Author Topic: Possible fix for hardware deviation  (Read 121 times)

snowmanjax

  • Newbie
  • *
  • Posts: 19
Possible fix for hardware deviation
« on: October 14, 2017, 06:26:00 PM »
Devs and Forum Users Alike,

          I think there may be a fix for a few issues found with the PE. Loopback not found, and miswires may be fixed by this. What I've found is the jack on the PE itself is slightly deeper and wider than most jacks. If you plug in a premade cable with the "nub" boot on it (usually much shorter connector length), you may find you have to push harder to get the clip to click. When I remove the boot, and push the conductors of the cable towards the conductors on the PE jack, sometimes the issues disappear. That leads me to believe my findings are true.

          The following solution solves 5 out of 10 miswires (the others were resolved by either a firmware update, or RMA through the wonderful PE team). Quite a few of my colleagues have PE devices for themselves, and this is how I replicated the issues.

          The fix is using CAT6 cable, with a CAT6 connector end (no boot, as that end plugs into the PE), and a CAT5E keystone jack (for your cable connections) as an extension. This gets rid of a good amount of hardware deviation between all of the different cables and jacks, as I've found not all cables and jacks are alike. Also, it would lead to less damage to the PE jack.

          Can a dev/moderator tell me if the following cable would serve as a decent extension?

 https://www.amazon.com/UGREEN-Ethernet-Extension-Shielded-Connector/dp/B00QV0K2DI/ref=sr_1_5?ie=UTF8&qid=1507996475&sr=8-5&keywords=rj45+male+to+female

          I don't know if the resistance, reactance, delay, and/or impedance of the above premade extension would make the PE less accurate. TDR uses cable reactance if I remember correctly, and the PE wiremap/loopback adapter uses resistors and impedance to give signal back to the PE device.

          If anyone (devs and forum users alike) could please let me know if I'm completely incorrect, or if my findings could be accurate? Thank you all for your time in reading this, and hopefully this sheds some light on a few PE issues.


     -Snowman

JP

  • Newbie
  • *
  • Posts: 20
Re: Possible fix for hardware deviation
« Reply #1 on: October 18, 2017, 04:18:17 PM »
Not all CAT5/6 RJ-45 plugs are the same, there are MANY deviations in manufacture quality.  Specifically RJ-45's are sold specifically for stranded vs. solid copper wire, rounded vs. flat cable, CAT5e can be 24-26AWG, CAT6 can be 22-24AWG...

Or there is my personal favorite CAT5e version that seems to always crimp correctly regardless or stranded or solid, it's even advertised as so...
https://www.amazon.com/TRIPP-Stranded-Conductor-4-Pair-N030-100/dp/B008F533EC

I would think if something is not fitting in the jack on the Pockethernet it's because the RJ-45 plug is of questionable quality, not the jack in the Pockethernet which seems to be pretty good, at least on mine.  Note that it is a shielded jack, which means it has the little grounding tabs on the right/left sides, expecting a shielded RJ-45 plug to match, which *should* be the same exact size...
According to my calipers; Tripp-Lite standard RJ-45 is 11.58mm / Tripp-Lite shielded RJ-45 is 11.69mm
...so yes, shielded would be a little more "snug"

The cable you identified from Amazon should be fine, realize it may skew the TDR distance calculation a little (inches?) due to now having to go through additional connectors.  The wiremap test shouldn't care, unless the connector pins were extremely dirty, or the wire inside that short cable was the cheap aluminum/copper clad type (something China loves to do now) resistance should be close enough within tolerance that it passes the test.  Personally I would cut that UGREEN cable open to see if it's even manufactured to the specs shown in the thumbnail (foil wrap over each pair, separate ground line, all copper stranded conductors, unfortunately they don't specify the AWG of the wire)

Honestly, sounds to me like you may be using cheap patch cables, or cheap RJ-45 connectors and that's why things aren't consistent.