Laser Noises and other issues; (to choose optimal DC current)

[LandLack]

Hi guys,
I would like to insert inside a Direct Modulated Laser a multitone RF signal;
the laser is this: http://www.suncityinc.com.cn/IMAGE/pro/PDF/DC-18GHz_DFB_LD_FIN.pdf .
Bias current has a different input than the AC signal current, so I would like to inquire the best DC current to give. The only thing I understand is that the DC current should have a high enuogh value to avoid "Clipping", but I don't know the other things to take into account.

1. Can you tell me the main issues of a laser, from this point of view?​

For what I've been told (in general) the main issues are "Carrier power/Noise" and "IMN" (maybe Intermodulation?);

2.Can you tell me what are the sources of noises in my situation?​

3. Also, is this Intermodulation referred to the combination of the (frequency) tones of the input, or the various wavelengths? If the second one is true, shouldn't I be safe from this point of view (since I'll use only a wavelength to transport the multitone signal)??​

Sorry for being long, but I've got a general theoretical knowledge and no experience on the subject.

Best Regards,
LandLack

 

[FvM]

You'll want to use a laser driver with feedback for the DC bias to keep the average laser power constant. Modulation current will be directy superimposed to the laser current.

 

[LandLack]

You'll want to use a laser driver with feedback for the DC bias to keep the average laser power constant. Modulation current will be directy superimposed to the laser current.

I have managed to prepare a good DC bias source (still not tested), but I would like to know what level should I use, and what are the implications; for example, I've seen that the Ith (threshold) is 20mA, and the maximum suggested Idc (DC current) is 110mA.
The only thing I've understood is that since the signal's maximum input power is 13dBm, which means 20mA (Zin= 50ohm), so to avoid Clipping issues, I need to give at least 40mA to DC current.
I would like to know something about the other issues, since I would like to have the best estimation infos about C/N + IMN (Carrier power to noise ratio and the intermodulation).

 

[drkirkby]

I've modulated lasers with RF by setting the DC current half-way between the threashold and maximum. In your case the difference betweeen threshold and maximum current is 110-20=90 mA, so set the current at 20 + 90/2 = 20 + 45 = 65 mA. Then you can get close to 100% modulation depth using a peak current of 45 mA, which is an RMS current of 32 mA.

I'm a bit confused by some of what you wrote, but if you bias it at 40 mA as I think you are implying, and have 20 mA RMS current for modulation, then the peak modulation current is going to be 28 mA, so the current would attempt to drop to 40-28=12 mA and so be below threshold.

Be careful if running lasers at their maximum current. They make extreamly fast fuses. You need to ensure there are no transients on your power supply.

 

[LandLack]

I've modulated lasers with RF by setting the DC current half-way between the threashold and maximum. In your case the difference betweeen threshold and maximum current is 110-20=90 mA, so set the current at 20 + 90/2 = 20 + 45 = 65 mA. Then you can get close to 100% modulation depth using a peak current of 45 mA, which is an RMS current of 32 mA.

I'm a bit confused by some of what you wrote, but if you bias it at 40 mA as I think you are implying, and have 20 mA RMS current for modulation, then the peak modulation current is going to be 28 mA, so the current would attempt to drop to 40-28=12 mA and so be below threshold.

Be careful if running lasers at their maximum current. They make extreamly fast fuses. You need to ensure there are no transients on your power supply.

Sorry, I've descibed it badly, and wrongly; let me try again: the electrical specification of the the laser said that the 1dB compression point was 13dBm. I've created an electronic system with a peak to peak value of 3V, with 50 ohm (around exactly the limit of 13dBm)... So, as you said, the current changes are -30mA to + 30mA, so to avoid clipping I need to give at least 50mA of DC current.
If this was only a 1bit signal, to maximize the Extinction Ratio it would have been better to have Idc as low as possible (so 50mA, as said before), but since I'm using many different tones together (each representing a bit, actually), how should I behave?
Also, I've found the formula for RIN noise and (probably) how to find IMN (through IIP3, right?); can you tell me the other possible noises?

Best Regards,
LandLack