SparkFun Forums

[alex.forencich]

I have been mulling this over for a while and I think it may be possible to build a high resolution, high accuracy distance sensor (better than 5mm) with a high update rate (more than 1,000 samples/sec) at a relatively low cost (possibly $200-300) in the form of a laser rangefinder. Operation would be based on transmitting a high frequency modulated laser beam and then measuring the phase shift of the returned light. I will have to work through the design to be sure, but it could well be doable on a single board of about 3x5 in.

By the way, you're not the first person to think of this (I myself followed this research some months ago):

http://forums.parallax.com/showthread.p ... nge+finder

And:

http://forums.parallax.com/showthread.p ... evaluation

Its good to note, this project is underway in South Africa. So the founder is having a bunch of problems with US import laws regarding his system. Hes using a high power laser thats not easily available here, so this may be an issue in your project. Im excited to see progress on your project; maybe we can bring the price tag down AND make it available to folks here in the US.

I saw that particular product already. I think the links were posted earlier in this same post. It does look like a pretty nice product, but I'm thinking about trying something rather different; using a phase measurement instead of a direct time of flight measurement. I am primarily targeting high precision, short-range measurements so the design space is a little different. For one, I want to use a very basic laser diode. If I can get it to work with a $1 made in china laser diode, then that would simplify construction and part sourcing as well as reduce the price.

I have been able to modulate very cheap laser diodes at 500 MHz on my workbench with an RF signal generator and receive the signal with an OE converter. The next step is to receive the reflected signal with a photodiode detector and low noise amplifier chain. However, isolation and filtering are a big problem and the filter I built did not work (very high loss) because the component values in my assortment were not correctly marked. I got sidetracked trying to come up with a way to accurately measure caps on the order of 1 pF so I could build another filter with the correct component values.

[jremington]

A couple of years ago SparkFun issued a challenge to hack the Fluke 411D handheld laser tape measure (accurate to 5 mm up to about 25 m) but so far no one seems to have succeeded. See http://www.sparkfun.com/tutorials/323

However, more recently a bunch of cheaper imitations have become available, and one of them, the Uni-T UT380B (available for $60 from dx.com), has an easily accessible serial port that outputs the measurement. Range finding can be initiated via the keyboard connector. Here are the details: http://blog.qartis.com/arduino-laser-distance-meter/

Edit: the UT380B is supposedly accurate to +/- 2 mm, up to 45 m.

I've ordered mine!

[UNTEngineer]

Oh. Ok.

Well my application was a security system with long range motion and direction tracking. I stopped work on it for now as its not my area of expertise.

[alex.forencich]

A couple of years ago SparkFun issued a challenge to hack the Fluke 411D handheld laser tape measure (accurate to 5 mm up to about 25 m) but so far no one seems to have succeeded. See http://www.sparkfun.com/tutorials/323

However, more recently a bunch of cheaper imitations have become available, and one of them, the Uni-T UT380B (available for $60 from dx.com), has an easily accessible serial port that outputs the measurement. Range finding can be initiated via the keyboard connector. Here are the details: http://blog.qartis.com/arduino-laser-distance-meter/

Edit: the UT380B is supposedly accurate to +/- 2 mm, up to 45 m.

I've ordered mine!

Right. How many measurements per second can you take with those sensors? The one I am planning on building should be able to trivially take at least 1000 full resolution measurements per second. With some high speed ADCs and an FPGA, it may be possible to get up to 1 million measurements per second. Most other rangefinders that I have seen run at far less than 10 measurements per second. Perhaps I should not call this a rangefinder so much as a lidar unit.

[jremington]

The blog I linked suggests that 3 measurements/second are possible with UT380B. I'll know for sure after mine arrives, but that data rate and price are just fine for my particular autonomous robot application.

[Gnibur]

As a hobby-project I'm trying to build an outdoor mapping robot. To be able to do this, I therefore need a range finder which can measure distances up to 30m. In addition the measurements have to be fast, which makes it possible to rotate the sensor at some speed.

[alex.forencich]

Well, how fast would you like to be able to make measurements and what sort of a budget are you looking at?

[Gnibur]

To be able to move the robot at a reasonable, one circular reading per second would be good enough. Therefore an updaterate of about 350Hz would be enough.
As I'm doing this as hobby-project, I don't think I will spend more then 200-250 usd for this sensor.

[fusedlightning]

Alex,

Sounds like an interesting project, to say the least.

Looks like something similar is patented:
http://www.google.com/patents/US5164733

Basically, they use a LASER as a carrier wave for a set of modulation frequencies, and observe phase differences on the frequency bands to get distance. I will be the first to say that I don't know enough about the hardware behind this to design such a system!

However, I do have an Imaging background, and see a possible extension of what you suggest. Instead of using a single modulation frequency on top of your carrier (the LASER frequency), you could use a distributed set of frequencies (for example, a chirp function). This (partially) removes the problem of phase ambiguity - you now have information distributed across a range of frequencies, at the expense of temporal resolution. Depending on the high and low frequencies of the chirp that you select, you would be able to change your maximum range, as well as your granularity.

Depending on how you were planning to do your phase measurement, the same technique may still work. Possibilities would be an optical correlation, an analog hardware correlation after a photodiode, or, if you can sample fast enough, a Fourier or Wavelet analysis inside a DSP or ASIC/FPGA.

There is a reason systems like this aren't used in industry: Power. Most long-range LiDAR systems use pulsed LASERs because of the power requirements of running a continuous wave LASER at any sort of range. The amount of return on a standard pulse is miniscule, because most objects do not have very directed backscatter. This means that while the laser energy is relatively unattenuated on the path to the object, a much larger amount is lost (close to d^2) on the return path. Further, the amount of radiation necessary could make a continuous wave LASER very dangerous for large distances.

That being said, this may still be a great way to build a low-cost LiDAR system for moderate distance applications.

[rxtx]

There is an interface board for the Fluke 414D available from Porcupine Labs. The board adds a USB port and a TTL serial port to the laser rangefinder. http://www.porcupinelabs.com/lr4/