ultrasinic accuracy

[alireza]

hi everybody
I want to build a device to measure the distances with ultrasonic sensors. It sould work with the time of flight method. The accuracy of the measuring must be 0.05 mm.
Is this accessible?
How much accuracy can I achieve it?
What will be the maximum accuracy of this device?
What should be the frequency of the sensors?

 

[blueroomelectronics]

0.05mm not with ultrasonics. Maybe a few cm.

 

[stevez]

The accuracy will be governed by a number of thing. You might get a sense of how challenging your specification is by doing a little math. Take the most likely situation and determine, approximately, the time period you need to measure. With that approximation you would then need to estimate how precisely you might actually be able to measure that time - within the limits of your resources. That would allow you to be able to calculate the variation in distance from just the time error alone. After that you need to understand how variable the speed of sound/ultrasound is for the conditions you might encounter - this will add to the challenge.

 

[dknguyen]

Accuracy gets better as the frequency gets higher. But range gets shorter as frequency gets higher. So really high accuracy only works are really REALLY close range. So a transducer with 0.05mm accuracy probably won't work at distances farther than 5mm.

For reference, there is a 2MHz transducer out there that claims 25 micron resolution (not accuracy) in air, but does not say in what conditions.

Also, at longer range there's much higher chance for worse conditions which makes readings much less accurate since there is more distance for noise and disturbance to get into the signal.

Needless to say, transducers that can do this are very expensive.

 

[alireza]

Accuracy gets better as the frequency gets higher. But range gets shorter as frequency gets higher. So really high accuracy only works are really REALLY close range. So a transducer with 0.05mm accuracy probably won't work at distances farther than 5mm.

Hi dknguyen
What is the relation between the range and the accuracy?
Maybe I want to optimize them.

 

[Leftyretro]

I would think that ultrasonic waves being high frequency sound waves would see variations in air temperature and barometric pressure as one limiting factor in accuracy.

Lefty

 

[dknguyen]

I don't know the specific relationship.

 

[alireza]

I would think that ultrasonic waves being high frequency sound waves would see variations in air temperature and barometric pressure as one limiting factor in accuracy.

Lefty

Hi Leftyretro
If I can measure the speed of sound exactly and simultaneously online, what will be the best accuary?

 

[Hero999]

I would think that ultrasonic waves being high frequency sound waves would see variations in air temperature and barometric pressure as one limiting factor in accuracy.

Lefty

It's possible to compensate for pressure and temperature.

I suppose it won't be as accurate if the air is moving.

 

[JimB]

I suppose it won't be as accurate if the air is moving.

There are such things as ultrasonic anemometers,
Ultrasonic Anemometers | GILL
which work by measuring the travel time of an ultrasonic pulse in air.
So, any air currents in the vicinity and the 0.05mm accuracy goes straight out of the window.

JimB

 

[alireza]

So, any air currents in the vicinity and the 0.05mm accuracy goes straight out of the window.

JimB

Hi JimB
I dont understand your mean. would you explain it?

 

[duffy]

The ultrasonic waves travel faster with the air, and slower against it. .05mm resolution is asking a lot from ultrasonics.

 

[JimB]

It means you have no chance of 0.05mm accuracy.

JimB

 

[Mikebits]

.05mm is... Well... wow...

 

[alireza]

Yes, I satisfy that I cant reach the 0.05mm accury.
So, what is the best accuracy that I can achieve it?

 

[duffy]

I would imagine if you are careful you could get at least one wavelength. At 40khz, standard air pressure and humidity (not sure how Iran compares) that would be about 8.5mm.

You might be able to get less, but probably not much less. To increase accuracy, you probably want a reference beam to constantly compare the measuring beam against. This would help you compensate for changes in atmospheric conditions and airflow.

 

[blueroomelectronics]

It's got to be for his version of one of these.
Coordinate-measuring machine - Wikipedia, the free encyclopedia

Why not use lasers as the modern units in the article use?

 

[Hero999]

Yes, I satisfy that I cant reach the 0.05mm accury.
So, what is the best accuracy that I can achieve it?

0.05mm is possible at sort distances and only if you want to spend a lot of money. The actually accuracy you'll get will depend on the frequency used, the wind speed and other factors such as humidity.

 

[duffy]

Looking at specular reflectivity, a .05mm feature size would equate to a wavelength of about 6.5mhz. Will air-coupled ultrasonics work at this frequency? I couldn't even find a graph that went that high. This page shows one to 2mhz -
air-coupled ultrasound
- and it shows a crazy SEVEN HUNDRED db/m attenuation at 2mhz. Looking at that curve and trying to project upwards to 6.5mhz, seems like you would have a better chance of detecting a gnat's fart on Jupiter.

 

[Hero999]

You can measure distances to a greater accuracy than one wavelength. The phase shift can also be measured which will give information of a higher resolution than one wavelength: for example if a wave takes 2.5 periods to return it will be 180° out of phase.

As far as I'm aware the attenuation of sound remains the same over a constant number of wavelengths. For example a 1Hz infrasonic wave can travel one million times as far as a 1MHz ultrasonic wave given the same level of attenuation.

EDIT:
According to the grap it becomes exponential at higher frequencies, so I stand corrected.