Ultrasonic tank level measurement (SRF005 ?)

D

davidwf

Senior Member
I am investigating the possibility of using a SRF005 based ultrasonic module to accurately measure the amount of heating oil in my tank and send it via a wireless link to a receiver - yes, I am aware that commercial units are available but their resolution is extremely low (e.g. in tenths of the tank) and they are quite expensive.....and of course I am a PICAXE fan !

The spec sheet says ....The length of the echo pulse is then divided by 5.8 to give a value in cm.......could it I wonder be modified to give an output in mm or tenths of mm perhaps ?
My 1385 lt tank is 6ft x 4ft x 2ft which equates to approximately 1.1361 lt/mm or 0.8801 mm/lt.

My idea is not only to indicate the actual level of oil remaining in the tank but also to be able to plot the consumption on a graph (probably via Excel) so the measurement head needs to have quite high resolution.

Anyone had any similar ideas ?
Any other suggestions.....apart from to get a life that is :)
 
Last edited:
A

Andrew Cowan

Senior Member
I'd imagine the sensor lacks the resolution to produce mm-accuracy results. I've found that it gives pretty repeatable results in cm mode.

A
 
E

eclectic

Moderator
@David.

1. I'll concur with Andrew re mm.
2. The idea has occurred on the Forum before.
3. Caveats:
The SRF005 is NOT waterproof, and it cannot
AFAIK, be covered in any way.

e
 
D

davidwf

Senior Member
I appreciate that it is not waterproof, I was thinking to mount it facing downwards so it would reflect off the top surface of the oil and could therefore be covered and sealed on the remaining sides & base.
 
E

eclectic

Moderator
davidwf said:
I appreciate that it is not waterproof, I was thinking to mount it facing downwards so it would reflect off the top surface of the oil and could therefore be covered and sealed on the remaining sides & base.
Click to expand...
Go for it. :)

I'm sure that there are others on the Forum
who'll appreciate your research.

I'm assuming that the tank is static,
and I don't know what the legal/safety position
is with regards to "electricity" in oil tanks

e
 
D

Dippy

Moderator
"Would it be possible to measure the amount of oil through weighing the entire tank?"
- Of course it would.
So, you'll want something tough and reliable and weatherproof and then make some new mounting points for the tank.
Now, what was the budget again?


Don't you have a transparent tube showing level outside the tank?

I think any home-made U/sonic device will be low grade.
Durability?
How will you protect the transducers?
There is (was?) an article on the Devantech site about SRFs - it mentions that swapping those t/ducers for weatherproof versions didn't work in their tests.
Please don't suggest clingfilm - if you then I'll know you are mental :eek:

How about white noise or sweep Fq then doing a resonance analysis of the returned signal? That will give you a value of the enclosed 'air' volume.(Kidding).
Or viewing a transparent tube?
Can you use something capacitive?
Maybe a Float on a rotary encoder or hall effect sensor? (something robust)

Think of something good ... quickly - or else some nutter will suggest a float on a stick connected to a 5p Ebay potentiometer :)


Whatever you do, please avoid basing it on the moths in your wallet.
For reliability and trust-worthiness DO NOT buy the Eb-cheapest bits out of a Beijing Wheelie Bin.
 
J

Jarubell

Senior Member
My 2 cents would be to use load cells, that is going to be my plan one day but I have much more to learn first.:) Oh, and I'm thinking about just weighing half the tank.
 
Last edited:
techElder

techElder

Well-known member
Different materials have different reflective characteristics to "ultrasound", so it will be well worth you experimenting outside the tank to see if you can get an adequate reflection from the oil.

My experience is that the tank of oil will be a poor reflector compared to a hard surface and will require some adjustment to either the transmitter's power or the receiver's gain to maintain some linearity in the measurements.
 
H

Haku

Senior Member
Well, so long as my ideas aren't being taken seriously, what about this:

A plastic tube with a float inside, inside the float a small rare earth magnet, stuck (& sealed) on the outside of the tube a few hall effect sensors to help measure where the magnet is. The whole tube stands in the oil with holes at the bottom to let the oil in.

