Greetings, Just musing here. How might I couple a fiber optic cable to the optical window on a Lidar chip? Wondering if it's possible to do Lidar through a twisty path, like routed around an arbitrary light blocking "thing". Are there going to be intractable reflections at one or both ends of the fiber? I do have a use for this if it is possible... Regards, Bill M.
OT? fiber optic question
Dec 06, 2025
Last reply: 7 months ago
16 Replies
If the lidar chip has a window, and the beam is expanded to some modest diameter, you'd need a lens to get any amount of useful coupling.
Got a link to the part?
There are lots of cheap OTDRs around, so there must be fiber-compatible parts.
John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics
There are. What are used for coupling are ball lenses or GRIN lenses.
Aligning these things by hand is possible, but very fiddly, so most folk buy a pigtailed laser as the optical pulse source. The LIDAR chip drives the source and listens to the receiver (also pigtailed). And so on.
Joe Gwinn
The Lidar chip is ST VL53L4CD, a low power gadget for 1 to 1200mm ranging. Nothing elegant, but it's cheap. I only need about 1.5" worth of sensing range, just monitoring the position of an air cylinder shaft. Yes, there are cylinders with built-in position read out, but I don't get to use those...I'm looking at an add-on to existing installation(s). I'm suspecting that the Lidar to fiber connection is not the real problem here, the moving target reflector may be the bigger issue, I can't introduce a critical mechanical alignment. It's going to get shaken vigorously...in a race car. Looking into the optical fiber as a way out of the also nasty electrical noise environment. Alternative is to level shift the Lidar chip signals to 12v and mount the chip on the air cylinder, not too keen on running an i2c bus over 3 or 4 feet of cable at low level signal. Anyway, just kicking around possiblities for now. Have a couple of the Lidar chips to play with, that may drive the solution one direction or another...or nowhere. Thanks for the reply!
-bill
Cool part.
You might be able to use the clunky large-diameter plastic fiber, the ones used in audio links, and get usable coupling. Toslink. The core diameter is about 1 mm.
You could glue the fibers right onto the tx and rx holes in the top of the chip.
You can also buy glass fibers with a moulded-in teardrop lens on both ends, the GRIN that somebody mentioned. That might work on both ends.
formatting link
Ignore the Thorlabs pricing. They are insane.
The other end, at the cylinder, would be interesting. Could the fiber ends face one another, with no reflector? That would give tons of signal with no alignment issues.
John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics
< That was my first thought, for a cheap experiment with easy to get parts. > < Thinking to make a little alignment block that sits on top of the chip, so I can play with the air gap distance to the fiber. Maybe overkill? I can do the machining, so it will only cost a bit of time... > < Insane? Close, I was thinking "Obscene"! :-) The individual parts look like the way to go, but no way can this little fun project afford them. > < That would be cool, but mechanically intractable, as the cylinder rod may rotate a little bit as it travels in/out. I can't guarantee it will stay aimed. >
I have seen a variant that I thought was a very creative use of this Lidar technology in kitchen design. A friend in the glass cutting business (think laser cut bespoke kitchen splashbacks) has a black box gizmo with a reeled fibre optic sensor probe. You put the probe onto the wall and get back an x,y,z position relative to a central datum.
Take it home run some clever software and it generates the exact CNC cutting profile for their laser cutter and just press go.
Compared to hand measurement it is incredibly quick and absolutely spot on to mm accuracy so there is no fiddling about to fit it later. It also spots walls that are bowed in some way or not truly flat and works out the most reasonable compromise fixup solution to adjust it to be so.
He's very pleased with it. Complex installations with lots of cutouts that used to take all day and with a slight risk of damaging the piece of glass doing manual final adjustments now take a couple of hours.
A retro-reflector might be useful. It would self-align.
There is paint and tape with retro-reflective glass beads. Amazon has some.
So: toslink plastic fibers and retroreflective tape. Cheap and rugged.
John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics
< That's a good starting point for sure. My mental image for the reflector is a stiff disk with a hole in the center for the air cylinder shaft. The paint option looks like the best bet. How about just temporarily connecting the two fibers end-to-end to get a calibration delay for the cable? That chip has some interesting possibilities.
The best ScotchLite and Reflexite material is good for 70 dB (electrical) signal increase compared with a Lambertian white surface.
Cheers
Phil Hobbs
70 dB sounds pretty good. The plastic fiber is sorta lossy and coupling in and out of the lidar chip won't be spectacular either.
John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics
I was imagining something like that, unless there is accidentally some moving thing handy.
If the fibers are close together and parallel, basically the transmit and receive antennas, why not measure the distance from there to the reflector and use that as the cal point?
It would take a Hobbsian level of calculation to see what the optical power budget might be. Or a couple of experiments.
A lens in front of the receive fiber might help. A 1 mm diameter fiber is a pretty small target.
Why lidar? Fun but maybe overkill.
Try Amazon for linear position sensor
John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics
< Something like this? "Red Reflective Tape - 3M 3930 HIP Air Backed MicroPrismatic Film type4"
It doesn't mention IR specifically; I see others that claim red & IR.
980nm is the wavelength in use here. >
< True, it is small, but not much different from the window in front of the chips receiver. Would be nice to make up for some of the loss from the cheap fiber though. Yet another mechanical challenge! > < mostly because that chip lets me program it to interrupt when the target is "in range", as defined by the host micro. Then I don't have any concerns about asking for a readout over i2c frequently enough to not miss the start or end of motion. Also it is small & inexpensive compared to everything else I have looked at. It kind of matters that the reflector/target can be very low mass as well. >
Scotchlite 3000x (corner cubes) and some Reflexite SKU that I don’t have handy (silvered glass beads) are the best. Even the crummy stuff will still get you 40 dB or thereabouts.
Either should work fine at 980–the dyes in the colored stuff transmit NIR fine.
For a one-off, try whatever you can get easily, and see if it’s good enough.
Cheers
Phil Hobbs
< Well, I found a couple of sources for modest quantity of the 3000x tape, one in Washington state, the other in Canada. Amusingly (or not) the one in Canada was about 20% less expensive, but will not take orders to ship to a USA address! Imagine that.
Try Amazon.
I just ordered some rectangles of 3M stuff to put on the insides of my car doors. The streets are very narrow here and an open door isn't very visible to cars or bikes.
Phil, would a lens, maybe a plastic frenel, help the tx/rx gain? I can imagine a 3D printed tube thing with two fibers glued into one end and a bit of lens on the other.
Hey, could one 3D print lenses?
John Larkin Highland Tech Glen Canyon Design Center Lunatic Fringe Electronics
Join the Discussion
Have something to add? Share your thoughts — no account required.
Didn't find your answer?
Ask the community — no account required