Re: 1 ns red laser diode pulses by 6 ns electrical pulses
Sulaiman, Sun Dec 04 2016, 11:59AM

the first half of the 6ns pulse is a classic (1-1/t) shape so
. something needs 'charging up' before it operates
. the source (sig-gen) and load (diode) impedances are not matched

e.g. 50 Ohm system, t=1ns for C = 20 pF etc.

Re: 1 ns red laser diode pulses by 6 ns electrical pulses
klugesmith, Sun Dec 04 2016, 03:07PM

Is this a puzzle question, where you already know the answer?
Let's assume not, and that you are looking for help to understand the behavior.

What happens to optical pulse W and H when you change the electrical PW?

As Sulaiman said, there's an impedance mismatch. 75 ohm cable with 50 ohm scope input and/or pulse generator output? We need more detail about the electrical setup, including cable lengths and other things you might think don't matter. What's the electrical pulse amplitude, in volts or amperes, taking oscilloscope and probe (?) scale factors into account?

What kind of LD do you have, and what wavelength? The ones used for optical fiber communication, and in my experience optical disk reading, are easily switched on and off (at least 20 dB power ratio) at GHz rates.

Re: 1 ns red laser diode pulses by 6 ns electrical pulses
Physikfan, Sun Dec 04 2016, 05:54PM

The pulse generator is a HP8082A with 50 Ohm output impedance, maximum amplitude is 5V from 50Ω into 50Ω.
The voltage amplitude was about 5V plus the full postive bias from the HP8082A.
The oscilloscope was a Tektronix DSA 602 Digitizing Signal Analyzer (1 GHz) with 50 Ohm input impedance.
I was using a 50 Ohm coaxial cable but I agree, you can see a small impedance mismatch.
I will improve the impedance matching by replacing the resistor of about 38 Ohm used now by a better value.

But the interesting fact is, that a small electrical pulse width (lower than 3 ns but the same amplitude) generates almost no laser diode (visible red) intensity.
Further information (laser diode specs, cable length, etc) will follow.

In the near future I will show the laser pulse width as well as the intensity as a function of the electrical pulse width.

Re: 1 ns red laser diode pulses by 6 ns electrical pulses
johnf, Sun Dec 04 2016, 06:24PM

That is because you are dividing the pulse energy into the junction capacitance of the laser diode and not getting above the lasing threshold

Re: 1 ns red laser diode pulses by 6 ns electrical pulses
Physikfan, Sun Dec 04 2016, 06:49PM

Hi

What are your opinions about my working hypothesis:
time delay caused by a slow increase of the light amplitude (laser oscillation of the laser diode)?
This time delay should depend on the Q of the diode laser cavity.

Ad Klugesmith:

Please could you explain:
"What happens to optical pulse W and H when you change the electrical PW?"

Ad johnf:

Please, could you explain more details of:

"That is because you are dividing the pulse energy into the junction capacitance of the laser diode"

"not getting above the lasing threshold" below 6 ns electrical pulse width?

Re: 1 ns red laser diode pulses by 6 ns electrical pulses
klugesmith, Sun Dec 04 2016, 07:38PM

Regarding the first question, the delay you report is millions of optical cycles. Hard to imagine any real physical oscillator being that "slow" to ring up.
Regarding #2, sorry about that. I meant to abbreviate optical pulse Width and Height, and electrical Pulse Width. For the first two I picked letters from FWHM but got that wrong.

Re: 1 ns red laser diode pulses by 6 ns electrical pulses
johnf, Mon Dec 05 2016, 06:34AM

Alright a quick reply
your pulse if the area under it is intergrated that is the loule amount
now put that amount into an unknown capacitor q=CV so now you can estimate the capacitance by reducing the area under the pulse so that V or voltage does not get above the lasing threshold of your device.
dont forget the capacitance includes the cables pcb and the laser diode itself.
work backwards you do not need too much capacitance upset the diode output
QED

Re: 1 ns red laser diode pulses by 6 ns electrical pulses
Physikfan, Mon Dec 05 2016, 06:20PM

Hi Sulaiman

Please, what do you mean with:
"a classic (1-1/t) shape"?

the shape of the voltage of a capacitor C as a function of the time charged via a resistor R:

Uc(t) = Uo (1 - exp(-t/RC)) ?

Re: 1 ns red laser diode pulses by 6 ns electrical pulses
Sulaiman, Mon Dec 05 2016, 09:17PM

yes, that is it, I was sloppy :(

Re: 1 ns red laser diode pulses by 6 ns electrical pulses
hboy007, Sat Dec 10 2016, 05:39PM




I see nothing wrong with that lasing spike, depending on the structure of the LD, this might even be desired. Extending the electric pulse should lead to more spiking, followed by CW operation.
As for the timing - At 5ns/m propagation delay every cable segment matters. Have you re-checked all cable lengths involved?
Impedance also matters. I more than once hat to resort to terminating the signal cable near the LD and adding in a buffer driver with only millimeters to the LD. Just to avoid any reflections summing at the LD...

Re: 1 ns red laser diode pulses by 6 ns electrical pulses
Physikfan, Sun Dec 11 2016, 11:01PM

Hi hboy007

Very interesting figure!

"adding in a buffer driver with only millimeters to the LD."

Please, could you give me the source or the circuit for this buffer driver.
Fast enough for almost subnanosecond pulses?

Regards

Physikfan