Optics and laser beam width, need help please. / Electromagnetic Radiation / Forums

Forums

4hv.org :: Forums :: Electromagnetic Radiation

« Previous topic | Next topic »

Optics and laser beam width, need help please.

previous 1 2 3 4 next last

Move Thread

LAN_403

Carbon_Rod

Tue Nov 27 2012, 08:16AM

Registered Member #65 Joined: Thu Feb 09 2006, 06:43AM
Location:
Posts: 1155

Round-trip-times for time-of-flight systems, and clock speed of counter.

1 centimeter -> 0.066ns -> 15.14 GHz
3 centimeter -> 0.200ns -> 5 GHz
15.24 centimeter -> 2.00ns -> 500MHz
1 meter -> 6.67ns -> 149.8MHz

$35 Clock speed of regular Raspberry Pi = 700Mhz
$24 1Gbps Optical Fiber Transceiver replacement module
$20 GRIN lens
$5 phase detector + digital down-converter
$12 fpga for burst-rate RTD counters

The dollar sign is in metric too.... =P

In theory, you could achieve 7.62 centimeter resolution, but those with too many detractors never prioritize winning.

Pinky's Brain

Tue Nov 27 2012, 09:01AM

Registered Member #2901 Joined: Thu Jun 03 2010, 01:25PM
Location:
Posts: 837

He isn't using time of flight, he is simply sending a laser beam straight ahead and then has a camera off axis. Reflections from different depths will appear at different lateral positions from the point of view of the camera.

He needs a large dof and infinity focused camera so he can't use a large lens (focal length has to be small).

Carbon_Rod

Tue Nov 27 2012, 09:25AM

Registered Member #65 Joined: Thu Feb 09 2006, 06:43AM
Location:
Posts: 1155

Newer cameras have lower resolution "color pixel group" tiles that are approximating the real image in software. A surface "pixle group" located far way has its "color" become indistinguishable from its neighboring cell.

There are some ambiguities in this method that will become self-evident.
=)
Even when "working" the limitations are clear:

Shrad

Tue Nov 27 2012, 10:13AM

Registered Member #3215 Joined: Sun Sept 19 2010, 08:42PM
Location:
Posts: 780

why not use a cheap DLP and project a raster of points with a pulsed laser, record output with a reverse biased PIN photodiode, and capture phase delay in sync with the laser pulse?

a sampling rate of at least 10 frames per second would be achievable, me thinks

Pinky's Brain

Tue Nov 27 2012, 12:16PM

Registered Member #2901 Joined: Thu Jun 03 2010, 01:25PM
Location:
Posts: 837

Phase measurement with squarewaves has problems with walk error, the point on the edges of the square waves where your comparator triggers depends on received signal strength ... there are solutions, but it's not as easy as all that. Also you need extra high frequency circuitry to do an accurate clock count until the pulse train arrives for disambiguation.

If you're going to put that much work into it you might as well try to make a single pulse TOF circuit ...

Shrad

Tue Nov 27 2012, 01:45PM

Registered Member #3215 Joined: Sun Sept 19 2010, 08:42PM
Location:
Posts: 780

signal strength should not be a problem with a dielectric filter with a sharp bandpass coating centred on the wavelength, and a schmidt trigger at the PD output, no?

after that, I have absolutely no idea about the high frequency requirements, but if I was to make such a device, I'd use a similar technique

even with slow measurements, I guess you could raster a zone easily with less than 3cm cube of components, and get signal back in approximately the same space

there are ways of modulating laser diodes at high speed "easily" and at HF there are some lasers that behave like if their quantum structure was modified, and produce extra short pulses (between 300Mhz and 500Mhz usually for what I heard)

so a pulse train on a biased laser diode would be easily produced at interleaves you could sync with the phase detector

the raster process in itself can be rather slow (for instance, a 100x100 points area of 15 milliradians in 100 milliseconds), and the output of the phase detector can be sampled by a standard microcontroller in sync with the DLP

all the high speed electronics would be laser diode driver (a standard fiber channel transceiver could be partly reverse engineered, or a driver designed) and the phase detector

thats just an idea, but in my head it seems pretty enough :)

Pinky's Brain

Tue Nov 27 2012, 04:38PM

Registered Member #2901 Joined: Thu Jun 03 2010, 01:25PM
Location:
Posts: 837

wrote ...

signal strength should not be a problem with a dielectric filter with a sharp bandpass coating centred on the wavelength, and a schmidt trigger at the PD output, no?

