Ultrasonic calibration

Frequent questions asked online, offline, in forums are answered here
Post Reply
palkovics.denes
Posts: 2
Joined: Fri Apr 09, 2021 11:59 am

Ultrasonic calibration

Post by palkovics.denes »

I am working in a self-driving car project and we want to use indoor "GPS" to track how well the cars (1:16 RC model cars) follow road lane.
We stumped upon a problem that sometimes the car calculated position jumps approximately 5-6 cm and in further investigation we found a strange phenomenon that the echo of the signal is stronger than the source. Clear eye-sight is provided when this happens. We looked for external noise but did not find any so we think this is not the case.

We did an experiment with these steps:
1. Put one beacon (say it's address is 13) on the wall.
2. Put another one (address is 3) on the ground.
3. Rotate the second beacon by 90° multiple times.
4. At each rotation look at the oscilloscope graph

The graphs of the oscilloscope after each 90° turn:
Image
Image
Image
Image
You can see on the last image that the echo is as strong as the source signal. Our assumption that the rx5 sonar output volume not as "loud" as the others and maybe the echo of other sonars seen on the oscilloscope.

My question is 1) our assumtion is correct and 2) how can we tackle this issue?

smoker77
Posts: 366
Joined: Sat Feb 06, 2016 4:03 pm

Re: Ultrasonic calibration

Post by smoker77 »

Hello,

1. Yes, your assumption about less loud RX5 is correct in case of HW4.9 beacons. Super-Beacon's emitting sensors are equally loud.
I think you are using HW4.9 beacons, yes?
2. Your signals look pretty weak for so small distance. Did you enable all sensors RX1...RX5 for transmitter and receiver?
3. Try to enable only central RX4 for transmitter and receiver. Orient the beacons face to face and check the signal.
Then enable other sensors and check the signal again.
4. In some cases when multiple sensors are enabled for RX and TX, signal can become more weak in rare points because transmitted or received signals kill each other because they are coming in opposite phases. This is why we recommend to disable unuseful sensors, for example sensors of stationary beacons oriented to ceiling or walls. Super-Beacons or mini-Rx don't affected by this on RX side, because their receiver is a single digital microphone instead of distributed RX1...RX5 sensors as for HW4.9
5. We don't recommend use 'Colored frequencies' mode in oscilloscope, it is exprerimental mode and have low resolution. Please use 'normal' mode.

palkovics.denes
Posts: 2
Joined: Fri Apr 09, 2021 11:59 am

Re: Ultrasonic calibration

Post by palkovics.denes »

1. Yes, we are using HW4.9 beacons.
2. About the weak signal: for looking the oscilloscope graph We set the gain to a fixed value instead of using AGC, and all we enabled all sensors.
4. If I understand correctly you suggest that switch off certain sensors on stationary beacons. What about the hedgehog beacon? If we turn on the RX4 only, then it sound towards the ceiling and not towards the stationary beacons on the wall.

Can we adjust the volume of each sensor so that the volume of the sensors becomes almost the same? I found a feature in dashboard: view -> Ultrasound self-calibration. Is this feature can be use for this purpose?

Post Reply