Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Brian Jojade]

I was pointing out what Behringer actually lists as a spec for their cable.  They Spec shielded cat5e wired to T568B.  I have never run into any other device that specified T568B alone as the wiring pattern.  I have seen devices spec both T568A or T568B as options.  I've never run into an instance where it actually mattered.  You'd think that the different rate of twists in the cable wouldn't matter.  But we have at least one reported instance on this thread where it seems to have now.

Point is, they have a specific listed spec.  If you vary from that spec, it may work just fine.  However, if you vary from the spec and have problems, the best advice is to get back to spec and see if the problems go away.  Sometimes what seems better isn't always better.

[Paul M Thomas]

Thanks for all the info.  I haven’t been with the rig to test since my first post. 

However, I did take the ends off a new ProCo Ethercon cable I had and noticed they are no longer soildering the drain wire to the metal of the RJ45.  I don’t spec the x32 all that often so it has been a few years since I have been down this road.  I know the ProCo Ethercons of a few years ago had a small drop of soilder on the drain....

[Dan Mortensen]

However, I did take the ends off a new ProCo Ethercon cable I had and noticed they are no longer soildering the drain wire to the metal of the RJ45.  I don’t spec the x32 all that often so it has been a few years since I have been down this road.  I know the ProCo Ethercons of a few years ago had a small drop of soilder on the drain....

This is significant, and I suggest you solder both ends of one of the cables but not the other cable, then see if swapping cables makes a difference.

Good luck!

[Paul M Thomas]

This is significant, and I suggest you solder both ends of one of the cables but not the other cable, then see if swapping cables makes a difference.

Good luck!

Ok,  got good news. 

I broke down and got the rig back in the shop.  I ripped apart a BBQ lighter to aid in performing repeated tests as Bryan did in his YouTube videos.

Had 3 ProCo Ethernet cables:
-2 that we’re prevuosly used in the field and both had produced failures.
-1 virgin cable

All 3 measured high impedance from ethercon shell to ethercon shell.

I soildered 1 previous cable’s and the virgin cable’s drain wire to the RJ45.  Shocking the ethercon socket frame and latch gave a failure rate in the neighborhood of 1 drop per 5 hits.  I swapped to the unsoildered cable and saw similar results of 1:5.  For me this concludes that the crimped connection must provide adequate connection to the drain wire,  however I’m going to say soildering provides a more robust connection.

Next I turned to the research on this board and others that state ethercon shell connection to cable shield is critical.  I soildered the shells on both ends of the virgin cable and a previous cable to the RJ45.  Measured impedance from inside of the ethercon shell to the other and verified good connection.  Performed esd tests with the lighter again, and saw 1 drop out of 100+ hits of the virgin cable, but the previous failed cable still netted a high ratio of drops (around 1:15) Puzzeled by this I reverifed connection from shell to shell, and found that it had failed and was measuring in the MegaOhm range. 

Resituated the cable and retested impedance to find low impedance again.  Decided the layout the cable and with the help of my wife,  was able to track down a specific spot in the cable were applying pressure would drop the shield connection.  Ok, so bad cable.  Caused by myself or from the factory,  it appears the shield and drain are damaged.  For sack of being thorough, I laid out the other 2 cables and repeated the test.  This time finding the virgin cable of food quality, however the additional previous failed cable also showed a spot where applying pressure caused the shield to go high impedance. 

So...  relating back to my original post,  it appears that my 2 cables used in the field with this rig both had a defect in shield continuity through the cable AND suffered from the known issue of no connection to the ethercon shield. 

Wanting to take this a step further, as I had used this brand of cables on 8-10 X32 installs and portable setups in the past without any KNOWN failures,  I contacted a close friend who had easy access to a portable setup and he measured continuity from ethercon shell to ethercon shell of both his primary snake and “backup” snake.  This somewhat proving that either ProCo cables used to have continuity through the shells,  or years ago I cashed in a bank load of karma and just received cables made to a higher, random standard. 

The E.E. in me wanted to know more about this failure mode, so I decided to remove the shell of one end of the ethercon cable and repeat the above tests.  This would deduce if continuity from one device to the other was required, or purely if only one device had to provide adiquite ground to the shield THROUGH the ethercon shell.  This test showed that shocking the connection with the ethercon shell produced virtually no drops (less than 1 in over 100 strikes). However shocking the exposed metal RJ45 of the other end,  produced around 1 drop per 2ish strikes.  This concludes, to me at least, that the ethercon shell must be providing a better shielding to the connection than the metal RJ45, and its role is more of an additional shielding to the “untwisted” portion of the cable than providing a better ground to the cable shield. 

