• State
  • Locked Locked
  • Replies 10 replies
  • Subscribers 34 subscribers
  • Views 1831 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

DS92LV18 driving SFP for FO communication

savino giusto
Prodigy 180 points

Other Parts Discussed in Thread: DS92LV18, DS92LV2412, DS92LV2411

Hello,

I'm trying to interface a DS92LV18 serdes with a common ethernet SFP-tranceiver (http://www.apacoe.com.tw/datasheet/60/LS38-C3L-Tx-N-V1.1.pdf). In my project i need to extend some commands and contacts over fiber in order to cover longer distances and to reduce the number of cable wires.

By now i managed to get the serdes to work on a copper pair. I can transmit and receive data correctly accross two prototype boards i made.

The problem is that i haven't been able to drive correctly the sfp. 

According to the datasheet the SFP drives/receives LVPECL levels, so i arranged a passive AC-coupling circuit as described in your application report (http://www.ti.com/lit/an/scaa059c/scaa059c.pdf).

Is there something i'm missing?

Do you have some alternative optical tranceiver that i could use along with DS92LV18?

  • 0
    TI__Expert 3175 points

    Hi

    My apology to get back to you late.

    AC coupling is correct. Can you send us a drawing that show how you connect DS92LV18 to the SFP?

    We will review and provide feedback.

    Regards,

    TK Chin

  • 0 in reply to T.K. Chin
    Prodigy 180 points

    2022.schematico.pdf

    Hi,

    Thank you for your reply. Attached you can see the schematic of my prototype board. I made two identical board and i'm trying to let them communicate.

    Best regards

  • 0 in reply to savino giusto
    TI__Expert 3175 points

    I take a quick look at the schematic. A few notes:

    1. Please double check TD+- is the "input" to the optical module driven by the LV18, and RD+- is the "output" of the optical module driving LV18.

    2. Suggest to change 100pF to 0.1uF for ac coupling capacitors

    3. Recommend to use a differential probe to check the signals at this nodes: DO+-, TD+-, RD+-, RD+-.

    regards,

    TK Chin

     

  • 0 in reply to T.K. Chin
    TI__Mastermind 43795 points

    It sounds like you are able to use a copper cable to connect between SFP+ cages.  Please verify. 

    With a balance data payload, like PRBS7 on each parallel input, the LV18 can communicate over optical modules.

    The optical modules will require a balanced - encoded pattern to work well. Otherwise a DC offset will be present, reducing any margin available in the system. If the DC content is strong enough, the system will not work at all.

    Regards,

    Lee

     

  • 0 in reply to T.K. Chin
    Prodigy 180 points

    1. According to the SFP-MSA seems that TD+- are the inputs which drive the laser diode and RD+- are the receivers. 2474.INF-8074-28.pdf

    2. Actually i tried a very large range of AC coupling capacitor. I spanned from 100nF to no capacitor at all but according to the frequency of the signal 100p seemed a more appropriate value.

    3. I do have a differential probe. Do you think a 60MHz oscilloscope would be fast enough for that signals or will i need a faster one?

  • 0 in reply to Lee Sledjeski
    Prodigy 180 points

    Lee Sledjeski said:

    It sounds like you are able to use a copper cable to connect between SFP+ cages.  Please verify. 

    With a balance data payload, like PRBS7 on each parallel input, the LV18 can communicate over optical modules.

    The optical modules will require a balanced - encoded pattern to work well. Otherwise a DC offset will be present, reducing any margin available in the system. If the DC content is strong enough, the system will not work at all.

    Regards,

    Lee

     

    Thank you your response.

    I'm not sure i understood what is the verification you're suggesting. Actually SFP cages are already electrically connected because they're tied to GND and the two boards are supplied by the same PSU.

    As for the consideration about payload this was something i was afraid of. In my application i need to control an monitor a Laser from a small (15-25 meters) distance. The signals i need to transmit and receive are in the form of short during (few milliseconds) steps. Will i be able to do this using standard SFPs?

  • 0 in reply to savino giusto
    TI__Mastermind 43795 points

    If it is not possible to guarantee DC balance directly, I would suggest using a device which can provide the encoding necessary to AC couple and work with optical modules.

    http://www.ti.com/product/ds92lv2411

    Regards,

    Lee

  • 0 in reply to Lee Sledjeski
    Prodigy 180 points

    Lee Sledjeski said:

    If it is not possible to guarantee DC balance directly, I would suggest using a device which can provide the encoding necessary to AC couple and work with optical modules.

    http://www.ti.com/product/ds92lv2411

    Regards,

    Lee

    That would require two separate boards, one for trasmitting and one for receiving. Am i right?

  • 0 in reply to savino giusto
    TI__Mastermind 43795 points

    Savino,

    Yes this would require two seperate devices, the DS92LV2411 and DS92LV2412 are needed to complete the serialization and de-serialization functions.

    Regards,

    Lee

  • 0 in reply to Lee Sledjeski
    Prodigy 180 points

    Thank you Lee for your answers.

    It looks like you have some experience with optical modules. Considering that the kind of data i'm going to trasmit. Will I have the same problems using a different kind of optical transceiver? (e.g. 0243.AV02-0176EN_DS_HFBR-x4xxZ_2013-12-09.pdf)


    Regards