URM37 V3.2 works fine but V4.0 doesn't- DFRobot Forum

2018-01-03 21:59:47 How can I solve the problem with the existing units? They are so unreliable that I could not say they are functional. Pribicevic Zoran

2018-01-02 11:09:23 V3.2 and V4.0 uses different ultrasonic probe, and it is easily influenced by the surface of the target, we are developing a new version of SEN0001, the new version would solve the problem. robert.chen

2017-12-28 08:07:53 I tried another URM37 V4.0, for the same soft target at about 1/2 m:
Distance=57.44cm
Distance=57.44cm
Distance=57.44cm
Distance=58.16cm
Distance=0.00cm
Distance=0.00cm
Distance=0.00cm
Distance=0.00cm
Distance=0.00cm
Distance=0.00cm
Distance=0.00cm
Distance=0.00cm
Distance=0.00cm
Distance=0.00cm
Distance=0.00cm
Distance=0.00cm
Distance=0.00cm
Distance=0.00cm
Distance=0.00cm
Distance=0.00cm
Distance=0.00cm
Distance=56.00cm
Distance=56.00cm
Distance=56.00cm
Distance=56.72cm
Distance=56.00cm
Distance=56.72cm
Distance=54.57cm

Port closed

Now URM37 V3.2 in the PWM mode, otherwise the same setup and your sketch:
Distance Measured=56cm
Distance Measured=54cm
Distance Measured=53cm
Distance Measured=55cm
Distance Measured=52cm
Distance Measured=59cm
Distance Measured=54cm
Distance Measured=54cm
Distance Measured=54cm
Distance Measured=54cm
Distance Measured=54cm
Distance Measured=60cm
Distance Measured=61cm
Distance Measured=61cm
Distance Measured=54cm
Distance Measured=54cm
Distance Measured=54cm
Distance Measured=55cm
Distance Measured=55cm
Distance Measured=53cm
Distance Measured=54cm
Distance Measured=54cm
Distance Measured=55cm
Distance Measured=55cm
Distance Measured=60cm
Distance Measured=61cm
Distance Measured=60cm
Distance Measured=60cm
Distance Measured=59cm
Distance Measured=55cm
Distance Measured=54cm
Distance Measured=54cm
Distance Measured=54cm
Distance Measured=56cm
Distance Measured=53cm
Distance Measured=52cm
Distance Measured=52cm
Distance Measured=52cm
Distance Measured=52cm
Distance Measured=52cm
Distance Measured=52cm
Distance Measured=51cm
Distance Measured=52cm
Distance Measured=51cm
Distance Measured=53cm
Distance Measured=52cm
Distance Measured=52cm
Distance Measured=52cm
Distance Measured=52cm
Distance Measured=52cm
Distance Measured=51cm
Distance Measured=51cm
Distance Measured=52cm
Distance Measured=52cm
Distance Measured=52cm
Distance Measured=52cm
Distance Measured=52cm
Distance Measured=51cm

Port closed

After hundreds of readings I didn't get any error. Pribicevic Zoran

2017-12-28 05:02:57 Yes, there is no problem with a perfectly sound-reflective surface, like a PC screen. Here are my measurements, very precise and stable:
Distance=13.64cm
Distance=13.64cm
Distance=13.64cm
Distance=13.64cm
Distance=12.92cm
Distance=14.36cm
Distance=13.64cm
Distance=13.64cm
Distance=12.92cm
Distance=13.64cm
Distance=13.64cm
Distance=14.36cm
Distance=12.92cm
Distance=12.92cm
Distance=12.92cm
Distance=13.64cm
Distance=13.64cm
Distance=12.92cm
Distance=13.64cm
Distance=13.64cm

Port closed

However, when I point the sensor into a less perfect target, like myself, at about 1/2 m, here are the results:
Distance=1.44cm
Distance=2.15cm
Distance=2.15cm
Distance=2.15cm
Distance=2.15cm
Distance=2.15cm
Distance=2.87cm
Distance=2.15cm
Distance=2.15cm
Distance=2.87cm
Distance=1.44cm
Distance=2.15cm
Distance=2.15cm
Distance=1.44cm
Distance=2.15cm
Distance=2.15cm
Distance=50.26cm
Distance=50.26cm
Distance=2.15cm
Distance=2.15cm
Distance=2.15cm
Distance=2.15cm
Distance=2.15cm
Distance=2.15cm
Distance=2.15cm
Distance=2.15cm
Distance=2.15cm
Distance=2.87cm
Distance=2.15cm
Distance=2.15cm
Distance=1.44cm
Distance=2.15cm
Distance=2.15cm
Distance=2.15cm
Distance=2.15cm
Distance=2.15cm

Port closed

That cannot be right. It should be possible to detect an obstacle of human size without so many errors. Try the very good Devantech SRF08 you sell and it will not have that behavior. Pribicevic Zoran

2017-12-27 11:20:57 I did a test with PWM trig mode, and the output distance is right and stable, I used a ruler and make the sensor measure the distance to the PC screen. So what is the circumstance of your measurement, you need to make the object clear and flat. robert.chen

2017-12-27 04:31:30 I was wrong. After more testing, I found the auto mode unreliable as well. For example, if a big object (me) is 40 cm apart, it will sometimes reliably report the distance as 2 cm, 30 times in a row. Sometimes it reports 0 cm. I used analog output this time. As it looked like an echo problem, I sent over the UART a sequence to increase the time between the measurements to 100 ms (0x44, 0x04, 0x64, 0xAC - I hope that is correct). There was no change.
BTW, your PWM example using Uno is not convenient since the only Serial is used for the communication to the computer and the sketch will not work. Pribicevic Zoran

2017-12-25 23:46:15 It works. Auto Measure Mode works fine. At least, when there is no close obstacle, it almost never reports one and I can say it is reliable. When the obstacle is more than 50 cm apart it is more and more sensitive to obstacle angle. At 1 m, even 20 to 30 degrees is a problem and it reports "Invalid" quite often. However, for my application, it is not so important. Pribicevic Zoran

2017-12-25 20:16:41 I tried the sketch without modifications, except trying Measure as true and false. Both results, analog and PWM showed similar anomalies. Therefore, only PWM trigger mode. I will try the Auto Measure Mode and will give you the results. Pribicevic Zoran

2017-12-25 11:07:53 Have you tried all three measurement mode, it seems the auto measure method works fine. robert.chen

2017-12-23 23:02:37 I have a number of V3.2 sensors and now ordered 10 pcs. of V4.0. However, the readings are quite unreliable. Your program in page https://www.dfrobot.com/wiki/index.php/ ... U:SEN0001), for PWM trigger mode, often repeatedly finds distance 10 cm where the actual distance is 1.5 m or returns invalid distance. Exchanging a V4.0 for V3.2 solves the problem and all the reading are fine. Pribicevic Zoran