Behringer X32 + SD16, no connection over 150’ cable?

[GaryH]

@Dan Mortensen From my looking into the insertion loss topic, I interpreted the info as the opposite case. The insertion loss figures looked to be way lower for Cat8 than the specs Midas and Behringer refer to as acceptable. They say lower is better I think. The reality may be my ability to understand what I was looking at. I will see if I can locate that graph again and post it here...and the midas info to refer to with it.

Still-my real life experience with Cat7/8 has come up positive in every case-so far.For a lot of years. BTW I don't think the cable you bought is the same as I tested. Only because I recall not being able to buy it in more than 100 foot lengths-thus my 3 hundred foot spools linked together and tested. I did buy some nice flat cable I liked a lot (Empanar) but never tried a very long version using it. So maybe you bought that brand? Also I recall adding a 4th 50 foot section on top of the 300 and that did NOT fly. Anyway good talking again!

EDIT: Insertion loss less than 24@100mhz for the length used per MT. Looks to me like it would be minuscule even if the length were multiplied x5. I may not read this correctly from the chart below but I see that figure as less than 15db? (2.8 x5?) Anyway:

[Dan Mortensen]

Thanks for posting that, Gary.

So my limited ability to comprehend that graph results in my impression that it is saying there is a 63' cable being tested by a Fluke DSX-8000 meter with two parts, a near end master unit and a far end (of the cable) unit. (The one I had on loan was a DTX-1800, at the US$20k end of the scale, and I assume this is the successor some generations removed.)

That page says the maximum Insertion Loss to pass the test (all the tests are pass/fail) is 32.7 (what units are those?), and this cable had an Insertion Loss margin of 2.8, which is to say it measured 29.9 on that scale. Maybe.

What I further maybe get from it is that since a 63' cable is barely over 1/4 the length of a 250' cable like I would like to use as a worst case for a test like this (my snakes are 125', 150', 200', and 250'), I would therefore assume that the longer cable will measure significantly worse (4x?) than the shorter cable, and it wouldn't pass the Insertion Loss test, if not other tests.

I have no way of knowing how much of that is correct. If my meters gave numbers for the various tests, I could see the effect of length and type on each parameter, but they only tell you what speed is measured and if the results of the tests pass or fail the Category minimums.

FWIW my meters are SignalTek II and Byte Brothers Real World Certifier. The former only goes up to CAT5e, and the latter goes up to CAT6.

[GaryH]

@Dan Mortensen My thinking in regard to the specs posted is that it is in reference to the max speed of 2000ghz. We will never approach that speed with aes50 devices will we? The cable doesn’t dictate speed the sending devices do right? So the chart on the upper right would be the place to look and then down closer to the 100 side. The lower left of the graph. Taking that approximate number, say 5db and multiply by 4 or 5 would keep it within the range or very close. I could of course be totally wrong but it can’t be too far off based on my real world results with 100-300. I guess I just made things more muddy

[Dan Mortensen]

@Dan Mortensen My thinking in regard to the specs posted is that it is in reference to the max speed of 2000ghz. We will never approach that speed with aes50 devices will we? The cable doesn’t dictate speed the sending devices do right? So the chart on the upper right would be the place to look and then down closer to the 100 side. The lower left of the graph. Taking that approximate number, say 5db and multiply by 4 or 5 would keep it within the range or very close. I could of course be totally wrong but it can’t be too far off based on my real world results with 100-300. I guess I just made things more muddy

@GaryH You are exactly right IMO about what is also my point, which is that we need cable that passes signal at a comparatively pokey rate, so if cable within the prescribed length (and the length limit seems fixed regardless of the kind of cable, at least as far as the testers are concerned) that was designed and works to pass massive data rates is used at the pokey rate, it should work at least just as well as the cable designed for our use.

My uncertainty about the graph is where on the bottom scale (MHz) we are. My Signaltek II, which I used again last night to confirm that it doesn't give a number for Insertion Loss (or any other of the parameters other than length), I think only goes up to 5e but I'm pretty sure it's showing a data rate for the tests of 1000MHz. I could be wrong, and am busy for the next few days and not sure if I can look at it again.

Ha, doing a quick search for CAT cable speeds shows that what my meter is probably showing is 1000BaseT, which is the Ethernet type for 5e, and that 5e is 100MHz, so you are right, I believe, about where on the line we are, which would suppose that your muddying of the water was in fact settling some of the mud.

