CWDM crosstalk model

[Thomas FreirParticipant]

Hi, I am investigating crosstalk in CWDM systems. Initially I am building a basic model so I can understand how the different components work. I have built a 2 channel CWDM model with the laser from one of the channels fixed at 1490nm and the second laser is nominally at 1470nm but it is being swept from 1420nm to 1540nm. I have had a result I cannot explain. I have carried out the simulation in EXCEL and they do not agree.

I have attached the optisystem simulation and the result I am querying.

In the picture “layout”. The top left laser is being varied by wavelength. I am interested in the effect it has on the other laser (bottom left) and its receiver optical power meter_2.

In the picture “result” the large curve is as I expected. It is the spike to the right that has me puzzled. Any help would be most appreciated.

Attachments:

1. 19-05-2015-New-CW-laser-CWDM-system-using-Gaussian-filter-1.osd

[Thomas FreirParticipant]

I had to attach the pictures separately.

Attachments:

1. 

   Layout.png
   

2. 

   Result.png
   

[Alessandro FestaParticipant]

Hi Thomas, what is the sense of sweeping the wavelength of CW Laser from 1420nm to 1540nm, if the WDM Mux 2×1 has the wavelengths fixed? I think you should have the first input of that WDM following the Laser emitted wl.

Alessandro

[Thomas FreirParticipant]

Don’t mind the range I have swept over. This is just so I can see how the filters perform. In practice the lasers will only be drifting by a few nanometers.
Remember in CWDM systems the lasers are allowed drift from their central wavelength by plus of minus 6.5nm. I want to explore how this wavelength drift effects adjacent channels.

[Alessandro FestaParticipant]

But how can you have results which make sense at Optical Power Meter 2, if the launch WDM has its wavelength which is not aligned with the CW laser?

[Thomas FreirParticipant]

The WDM mux and demux are band pass filters with a Gaussian response. As the CW laser has its wavelength changed some of the CW laser will leak through. The MUX and demux are not ideal.

[Alessandro FestaParticipant]

I am sorry Thomas but I do not see the point of your simulation…what result is puzzling you? Thanks for your patience!

[Thomas FreirParticipant]

Under ideal circumstances Laser 1470 would arrive at power meter 1 with a power of -10dBm. Laser 1490 would arrive at power meter 2 with a power of -10dBm.

The picture “Result” shows the power measured at power meter 2. You will see that the lowest power on this graph is -10dBm. This is the expected power from the 1490 laser. In addition to this there is a gaussian shaped bulge. This is an unwanted leakage of power from the 1470nm laser as its wavelength is swept over a range. I would expect this result and have confirmed it elsewhere.

The unexpected result is the spike of optical power to the right of the gaussian shaped bulge.

The reason I am building this model is that CWDM lasers are expected to drift in wavelength domain. I know the drift in this case is excessive but I am trying to validate the model.

Thanks for your time.

[Alessandro FestaParticipant]

Thanks for the clear extensive explanation.
The spike you mention happens at iteration 71, where both lasers are at 1490nm: since you are looking with PowerMeter2 at 1490nm, I would expect that this is the point where you receive the highest power!

[Md Hayder AliParticipant]

Hello,
I am from Bangladesh. I am working with Crosstalk effect of CWDM technology. I am using Optisystem 7.0. Unfortunatly it was for trial version. I need licence key for using it. Can you pls help me regarding license key? I need more help to compare crosstalk effect of DWDM and CWDM. I will communicate with you later.

Thanks
Hayder