SRF005 Ultrasonic Range Sensor

[clockwork]

The specification for the SRF005 states Max Range 3 metres. Also Sensiivity - Will detect a 3cm diameter broom handle at > 3 metres.

Does anyboy with practical experience of this Sensor have any idea how far away the sensor would be capable of reliably detecting say a wall
or door opening (with no doubt lower levels of accuracy?) Or is the maximum detection distance governed by time constraints? A maximum
length of output pulse from the sensor for example.

Clockwork

 

[westaust55]

The Robot Electronics website run by Devantec the manufacturers of the SRF005 and SRF008 Untrasonic sensors
suggests for the SRF005 that the range is up to 4 metres.

For the PICAXE using PULSIN, the value resulting could be up to 65535 but if using the Rev Ed calculation and a 4 MHz clock speed, the forumla is
PULSIN * 10 / 58
Hence the max acceptable result from PULSIN is 6553 otherwise there will be overflow (> 65535).
This gives a max value without overflow of 1129.8 cm (= 11.29 metres) with a 4 MHz clock speed.

If you operate at 8 MHz that max theoretic range halves to ~ 5.6 metres.

 

[AllyCat]

Hi,

I don't have any experience of those particular ultrasonic modules, but I can do a little maths (I hope accurately).

The speed of sound is about 300 metres/second, so the 30ms "timeout" of the modules is going to set an upper limit of about 10 metres (i.e. 5 metres to and from a wall). If they can detect a 3cm broom handle at 3 metres, then I would expect them to detect a flat object perhaps 5 cms across at a rather greater distance.

However, the wavelength of the (40 kHz) ultrasonic signal is about 8 mms. Consider a piece of string between the transducer and the nearest point on a wall 5 metres away. If you lengthen the string by about 2mm, it can touch a circle on the wall with a radius of about 140mm (using Pythagoras). Now any sound reflected back from points on that circle to the transducer would travel an extra half-wavelength and can potentially cancel out the waves from the centre of the circle.

I cannot say whether waves from that circle will actually be returned to the transducer, or if they would be reflected (like light on a mirror) to elsewhere in the room. But in either case, I think we can say that they're not "assisting" in the range measurement, so probably the wall is no easier to detect than an object 10 or at most 20 cms across.

Also, as you mention "opening a door", I do know that "ultrasonic intruder detectors" normally employ a rather different method. They normally transmit a "continuous" signal which sets up a "standing wave" pattern in the room and when any object moves, it changes the pattern and thus the signal strength at the ultrasonic receiver.

Cheers, Alan.

 

[Bill.b]

The Robot Electronics website run by Devantec the manufacturers of the SRF005 and SRF008 Untrasonic sensors
suggests for the SRF005 that the range is up to 4 metres

The link to this site appears to be off line or non-existant.

Regards Bill

 

[boriz]

I think the link was mangled during posting. Try this: http://www.robot-electronics.co.uk/htm/srf05tech.htm

 

[Dippy]

When I was having a play some years ago I found that model would easily sense my shed at 4m , but it got a little hit'n'miss at 5m.
But a) I don't know the acoustic properties of my shed and b) it was little windy.

The trouble is that ,as you approach the limit of any sensor, ambient conditions will have a greater effect on signal:noise.
Whilst maths is a great start the only true way is to try it in the conditions you wish to use it.
And if ambient conditions are variable then try and test it 'worst case'.

As to timing constraints, I really don't know. Maybe someone has checked this or maybe contact the manufacturers. Straight from horse's mouth is usually best.

Are you making a burglar alarm or occupancy detector or something like that?