Or failing that, gamma ray? ;)
 
E

eclectic

Moderator
Perhaps David can allay speculation
by telling us

a. The size of the tank
(Tiddly -Litres or Mega-litre industrial)

b. What sort of oil.

e
 
S

Svejk

Senior Member
From post #1

My 1385 lt tank is 6ft x 4ft x 2ft which equates to approximately 1.1361 lt/mm or 0.8801 mm/lt.
Click to expand...
Anyway, I'd go for the float and pot setup without knowing more about it. Perhaps a photo?
 
E

eclectic

Moderator
Svejk said:
From post #1



Anyway, I'd go for the float and pot setup without knowing more about it. Perhaps a photo?
Click to expand...
That was added as an edit.
It really would get VERY tedious to re-read
every post, every time one visited.

e

As in No chance.
This edit may be missed.
 
Last edited:
premelec

premelec

Senior Member
Assuming the heating oil is of fairly repeatable density [likely to not clog jets :-0 ] a pressure transducer and bubbler pipe to the bottom can work - you keep pumping a little gas down to the bottom and the pressure represents the oil height - the geometry of the tank is not stated unless it's rectangular [I've seen a lot of oval and cylindrical ones]. Anyhow if the height of the oil is a direct measure of volume no problem... other geometries more interesting... :)
 
J

Jarubell

Senior Member
A bubble tube would work great but now you need a constant air supply. I thought about that but my oil tank is in the house and a compressor going off in the middle of the night would suck.

I do agree with BB!
 
S

Svejk

Senior Member
@eclectic: no harm was intended, I apologise if any was done. I've seen that the post was edited, the "From post #1" was there just to point to the information.

Regards,
Svejk
 
premelec

premelec

Senior Member
Stirriing the pot?

Well the idea of a large air supply in a bubbler tube would be good for stirring the oil but not needed for pressure measurement - assuming slow use of the oil... one bubble per minute can do it... a 2 liter bottle storage unit and a salvaged battery operated blood pressure pump could be the fix - and if the compressor fails you get an empty reading and alarm which is kind of 'fail safe' - better locate the tube gravitationally above the oil - no nasty siphon effects then possible... and, yes, a vent on the tank to prevent pressure build up and probably several other measures I haven't considered... whatever works, tank volume sensing has been going on for hundreds, thousands... many years - and now with a PICAXE!
 
J

Jarubell

Senior Member
The few bubble tubes I have installed and/or played with in the industrial world had the air supply come from a 80psi+ header, stepped down then through a varible area flow meter, so the use of a much smaller compressor as you discribe has me interested in that setup.

Thanks for helping me take off my blinders!:D
 
H

Haku

Senior Member
There's always the "Waterworld" method - stick an old man in a rowboat with a measuring stick in the oil, then occasionally open up and shout down asking how much black stuff there still is.


I'll get me coat.
 
P

pilko

Senior Member
You don't need a Bubble Tube. Just stick the pressure sensor in a flexible, liquid-proof container, force it to the bottom and calibrate it for depth.------- Bob's your uncle.

pilko
 
D

davidwf

Senior Member
wow....23 replies in a day....:)

...some good ideas - and of course some comical ones - apart from my original ultrasonic idea the pressure transducer in a sealed tube seems feasible but I don't think it would give the accuracy or resolution I am after.
How about pulsing an i.r. beam onto the surface and measuring the time it takes to receive a reflected signal back ?

A weighted float on a piece of string attached to a rotary encoder i.e. potientometer which would sit above and outside of the tank seems to be the overall best suggestion - that is exactly how the mechanical gauges work and using a 10 turn pot or similar would, I feel be quite accurate and fairly simple to code.

The concerns over sealing the transducer are not insurmountable, my steel tank has 2 x 2inch filling holes right at the the top so the monitor could be fitted over one of them and sealed against the weather.

Thanks for the ideas....any more input welcomed

DF
 
D

Dippy

Moderator
Not only weatherproofing from the outside world , but sealing the transducer element from weather and fumes. Nice warm fuems on a hot summer day ... mmm.... lovely.:rolleyes:
Those usual cheap'n'cheerful ones have excellent performance for the cost, but are OPEN. If you are confident that they will survive long periods then good luck.
Maybe there are others using the closed/sealed type of transducer - looking around (or designing) is your job.

