Trosien asked 7 years ago
With your OPT05-8525xxx series, the optical specs (below) for the receiver sensitivity seem to have a narrow range of operation when designing link budget, eg is it correct interpretation that we have to guarantee to be between -8dBm & -16dBm to be sure of reception without overload ? The problem of course is that attenuators are commonly available in -5dB & -10dB levels, so with the variation in transmit power it may be hard to get the receive level correct with only a 8dB window (-8 to -16dBm). For our competitors products, their receivers for similar products have typically 20dB reception range to allow easy link budget design.
Trosien answered 7 years ago
I have been testing the attenuators today. First, I’m not sure why, but I don’t get 5dB attenuation, even at 850nm as sent. I then tested them at 1310nm to compare the performance.
Test method was using both a OPT05-8525 & OPT05-31205, with the transmitter at 100% duty (on).
I start with measuring the transmitter output : eg. -1.64 dBm, as per datasheet spec:
Then, zero the meter to read relative dB:
Then, insert attenuator to read attenuation value:
The cable used is 50/125um, but also have tried 62.5/125um with similar results:
The table of results is below.
Attenuator measurements 23 November 2016
Serial Number
Atten@850nm
Atten@1310nm
201611150154
-3.4
-2.5
201611150146
-3.3
-1.1
201611150148
-3.1
-2.6
201611150151
-3.5
-2.7
201611150138
-2.3
-1
201611150153
-2.85
-0.95
201611150159
-2.4
-0.9
201611150123
-2.5
-2.5
201611150125
-3.5
-1.7
201611150143
-2.1
-1
201611150145
-2.6
-0.9
201611150130
-2.5
-1.1
201611150157
dead
dead
201611150134
-2.9
-1.1
201611150139
-2.7
-0.8
201611150156
-2.3
0.85
201611150136
-2.2
-0.7
201611150133
-2.6
-1.4
201611150131
-2.35
-1.2
201611150141
-3.3
-1.9
201611150142
-2.3
-1
201611150128
-2.75
-2.2
201611150122
-2.35
-1.1
201611150158
-2.46
-1
201611150147
-3.1
-1.2
201611150135
-3.9
-3
201611150155
-2.86
-1
201611150132
-1.9
-0.8
201611150137
-2.76
-1.5
201611150144
-1.8
-0.8
201611150149
-2.3
-1.25
201611150152
-2.3
-0.8
201611150140
-2.1
-1.3
201611150127
-2
-3
201611150129
-2.7
-1
201611150150
-2.9
-1.6
201611150126
-1.2
-0.9
201611150160
-2.6
-1.5
201611150121
-2.2
-1
201611150124
-2.5
-0.8
All parts seem to have much less than 5dB attenuation; interestingly at 1310nm most results scale from the 850nm figure, but a few individual ones have similar performance at 1310nm.
Can you ask an engineer please if there is any reason the attenuators should be so low? Even with the results above, it would seem that the 850nm parts are not suitable at 1310nm.