If time allows, tomorrow,  I’ll try and test different combinations of using ethercon pass throughs in rack rails, and see if that hurts signal integrity.

I didn’t grab any video or photographic proof of these tests, and am posting purely to hopefully save someone else time and heartache.  However that being said, I attempted to be as systematic and methodological as possible to net accurate results in these tests.  I hope this helps others, but your mileage may vary...

ProCo cable on amazon seems to be a good choice for AES50,  however close inspection and slight alteration to the cable shells might be required.

[Dan Mortensen]

Ok,  got good news. 

Congrats! Glad you found a problem (two problems), and hope that solving it solves your problem at the events.

I think what you did with the Ethercon confirms Brian Wynn's findings, which is always good and better learned first hand.

Nice going!

BTW, your results with the RJ45 only (no Ethercon shell) would have been the same if you'd shocked the console, the stage box, or even a microphone with metal body connected to Pin 1 of the XLR. You don't have to directly hit the RJ45.

[Dan Mortensen]

This is anecdotal, but after building a cable to T568A and getting it certified via an Ethernet tester, and confirming continuity  between ethercon shells, my x32 and dl16 did not recognize each other. Wiring the same cable to T568B remedied the issue.

I would have said the same thing about T568A/B not mattering in the slightest before having issues with the cable..

Sent from my LG-H931 using Tapatalk

Picked up a 130' CAT6a cable wired to T568A and another wired to T568B today from Blue Jeans cable, which were made to order and tested on their Fluke DTX1800. The meter demands to be told if the cable is wired to A or B before testing; both passed certification under their standard.

I had the Blue Jeans guys test the A cable to the B standard while I was watching, and it passed and measured the exact same under the B standard as it had under the A standard.

That indicates that, as expected, there is functionally no difference in performance as long as both ends are wired the same.

I told them about your experience as described in your post and they felt something else must have been wrong with your cable that was fixed by redoing it.

We talked for quite a while about CAT wire, and I watched a technician put an end on in a fraction of the time it takes me to do the same thing. They are using a 3 piece RJ45: the main connector body; the insert to hold the wires in the proper order, spacing, and wire length; and a little piece that fits over the 4 pairs to hold them steady while you insert them into the insert. That was fun to see.

Further discussion about A vs. B: the only difference, as previously noted, is that the connector locations of the green/green-white and orange/orange-white pairs is swapped. Today I learned that there is NO standard for the locations of the pairs within the cable, so going clockwise the colors could be orange/blue/green/brown in one cable, and orange/green/blue/brown in another, and orange/brown,green/blue in another, and so on.

So having A vs. B does not give any advantage in putting colors next to or opposite some specific other colors.

I haven't had a chance to connect console/stage box together but it stands to reason that A or B will make no difference. I'll post if it does, but won't post if it doesn't.

[Brian Jojade]

Picked up a 130' CAT6a cable wired to T568A and another wired to T568B today from Blue Jeans cable, which were made to order and tested on their Fluke DTX1800. The meter demands to be told if the cable is wired to A or B before testing; both passed certification under their standard.

If it truly doesn't matter, then explain why the meter needs to be told if it's an A or B wire beforehand?  Why would it bother giving you that option?

My guess is that while the differences are extremely subtle, there may be enough difference that could throw one out of tolerance from the other.  In most instances it probably won't make any difference whatsoever.  However, at extreme ends of tolerance, there may be enough difference that it's going to cause problems.

[Dan Mortensen]

If it truly doesn't matter, then explain why the meter needs to be told if it's an A or B wire beforehand?  Why would it bother giving you that option?

Fluke Networks' Explanation

Probably the part about the US Government requiring all cabling to be done to T568A?

And since I'm posting again in this mostly-settled thread:

One of the things that's cool about getting cable from Blue Jeans (other than getting to chat with Kurt and Andrew) is that they give you a printout of the tested results for each and every cable from their Fluke DTX1800.

Attached below are the two printouts from the two cables under discussion; I've highlighted the 568A cable result, but it's on the top to compare.

Hmmm, the 568A performed a little worse than the B on a few tests. Could that be due to the difference in wiring?

Luckily, the project these were made for needs 3 cables, so I had a third one built and it was also to the B standard.

The second pic shows it, and it is more like the A than the other B, so the differences must be due to the fact that these are handmade.

So I'd have to say that no, the A and B are not functionally different given the functions that we think are important (i.e., ignoring backwards compatibility to an old AT&T standard and not being built on contract for the U.S. Government).

This also seems to show that there can be (numerically) considerable differences in specific test results that still result in the cables under consideration meeting spec.

Hope this helps.

[ProSoundWeb Community]