Don't get me wrong, I think it's the easiest and safest method, but as a designer, I like to look at all the options rather than just reaching for the first (cheapest) one off the shelf.



Have you done ANY calcs on how the SRF range accuracy/resolution impacts on the oil level accuracy?
What sort of accuracy/precision are you expecting? And why?
Surely you understand that with ANY sensing system you will be limited by the Sensor type and the physical mechanics and YOUR ability to extract , process and interpret the 'signal'.

Quick examples:-
http://www.sensor-systems.com/brochures/domestic/sonic/Watchmansonic_DL_HR.pdf
http://www.intellitank.co.uk/alarms.html
http://www.tankservices.co.uk/oil-fuel-tanks-for-sale.php?tanks=473


Time for speed of light?
I take it your Physics and Maths and Electronics experience is at the start of the curve?
For a home-build forget it.
Some time ago I discussed with Jeremy Leach about the possibility of doing a home-made laser range finder using phase and pulse logic. It's beyond the scope of the Forum I reckon but an analogy (and hippy likes those) would be the optical linear measuring scales used in DROs on milling machines etc.

Hasn't this question cropped up before years ago? - I can't remember if any conclusions were drawn or whether it faded out in true Forum style :confused:


Whatever you do must be robust , safe and reliable and NOT built on the cheap.
Where possible, avoid moving mechanical stuff, though good quality (NOT cheap) pots/encoders should be OK . Sloppy mechanisms made using Dad's pillar drill and bent CHCH drills should be avoided at all costs.
Unless, of course, you enjoy fixing it twice a week and then check oil level manually as you can't trust it. ;)
 
Last edited:
S

StigOfTheDump

Senior Member
davidwf said:
wow....23 replies in a day....:)



The concerns over sealing the transducer are not insurmountable, my steel tank has 2 x 2inch filling holes right at the the top so the monitor could be fitted over one of them and sealed against the weather.



DF
Click to expand...
Can you not just screw a pressure transducer in one of these holes?

Edit forget that. I thought the holes were at the bottom
 
Last edited:
P

pilko

Senior Member
Use a 0-2psi pressure sensor. It should give you enough resolution.
(Instal it in the tank --You will need to force it to the bottom, as it will try to float)

pilko
 
J

Jarubell

Senior Member
pilko said:
You don't need a Bubble Tube. Just stick the pressure sensor in a flexible, liquid-proof container, force it to the bottom and calibrate it for depth.------- Bob's your uncle.

pilko
Click to expand...
I know the range is small but would that transmitter require an atmosphere reference like a breathing tube?

Does anyone have concerns having these non rated electronic devices in the oil tank?
 
D

Dippy

Moderator
"would that transmitter require an atmosphere reference"
- eh? Are you on about the sensor really?
It may depend on the sensor type and the reference it uses.

Concerns?
Yes, but I've given up worrying. Mainly as none of you live anywhere near me ;)
It just pains me that people will penny-pinch and bodge to save 50 pennies, then spend ages fixing and repairing afterwards. False economy at the very least.


Buoyancy may be another option?


I'd be tempted to see what techniques are commercially available. Mainly because people with healthy brains and budgets have developed ideas that can be made at a resonable cost. i.e. what you all are trying to do but with the facilities to design.
Don't forget that this is all old hat and the problems have been solved, so best to base yours on a commerically available product.
 
premelec

premelec

Senior Member
Bouyantly

@Dippy... bouyancy - I like that - a long hollow sealed cylinder pressing up on a force sense unit or a solenoid servo balance where current in the solenoid relates to length of the cylinder immersed... some interesting equations... whatever works... Rube Goldberg lives!

OK - this seems to work better as a thought than a practice... long floaty cylinders in oil not that much fun to assemble.

There's still the two insulated parallel wires probe which changes capacitance proportionate to what length is in the oil - with a side sense unit checking the particular dielectric constant of the current batch of oil :)

I think a sight glass with a servo controlled up down slider to look at and report the sight glass level might be fun - and encourage cleaning of the glass more often... [BTW could be a siphoned sight glass if you don't want to puncture the tank sides].... so many opportunities for erronious readings so little time!
 
