FDDI fiber testing by ATMega microcontroller

Thread Starter

trixheim

Joined Apr 1, 2009
10
Hi

I'm thinking of a new electronic project to make a simple test unit for FDDI optical fiber (have a lot of fibers and need to find out who are good).
The plan is to use HFBR2115T/HFBR1115T tranceivers put into sockets because I also want to test different tranceivers. The interface to the tranceivers are PECL, so i will translate the levels to TTL by MC100ELT20 and MC100ELT23, and finally connect a ATMega88 running at 20MHz (because I have used it before).

The consept is to make a loop where I will send data out on the microcontrollers serial output. The data should go through the fiber and returning back to the serial input, and if there is no errors the fiber (and tranceivers) are supposed to be good.

Is it good enough to just use async serial communication?
Maximum data speed from the controller is 2,5 Mpbs @ 20MHz, but the FDDI is rated at 100Mbps. If the communication is ok at 2,5Mbps, will that be a good enough indication of a good fiber, or is it possible that it will fail at 100Mbps?

Regards
Tor
 

beenthere

Joined Apr 20, 2004
15,819
Depends on you definition of "good". If it's enough, just illuminating a fiber with a source and checking that the output is within limits might be enough. That's assuming that a nick in the cladding that would attenuate the signal will also affect total light through the fiber.
 

Thread Starter

trixheim

Joined Apr 1, 2009
10
Yes, maybe I'm thinking too much about how electrical cables are working. As long there is enough light transfered, the bandwith should not be directly affected. My plan is to transfer data continuously and use LED's to indicate data received (or maybe a small LCD-display to show numbers of bit errors).

thanks
Tor
 
Top