Other TE problem here. A couple years ago, I bought a W & G optical test set used from the UK (back when the UK was a country you happily bought things from instead of laughing at its narrow-majority-inflicted problems), around 100£ for source and meter. I've been worried about its accuracy for some time, because I've gotten strange readings from measuring through lines using the source into the meter, as one does. The symptoms were seriously low power levels recorded.
Yesterday evening I had my plan formulated, and went at it. I plugged a 10G LR transceiver with DDM (Digital Diagnostic Monitoring) into the free SFP slot on my living room switch, and took a look at the values:
STROMSCHRANK#sh interfaces te 1/0/1 transceiver detail
...
Optical High Alarm High Warn Low Warn Low Alarm
Transmit Power Threshold Threshold Threshold Threshold
Port (dBm) (dBm) (dBm) (dBm) (dBm)
--------- ----------------- ---------- --------- --------- ---------
Te1/0/1 -1.5 0.0 -1.0 -6.0 -7.0
Optical High Alarm High Warn Low Warn Low Alarm
Receive Power Threshold Threshold Threshold Threshold
Port (dBm) (dBm) (dBm) (dBm) (dBm)
------- ----------------- ---------- --------- --------- ---------
Te1/0/1 -40.0 0.0 -1.0 -16.0 -17.0
The important part we're seeing here is that it claims to push -1,5dBm out. I connected the meter up, after having done the entire IEC-mandated inspection and cleaning procedure, and got a reading of -1,62 dBm. That is entirely acceptable; losing 0,12dB over two splices is nothing to cry about, au contraire! It is quite reasonable.
OK, so we know it's gotta be the the source. Source then got inspected and cleaned according to the manual, which means lifting the bulkhead FC connector from the enclosure, disconnecting the internal cable connector, and inspecting and cleaning the ferrule. Am not entirely happy with the end face on the 1310nm output, there are contaminations that I can't clean off. The 1550nm output has a pristine ferrule anyway, so that's not the major problem, because I'm losing 27dB compared to what the source's supposed to put out. On both wavelengths.
I need to get hold of documents, and dare opening the unit. Cal is 4 years overdue anyway.