Why did you write "they are using"? Thre was no team, it's my own project.
> What kind of details do you want to know?
What is the limiting factor on the precision - is it just the specs of the TDC-GP21 or something else?
How did you calculate the appropriate laser lens and standoff?
Is the CS mount lens's focal length important, or is there no need for a sharp image at all distances?
Why use one CS-mount lens and one M12 lens, rather than two M12 lenses?
Do you need any sort of dynamic gain or optical filtering on the received signal? Is it vulnerable to getting washed out by bright sunlight?
Did you need much fancy $$$$ test equipment to get the design working, or did you achieve it with tools mere mortals can afford?
Precision in this project is limited by rise time of the received signal. Meanwhile, this time is limited by rise time of laser's current. It is too long now - 10 ns.
Theoretical TDC resolution is ~13mm per BIN.
You could see real results here: https://github.com/iliasam/OpenTOFLidar/wiki/Resolution-of-t...
> "How did you calculate the appropriate laser lens and standoff"
I can't say that there was too much calculations.
The is some information here: http://www.ti.com/lit/ug/tiduc73b/tiduc73b.pdf (part 2.3.5)
and here: http://www.ti.com/lit/ug/tiducm1b/tiducm1b.pdf (part 2.3.2)
Laser's lens should have maximum possible focal length, but it must not crop light beam from the diode.
Photodiode's lens should have maximum Entrance pupil
> "Is the CS mount lens's focal length important, or is there no need for a sharp image at all distances?"
You need to have sharp image at big distances. At short distances it is not important - light signal is high enough.
> "Why use one CS-mount lens and one M12 lens, rather than two M12 lenses?"
I think that it is easier to find lens with big "Entrance pupil" with CS mount that with M12 mount. Bigger mount - bigger lens diameter, bigger "Entrance pupil".
> "Do you need any sort of dynamic gain"
I don't have any kind of electrical gain control. There is an ability to change APD gain, but I don't use it.
> "optical filtering on the received signal"
I don't have filter in my LIDAR, but it is necessary to have it if you are going to run LIDAR at sunlight. I give links to a several interference filters in "LidarTotalBOM.xlsx".
> "Did you need much fancy $$$$ test equipment to get the design working".
All that I have is Tektronix TDS540D oscilloscope (bought at the Ebay for $300) and multimeter.
"This use of they isn't ungrammatical, it isn't a mistake, it's a feature of ordinary English syntax that for some reason attracts the ire of particularly puristic pusillanimous pontificators, and we don't buy what they're selling."
There's not a man I meet but doth salute me
As if I were their well-acquainted friend
Peak power vs. SNR favours pseudorandom sequences with a peaky autocorrelation function. In practice, an LFSR is typical, but if you use e.g. AES-CTR, you can get quite a lot of resilience against (intentional) jamming and be potentially undetectable as long as the sensor doesn't move directly into the beam.
There is no need for an FPGA, but you will rather quickly throw FFT correlation processing at it.
This would not use any delay lines. You'd just need to lock the sampling of your ADC to the signal generation, because 2ns jitter are 30cm/1ft jitter in distance.
For the source they are using an osram SPL PL90_3 laser diode which provides 75W peak output at 900nm, with pulse duration roughly 10ns and costs $10.
For the detector they are using a MTAPD-07-013 avalanche photodiode, which has a quantum efficiency of about 80% and an internal avalanche gain of 100 and 0.6ns rise time. Also costs about $10.
For the time measurement they are using a TDC-GP21 time to digital converter with 22ps resolution, although the author has it configured for 90ps counts. Costs a bit over $5.
It is really a testament to the amount of development work that has gone into the lidar and related field, that for under $100 you can build a full lidar system out of parts from digikey including custom pcb's and scan mirror. When I was in grad school we paid $7500 for our avalanche photodiode, $25k for our pulsed laser, and used a $50k oscilloscope to read it out.