The nice thing about this method is that it's all passive components.

I might still rather do the switch at top with PoE in+out and VLAN per SM.  Without actually adding everything up I'd wager that the cost was about the same....or comparable at least.  The operational difference is in whether you have more faith in the brains out in the field or more faith in the brains in the office configuring switches.  And if the building had more than 2 units it might matter that you can add a 3rd or 4th SM this way.

Or the ultimate lesson to take away from this:  If you're running a difficult and time consuming cable path then pull extra cables at the same time.  It's hard to get employees to look past the current job and think about how they can help themselves in the future, but it's nice if they can.


On 1/29/2020 12:52 AM, Forrest Christian (List Account) wrote:
The differing twist lengths are engineered to minimize crosstalk between the pairs.   If all of them have the same number of twists per inch then you will find that the same wires tend to be next to each other down the length of the cable.   If instead you have each of them have a different number of twists such that over the length of the cable the amount of time each is in contact with each other tends to be more even, reducing crosstalk.

I suspect in some cases having two separate links running through the same cable will hurt performance because you will get crosstalk from the other link which you may not be able to cancel out using an echo canceller.  Probably depends on the length and specifics of the link.

On Tue, Jan 28, 2020 at 8:33 PM Ken Hohhof <af...@kwisp.com <mailto:af...@kwisp.com>> wrote:

    Each pair has a different number of twists per inch.  In Cat5 and
    Cat5e cable I observe the green and orange pairs, which are the
    data pairs, have the tightest twists.  I don’t remember if Cat6 is
    similar.  This leads me to believe the blue and brown pairs may
    have inferior crosstalk performance.  But GigE uses all 4 pairs
    for data, so my theory is probably wrong.  I guess the important
    thing is that none of the pairs have the same number of twists.

    The reason I mention this is sometimes I see people assert that if
    you split a Cat5 cable into 2 as is being discussed, it will hurt
    the performance.

    *From:* AF <af-boun...@af.afmug.com
    <mailto:af-boun...@af.afmug.com>> *On Behalf Of *Forrest Christian
    (List Account)
    *Sent:* Tuesday, January 28, 2020 5:24 PM
    *To:* AnimalFarm Microwave Users Group <af@af.afmug.com
    <mailto:af@af.afmug.com>>
    *Subject:* Re: [AFMUG] Cambium POE Splitter

    This is actually pretty simple:

    Split the CAT5 into two, two pairs per radio, put the pairs on the
    data line.  At the bottom use a 24V gigabit capable injector which
    puts the power on the data pairs.  We have a couple at PacketFlux,
    Chuck makes a couple, and there are others available.   The goal
    here is to get the 24V riding the data line along with the data. 
      So effectively you have two 10/100 capable links up with power
    on them.

    At the top, you reverse the process....  get a device which will
    pull the power off of the data pairs, probably one of them from
    Chuck.   (See 800-GigE-PoE as an example).   Plug the cable from
    the bottom in the PoE port, then build yourself a cable for the
    radio which puts the extracted power on 4,5,7,8 and the data pins
    where they belong.

    You could also use a single midspan Gigabit PoE injector at the
    bottom with power on all 4 pairs, then remove it using a similar
    one at the top.   Then your long CAT5 stays unsplit, and the
    splitting and PoE mess is all in a single cable harness.   To do
    this you'd take two cat5 cables, and then wire the 1,2,3,6 pairs
    from each cable into a single RJ45 (putting one on 1,2,3,6 and the
    other on 4,5,7,8) which gets plugged into the non-PoE side of the
    extractor.   Then the remaining 4,5,7,8 wires you'd connect to the
    power which came out of the PoE extractor at the top.     The
    bottom harness would be similar but for simplicity you can just
    put 24V in the injector and not connect 4,5,7,8 on either CAT5. 
     Now I think about this, this is what I'd probably do and just use
    a single 800-GigE-PoE top and bottom.

    On Tue, Jan 28, 2020 at 9:17 AM Matt <matt.mailingli...@gmail.com
    <mailto:matt.mailingli...@gmail.com>> wrote:

        I have a case where I need to power up two separate Cambium 24
        volt
        SMs on rooftop but only need 100base to each.  Its very
        difficult to
        run the second wire at this location which I need. Anyone know
        of way
        to split the cat5 at bottom and top to do this?  Not likely but
        thought I would ask.

-- AF mailing list
        AF@af.afmug.com <mailto:AF@af.afmug.com>
        http://af.afmug.com/mailman/listinfo/af_af.afmug.com


--
    - Forrest

-- AF mailing list
    AF@af.afmug.com <mailto:AF@af.afmug.com>
    http://af.afmug.com/mailman/listinfo/af_af.afmug.com



--
- Forrest

-- 
AF mailing list
AF@af.afmug.com
http://af.afmug.com/mailman/listinfo/af_af.afmug.com

Reply via email to