Last edited:
D

davidwf

Senior Member
Dippy said:
........ people will penny-pinch and bodge to save 50 pennies.......
Click to expand...
hey, not me....never have, never will :mad: I KNOW it doesn't pay.

........ base yours on a commerically available product........
Click to expand...
There isn't one at an affordable price...... hence my interest !

Like I said earlier, the only commercial gauges that are generally available use a weighted float and piece of string (posh string) onto a geared dial type indicator and give a resolution of typically 1/4 or at best 10th tank intervals.

I therefore thought that a PIC based product would be ideal

Dippy said:
Have you done ANY calcs on how the SRF range accuracy/resolution impacts on the oil level accuracy?
Click to expand...
...no, I haven't...I was kinda hoping someone else had :)

Dippy said:
What sort of accuracy/precision are you expecting? And why?
Click to expand...
Whilst I don't need super accuracy I thought it would be nice to be able to see how much the central heating it is costing to run on a day by day basis - we have electric meters, we have gas meters we have water meters so why not an oil tank meter ?
 
D

Dippy

Moderator
Well, whatever floats your boat :)

Rube Goldberg? Is he the chappy who copied Heath Robinson?

Oh David.
You'll have the resolution spec of those sensors.
Therefore you'll know know the reported detected range resolution and variation.
You'll have the size of your tank.
If the reported range varies by (e.g.) 1cm you can easily calculate the variation in the volume measured.
Come on Mate, this is kid's stuff :(


Ah, at last... a reason for wanting to know.
Unless you go for a flow meter, then I can't see U/sonic being good enough unless you have something really fancy which I suspect is a tad beyond your technical experience.
Use one of the other suggested methods.

Do you NOT have a see-through level indicator on outside of tank?

I think I'd research various professional methods as opposed to bodging something together.
Pressure / weight / buoyancy / float-height would be the areas I'd consider, but you'll need some mighty fine resolution for real-time consumption metering.
I know you don't say Real-time metering at this stage, but if you ever do then I can't see anything better than genuine Flow-Metering for that.
 
S

SAborn

Senior Member
I know Dippy didnt favor the idea of using encoders and yes they can have problems.

A encoder disc i have used from a old bubblejet printer (canon) gives excellent readings, and with a little engineering and some posh string attached to a float in a vertical tube you would get your sub mm readings i would think.

I used one in an aneomometer i built so it could read very low wind speed over a short interval.

Heres a photo.

Edit:- should really call it a photo interupter rather than an encoder.
 

Attachments

Last edited:
P

pilko

Senior Member
If you just want to know your oil consumption,use your oil pump and time.
GPM X Minutes. Most oil pumps are gear pumps and are very consistant. The GPM can be obtained from the pump and the burner nozzle.
I disconnected my oil line and ran the oil into a container and measured it.
In my case I needed to know at what outdoor temp. it is more efficient to burn oil or run my Heat Pump.

pilko
 
Last edited:
D

Dippy

Moderator
"I know Dippy didnt favor the idea of using encoders and yes they can have problems."
Come on SABorn ... keep up mate!

Post #7 Dippy says "Maybe a Float on a rotary encoder or hall effect sensor? (something robust)"


Sorry, but I'd want something a bit more roughty-toughty in or near my oil tank than a plastic encoder (or interrupter) in a plastic mount. And you're asking trouble with a non-absolute sensor and novice programming.

Why bodge when you get something butch for a few quid? (This is what frightens me about this Forum sometimes. Bodgit & Scarper are the two most unreliable workers.)

Anyway, so many answers - so good luck with the project.
I hope you actually get around to doing it.
Many people get (understandably) duanted when their initial simple idea get's expanded . Then they throw their hands up in horror all the ideas are wasted.
 
E

eclectic

Moderator
E

eclectic

Moderator
P

pilko

Senior Member
You don't need anything fancy, just use the AXE to multiply the Gallons per Minute of the pump by the elapsed running time in minutes.
--I do it this way
--It works
--It is accurate
--It is repeatable

pilko
 
You must log in or register to reply here.
Top