Notifications
Clear all
Archived Suggestions
7
Posts
4
Users
0
Likes
1,123
Views
Topic starter
2022-01-18 8:29 am
All
I have pondered cheaply measuring longer distances with good accuracy using two part ultrasonic transducers.
The common transducers as I understand it generate a pulse and measure the time for an echo to be returned. Both transducers mounted close together. Some firmware (not too sure of the details - functional description a bit Chinese) to discriminate the pulses. I have tried these and they are OK for a few meters.
I have considered simply removing the receiving transducer (not as easy as it sounds - wrecked mine and have not got back to it) and hard wiring it to the object to be measured. Clearly not very handy but... It would then generate the pulse and measure the time for it to reach the receiver - no echo. I expect that this would be much more sensitive and more reliable as the signal is direct and it would be less effected by spurious echoes etc. I presume the firmware on a standard module would not object.
The more ambitious approach would be to use say an ESP32 or other radio module to sense the initiation by radio and time the much slower audio signal and transmit the time back to a master. This would allow much more flexible application. Selecting a radio with predictable and short latency may be a problem.
An application I have in mind is a large (10 + meters) cable delta actuator. I would like to get an accuracy of a few cm. Might need compensation for pressure and temperature. 3 poles with simple winches each reporting a distance to the load (with the transmitter). Software could then solve the movement commands to the winches (again not as easy as it sounds - solving the intersection of spheres) without needing encoders and compensating for cable winding errors. One of the cables may be a hose for precision watering.
Has anybody tried this approach? Of any general interest?
I am slack. Much easier to ask the question than to do it!
It could of course be all radio (think GPS) but I think the timing precision would require very special purpose hardware.
Thanks
Patrick
PWD
2022-01-18 1:19 pm
What about using a camera, and using some form of machine learning to estimate distance? Is that feature available in existing s/w?
2022-01-18 7:53 pm
have considered simply removing the receiving transducer (not as easy as it sounds - wrecked mine and have not got back to it) and hard wiring it to the object to be measured. Clearly not very handy but... It would then generate the pulse and measure the time for it to reach the receiver - no echo.
No reason to remove the receiving transducer. I assume you are referencing a HC-SR04 Ultrasonic Module? Just use the transmitter transducer on one device to send the signal and the receiver transducer on the other device to receive the signal.
Topic starter
2022-01-19 1:37 am
Thanks for your interest.
I guess you could use 2 full modules or just work harder at getting the receiver off (or transmitter - doesn't matter which.) I think on the module I tried the pins of the transducers were a tight fit into holes or I wasn't getting the solder off well enough.
The radio link is the tricky bit.
Thanks
Patrick
PWD
Topic starter
2022-01-19 1:39 am
I expect that that would be sledge hammer to smash a nut.
I would be after reasonable accuracy of a few cm.
Thankjs
Patrick
PWD
2022-01-19 12:29 pm
Thanks for your interest.
I guess you could use 2 full modules or just work harder at getting the receiver off (or transmitter - doesn't matter which.) I think on the module I tried the pins of the transducers were a tight fit into holes or I wasn't getting the solder off well enough.
The radio link is the tricky bit.
Thanks
Patrick
So true Patrick! Sometimes using a sledgehammer can be fun though. 🙂