In the other forum, people are asserting that "AES50 is not Ethernet", which is definitely true at least in that you can't use a network switch (like a signal splitter: 1-many in, 1-many out). Not sure how further meaningful that statement is, since data rate and signal flow are, or seem like they would be, givens. It's likely also not true for the iteration of AES50 in our consoles in that the signal flow can be disrupted easier for us by messing with the cable in the right (wrong) ways, as compared to Ethernet. One of those right (wrong) ways was demonstrated and solved by Brian Wynn. Another was me pulling and squeezing a snake in the wrong way, which seemed to be from cable construction (all 4 pairs in a tube jacket as opposed to a jacket with internal longitudinal spacers of some sort).

I need to stop writing about this for now, thanks for looking into what I think is an important but overlooked subject.

[GaryH]

@Dan Mortensen

"In the other forum, people are asserting that "AES50 is not Ethernet"

I got slapped with that statement (or similar) too, IMO who cares about that, I never said it was or was not. AES50 is still networking (AOIP) and Ethernet is part of networking and besides that cat5e is about as Ethernet/networking related as it gets. As is the rj45 on both its ends. It still uses 4 pairs (the same 4 pairs) as Cat6/7/8 it still passes the same data down them. In my exchange they were saying "backward compatible" didn't count because aes50 is not real networking. Say what? I had said EVERY Cat cable iteration from 6 on up says "backward compatible" (they do) to ANY that precede it. Except apparently in AES50 use? Why? because of magical qualities? They want to argue that Cat6/7/8 is not capable of transmitting the very same data down the same 4 pairs with the same shields and same outer covering yet 5E somehow is. Makes zero sense to me. Cat5E was, I am to accept, so forwardly advanced that it is the ONLY possible cable that can be used in equipment designed and made some 20 years later? I don't buy in. (sorry for the rant)

Further they argue twists per inch make the big difference-which makes SOME logical sense with regard to distance EXCEPT there is no standard twist per inch. In my investigating I learned that it is random manufacturer to manufacturer (Cat5 or otherwise) and the real purpose is really only to stagger the twists so as to reduce crosstalk and EMI/ESD interference pair to pair. Stolen from the old telephone company where they used to twist the wires from pole to pole for exactly the same reasons. EMI and ESD reduction both Cat7 and 8 do better in spades because of all the extra shielding per pair. They also say that sync is an issue for Cat6/7/8 as well. Why? It uses the same 2 pairs (4/5 transmit +/- and 7/8 receive +/-) and uses the same 4 copper wires. Again magic? Drives me crazy. Then was the all important (per Midas) insertion loss bit we are now talking about. I said (perhaps now wrongly-but don't tell anyone) that the Cat7/8 insertion loss specs were better. Again from real life experience, it seems to be at the very least, within acceptable limits. Now that MT says 80M is the max, it MUST be very close One of these fine days I am going to find someone who has that expensive and modern tester and I can test my own builds to find out.

[Dan Mortensen]

@Dan Mortensen

"In the other forum, people are asserting that "AES50 is not Ethernet"

I got slapped with that statement (or similar) too, IMO who cares about that, I never said it was or was not. AES50 is still networking (AOIP) and Ethernet is part of networking and besides that cat5e is about as Ethernet/networking related as it gets. As is the rj45 on both its ends. It still uses 4 pairs (the same 4 pairs) as Cat6/7/8 it still passes the same data down them. In my exchange they were saying "backward compatible" didn't count because aes50 is not real networking. Say what? I had said EVERY Cat cable iteration from 6 on up says "backward compatible" (they do) to ANY that precede it. Except apparently in AES50 use? Why? because of magical qualities? They want to argue that Cat6/7/8 is not capable of transmitting the very same data down the same 4 pairs with the same shields and same outer covering yet 5E somehow is. Makes zero sense to me. Cat5E was, I am to accept, so forwardly advanced that it is the ONLY possible cable that can be used in equipment designed and made some 20 years later? I don't buy in. (sorry for the rant)