The problem is not getting a strong enough signal. Lets simplify a bit and say the detector+schmidt trigger (to make a square wave you can use for phase detection) triggers at light intensity X ... a strong pulse will rise to X faster than a weak one even when the real latency of the pulse as a whole is identical. There are multiple solutions (AGC and differentiation for instance) but it's not as simple as I first thought if you want useful precision.

Shrad

Tue Nov 27 2012, 08:53PM

Registered Member #3215 Joined: Sun Sept 19 2010, 08:42PM
Location:
Posts: 780

I don't know how a PIN photodiode would behave at higher frequencies, but they behave in avalanche mode... wouldn't it be sufficient to have a decent slope for each type of pulse so there would be not enough rise time difference to induce errors? (if I express myself correctly... not my first language ;) )

Patrick

Wed Nov 28 2012, 02:11AM

Registered Member #2431 Joined: Tue Oct 13 2009, 09:47PM
Location: Chico, CA. USA
Posts: 5639

sorry i fell dark there for a day or so, dropped my laptop cracked the screen, retrieved Login, now im on another machine...
ok i saw my Physics profesor today, drew pinky's idea out, i think pinky is closest to what i mean, (in fact so close i dont think he knows how right he is.)

links and pics to follow.

My pic, from discussion with Professor Tom Mausulis, Shasta College, and me Patrick Coleman, Chico and Shasta student. nov 27, 2012... (the bottom most triangle is pinkys solution)

links, as per Pinky's points... (my research from a few weeks ago),As pinky points out, onbce you make some diecisions, you become constranied by others. thats also noted by the next two people.

WiiDAR ->

using a wiimote.Lassiter (DAR) ->

highly developed one with a count screen. (videos!)

i have some optics on the way, from anchor optics, theyll get me started, but i should have got a cylinder lens too. dam.three 12.7mm dia, planar convex, no coating, 33mm FL. and a IR filter.

edit:with a 5mm focal plane, and a senor 8.3mm wide at that plane-distance, im getting 71 degrees filed of view... Does that sound right? Im using "ray tracing" through the lens center. nevermind, i botched it...

It can be done! this video proves it!

EDIT2:
in the future, for scanning the beam, ill need a very wide mirror... which is why the Neato robot turns the whole dam thing!!! shit!

EDIT3:
I need all the angles to be very sharp, acute, or this idea wont work at all.

EDIT4:
it looks like the pixels should be canted off to one side more, or else you half half the pixels seing only 15% of the triangle.

EDIT5:
i can see now what other posters' meant now... i have a design that shows 75mm to 305 mm being measured by 1500 pixels, then 300 mm to 10,000mm being measured by just 497 pixels. so each pixel, if lit singly, would yield a 20mm distance increment. feild of view is 14 degrees.

Shrad

Wed Nov 28 2012, 07:50AM

Registered Member #3215 Joined: Sun Sept 19 2010, 08:42PM
Location:
Posts: 780

for EDIT2 : what you need is a scanning polygonal mirror from a laser copier... 3mm width per 40mm length, 6 or 8 faces, spinning at kilo rpm speeds and with the possibility to sync with a hall sensor

Polygon Motor Polygon Mirror Motor For Sharp

Polygon Motor Polygon Mirror Motor For Sharp

if you manage to get an entire scanning laser module, there is a laser diode, collimating optics, polygonal mirror and cylindrical lenses for correcting the focus already, with no adjustment to be made except maybe extend the output slit from the existing A4 range (or A3 if you have the chance to get a module from a BIG printer)

getting an old laser printer for guts is not that hard, and there are much nice components inside (and the heating lamp is quite luminous too)

previous 1 2 3 4 next last

Moderator(s): Chris Russell, Noelle, Alex, Tesladownunder, Dave Marshall, Dave Billington, Bjørn, Steve Conner, Wolfram, Kizmo, Mads Barnkob