Measuring an optical signal with nanosecond resolution

[malakh]

Hello,
I'm looking for a way to measure a difference between 2 light signals from 2 fibers. Does anyone have any idea?
Actually what i want it's DAQ card with an accurate resolution...
Thanxs

 

[Sceadwian]

The difference in what? The arrival of a pulse, the wavelength, phase? You need to describe what you're trying to do a little better.

 

[malakh]

thx Sceadwian,
my light will be converted in volts and measure the voltage. when a loss of light (so decreasing of voltage) occurs i want to measure the time between the receiver will detect this loss on the 1st and on the 2nd. theorically this time is in nanosec

 

[Nigel Goodwin]

I suspect the conversion to voltage is going to take a great many times longer than the period you're trying to measure.

 

[Sceadwian]

Yeah, it really depends on your light sensors responce, most common light sensors have response times of a couple miliseconds or micro seconds.

 

[ericgibbs]

Yeah, it really depends on your light sensors responce, most common light sensors have response times of a couple miliseconds or micro seconds.

hi,
Some IR detectors have nanosec response times, commonly used on laser ranging devices.

Even at those high switching speeds it very difficult to do 'direct' real time comparison of two light pulses using a counter.

EDIT: Example types.

 

[jpanhalt]

Neither datasheet (SFH205 and SFH203a) gives data on the reproducibility of the switching time. Would it really matter what the switching time was (within the limits of pulse frequency), if it were very reproducible (i.e., the s.d for switching time was << than the difference in arrival times)? Is that data available?

As I read the OP's question, he seems to mean voltage as a logic/digital signal, not as something that is quantitated.

John

 

[ericgibbs]

Neither datasheet (SFH205 and SFH203a) gives data on the reproducibility of the switching time. Would it really matter what the switching time was (within the limits of pulse frequency), if it were very reproducible (i.e., the s.d for switching time was << than the difference in arrival times)? Is that data available?

As I read the OP's question, he seems to mean voltage as a logic/digital signal, not as something that is quantitated.

John

hi John,
I used the SFH203FA as detector, using a LD163 904nm laser diode in my commercial range finders.

The laser units had to be calibrated to to correct for the variance in 'sum total' response time of the laser diode, SFH and following amplifiers.
Once calibrated the laser unit were very stable.

The laser measured 'range' had to be corrected also for the 'strength' of the returned echo.
The stronger the echo, the faster the rise time and the 'apparent' shorter range.

 

[Sceadwian]

All things being equal a fast response time is going to equate to high reproducibility.

 

[jpanhalt]

hi John,
I used the SFH203FA as detector, using a LD163 904nm laser diode in my commercial range finders.

The laser units had to be calibrated to to correct for the variance in 'sum total' response time of the laser diode, SFH and following amplifiers.
Once calibrated the laser unit were very stable.

The laser measured 'range' had to be corrected also for the 'strength' of the returned echo.
The stronger the echo, the faster the rise time and the 'apparent' shorter range.

Great. Now I have a winter project.

Thanks for the information. I suspected intensity, temperature, and ambient light(?) might be factors. It is easy to see how echo intensity might be corrected for. Is temperature a factor too?

John

 

[ericgibbs]

Great. Now I have a winter project.

Thanks for the information. I suspected intensity, temperature, and ambient light(?) might be factors. It is easy to see how echo intensity might be corrected for. Is temperature a factor too?

John

John,
The method used is:
At TX time [about 250pps] the LD163 30Watt is lased, using a transistor in avalanche mode, for a pulse period about 25nSec, a TX F/F is set.
The TX F/F setting enables a high linearity ramp down voltage, as the echo is detected the TX F/F is reset and the ramp voltage is run upto a zero detect comparator.
The ramp run up time is 50 or 100 times longer than the ramp down rate.

During the ramp up a high speed xtal clock is gated into a counter and stopped at zero crossing of the ramp.
By choosing the xtal freq as multiple of 15Mhz. [ assuming the laser beam travels at 300,000,000 mtr/s , round trip of the pulse is double the range]
you can get a direct count of the range.

If you capture the peak value of the returned echo and digitise it using a8 bit high speed adc you can use this 8bit value as an address pointer to the signal strength table and adjust the correct measured range.

The cal table resides in non volatile memory and is created during calibration in production.

The range spec for the laser is 5000mtrs from cooperative targets with a resolution of 0.1mtr.

If you want more info, just ask.

 

[jpanhalt]

Thanks again. Useful information for me and the OP, I hope. If I have more questions, I will start a new thread so as not to hijack this one.

John

 

[malakh]

Thank you guys. I think to another way to get this time I want measure.
John I am interested by your questions. I'm following your thread