Further they argue twists per inch make the big difference-which makes SOME logical sense with regard to distance EXCEPT there is no standard twist per inch. In my investigating I learned that it is random manufacturer to manufacturer (Cat5 or otherwise) and the real purpose is really only to stagger the twists so as to reduce crosstalk and EMI/ESD interference pair to pair. Stolen from the old telephone company where they used to twist the wires from pole to pole for exactly the same reasons. EMI and ESD reduction both Cat7 and 8 do better in spades because of all the extra shielding per pair. They also say that sync is an issue for Cat6/7/8 as well. Why? It uses the same 2 pairs (4/5 transmit +/- and 7/8 receive +/-) and uses the same 4 copper wires. Again magic? Drives me crazy. Then was the all important (per Midas) insertion loss bit we are now talking about. I said (perhaps now wrongly-but don't tell anyone) that the Cat7/8 insertion loss specs were better. Again from real life experience, it seems to be at the very least, within acceptable limits. Now that MT says 80M is the max, it MUST be very close One of these fine days I am going to find someone who has that expensive and modern tester and I can test my own builds to find out.

Did I write that post without knowing it?



@GaryH

[pmaillot]

The difference between protocol, transport and standard ... Keep this for the academics and just do Music!

[GaryH]

@Dan Mortensen As @pvannatto often likes to say, great minds think alike ...well in my case, not all that great but persistent

On FB (private users group) it becomes a pile on session with those "Cat5 only" folks. Most of them are so ill informed its quite sad. Some well, just plain stupid (sorry). Some need the last and loudest word. In those cases you can never "win" because they will never say you are right and will ALWAYS be the last (often rudest) response. I lose energy and get weary in so doing the 'battle'. Also a bit of "sweet lemons" is afoot there-what I bought HAS to be the best because well, I already bought it and use it. There are also many users there who agree with my views (and the number is growing) but don't want the flack so don't join the fray. I don't blame them. I don't always respond (or even see) the posts, only once in a while when feeling brave. Sometimes I just enjoy stirring the pot a bit myself. Honestly Cat5e does work (I never said not) so I keep silent sometimes. As we know shielded cable with sheilded rj45's (shields bonded at BOTH ends) with ethercons are really the only important points to adhere to.

BTW one of the last documents from MT I found was this (and kind of interesting). It's from their "knowlege base" I'll summarise because I can't find it at the moment (fake quotes follow) "Lots of users are finding the manufacturers no longer carry cat5e and have had to switch to Cat6 and have reported positive results. Though Cat6 does work well we officially only support the tested use of Cat5E" Say what? It sounds to me like since they have not tested other cable they will just stick to Cat5E support only. However I think the most compelling official shift in direction by MT is the inclusion with all new mixers and stageboxes (anything that has AES50 ports on it) of a BRIGHT yellow label over the ethercon ports that says "Use Shielded cable ONLY" Conspicuously missing is the word Cat5E or any category type at all. Progress? Maybe.

[Dan Mortensen]

@GaryH Here are two links to MT articles about this subject:

https://community.musictribe.com/kba/ar ... musictribe
https://community.musictribe.com/kba/ar ... musictribe

I believe the second one is what you sort of quoted, more or less correctly.

What's missing here are the dates they were posted. The first one is most likely from the very early days of the console, the second one reads like I imagine an intern or somebody would write to be both technically correct and to cover their ass, when they know that there isn't a chance of anyone spending money/time to work out what works and what doesn't.

Keep in mind that the protocol is called AES50 because it was created by some kind of working group operating under the auspices of the Audio Engineering Society, spending effort over years (probably) to come to an agreement by all concerned parties about what it should say, similar to the National Electrical Code in the US. The AES released it at some point, several companies made products based on it, Behringer/Music Group/Music Tribe/whatever-its-name-was-at-the-time bought it at some point and is apparently now the only company putting that Standard into products.

They are still making products so they obviously have a technical staff now, but those people are hidden away somewhere, at least compared to when the X32 was first introduced. They were frequently in the forums, and sent staff around the country/world helping with product implementation. One of them came to the Ethernet cable workshop I did in 2013.

Now, the only time we hear from them (indirectly) is when there's a new Firmware update, which thankfully they are still doing 12+ years on, at least once in a while.

You and I think similarly, and we probably don't need to continue, especially since @pmaillot , who has done as much or probably quite a bit more than any private citizen to improve the usefulness of the consoles, has asked us to stop.

I'm getting busy anyway. Thanks for this forum for keeping the pointed conversations about these cool mixers (that I still enjoy using) alive.

[GaryH]

@Dan Mortensen Yup to both those links! Not really sure what @pmaillot was getting at but I didn't take it as stop. Maybe I should have I was happy to discuss these twists and turns again with someone who also cared about figuring out the bottom line. Thanks for all the info. Was fun revisiting old topics.