already bought a air pressure sensor which i placed on a pipe end. Then i placed the pipe in the water so that as the water rose up it pushes an air column up to the pressure sensor. this did not work so well with plenty damping in place.

I am still experimenting and i will update as soon as i get results. Thank you so much for your interest.

Cheers!

Ram

I would use a simple film camera in a darkened room with a flash to capture the event.

The tank needs to have one side made of plexiglass so you can visually see the water. I would also use food coloring in the water to make observation easier.

You can directly measure the film after you develop it. If you have a digital camera, then print the picture.

You can also rent a high speed sports digital camera that can take pictures at 8 frames a second or find a friend that has one you can borrow.

A high speed film camera would be the cat's meow. You can capture lots of frames and see the propogation if you need that.

Hi Ram,

The easiest way to increase the sensitivity of a manometer is to incline it off the vertical.

If the measuring part of your manometer tube is say 10cm long but when in the vertical position you are getting say 1cm travel, then incling the manometer at an angle such that the top of the tube is only 1cm above the bottom you will be able to use the full scale of the manometer for your small (1cm) reading changes.

This should help you capture these small movements more easily on camera.

Someone mentioned capacitor, but that was it.

Have you thought about a sturdy metal rod (you already have something I believe), encased in say garden hose (as a dielectric) or similar, longer than the highest wave you want to measure and a similar rod, relatively close, but not touching, that is not encased. This would form a crude capacitor when both are in water. Using silicone as an excellent sealer. The water will be usually a conductor as long as you are not using distilled, of some degree or another. But even with distilled, the idea should work as then the water is a dielectric as well.....

It would need to be calibrated for the water type before usage......salt water conduction better than fresh for example.....

A small version could be built to test the viability of the system first.

Use this in a circuit of a frequency generator/oscillator of some sort to alter the frequency in time with the variations of capacity. Generally speaking, the lower the wave, the lower the capacity and the higher the frequency and vice versa.

The timing circuit could be a relatively simple 2 transistor oscillator or just as simple a NE555 circuit......

To make the circuit even more exact, the rods could be made much longer and set at say 45° to allow a greater frequency range for the same wave heights......

You might even be abble to make a capacitance bridge and then the capacitance will be a function of the value of capacitance measured....

You then need to calibrate frequency against wave height.

Something along those lines will be relatively stable, no mechanical pieces bobbing about and it should not have any latency at all. The drops of water clinging to both posts would not affect the capacitance greatly, as they are tiny in relationship to the mass of water in a wave.....

This simple sketch may help to make the idea somewhat clearer.

The wires connecting the oscillator to the sensor should be as short as possible to stop extra capacitance.

I have never built anything along these lines I hasten to add.......but theoretically, it should work just fine (as all theories do!!!)

Anyway, it should give you a further few ideas to be getting on with......

Let us all know how you get on......

Hi Andy Germany

I had someone do it before, we used a bridge to compare out differences caused by other factors. see my post 21.

I think it involved capacitance and not resistance as the surface of the PCB were laminated.

I will visit the lab soon to see if this 1972 "invention" still exist.

Good move, thanks.

Of all the suggestions so far, the Electrical Capacitance method by Frankston is be far the best. I have used this method for Hydrodynamics labs in the past with complete success. The sensing electrode can be very simple and if the tank is constructed of plastic, need only be a silver foil strip stuck to the tank wall and insulated with sticky tape. this would offer minimal effect to the wave motion. Capacitance Transducers are available giving an analogue output to feed a moving coil indicator or other device. The capacitance measured is that between the electrode and the electrically conducting water.

J.J.

Hi Ram44,

Freescale Semiconductor makes a variety of pressure sensors including differential and absolute. I believe they have DP sensors as low as 10 inH₂O, with some having outputs of 0 - 5VDC. Feed this to a multimeter that can give max observed voltage, and there you have your answer. I don't know if the response time of the sensor would be satisfactory. Anyway, here is a link to the pressure sensors.

http://www.freescale.com/webapp/sps/site/taxonomy.jsp?nodeId=01126990368710

Oh, and did I mention that you can get free samples of some of the sensors?

Mike

Simple graduation marks along the inside wall of the tank should give you your wave height. Alternatively, a plexiglass side with graduations will serve the same purpose without requiring you to bend over the tank.

For period, again, if the wave generator is fairly constant and you don't need high accuracy, a floating steel cylinder restrained in a guide can be used to trigger a timer with an inductive proximity sensor. Just position the sensor at the max height (or min. height - your choice) so that when the float reaches that high, it starts the timer. When it comes back up after making one cycle, the second pulse will stop the timer, and you can calculate your period.

If your wave generator keeps changing height and period, this idea won't work. But if it gives a fairly constant wave in terms of height and period, and you don't need super accurate figures, I think this one will do fine.

regards,

Vulcan

thanks to all of you!

i used a simple float inserted on a steel rod. i had marked the steel rod for measurements and just like you suggested i used a 20fps camera to capture the video. I used a software to get pics from the video at a rate less than half the predicted period of the wave. Also took advantage of slow replay of the video. 20fps camera did a good job. i got and plotted the readings to identify a significant wave height and peak period. the plotted wave gave indications of reflection from the walls of the wave tank. it was the most economical solution for me.

Thank you all.