We have a new site we deployed that is having issues passing sync over power from the CMM. The CMM Shows sync properly and is tracking 9+ satellites how ever all the aps connected to the thing show no sync from power. Distance from AP to CMM is 250ft.
Any one ever experience this and if so how was it resolved ?
Joe Camadine
Are you close to an AM tower?
Are you using shielded cat5? We have several towers with AP’s 300+ from CMM and have no trouble with timing.
Cable is not sheilded. I have been informed we have runs longer than this using a CMM. we are going to re-end cables and hope for the best.
Will update if this does not work.
Thanks
Joe Camadine
Oddly we did some troubleshooting. its not a cable problem as far as length goes it must be some kind of ingress signal or something. We used exact same outdoor cable we usually use and made a 260ft patch cable (exact length of runs to aps) and that works perfectly. We thought it might be the patch panel so we punched it down to that with no success. We removed surge suppressors, that didnt help. The only thing that makes sense here is some odd termination problem to RJ45 or some kind of cable interference causing this. I will probably end up whining to motorola.
supernull wrote: Oddly we did some troubleshooting. its not a cable problem as far as length goes it must be some kind of ingress signal or something. We used exact same outdoor cable we usually use and made a 260ft patch cable (exact length of runs to aps) and that works perfectly.
I may be attempting to teach grandma to suck eggs here, however when doing short cat 5 cables you can do a pin to pin match up and they will work fine, however for longer runs you need to flip a pair, IE use one lead of a pair and another lead from a second pair for your TX and RX signals.
Failure to do this on long runs (100+ foot) results in a cable that looks find on a cable tester but that will not pass traffic.
JayR
supernull wrote: I may be attempting to teach grandma to suck eggs here, however when doing short cat 5 cables you can do a pin to pin match up and they will work fine, however for longer runs you need to flip a pair, IE use one lead of a pair and another lead from a second pair for your TX and RX signals.
Failure to do this on long runs (100+ foot) results in a cable that looks find on a cable tester but that will not pass traffic.
JayR
I know I'm not alone here, but I've been using T-568A & B pinouts on 100-300' runs for over 12 years and never had a problem with them passing data. Would you be able to provide any reputable documentation to back up your claim?
[/quote:25ageei2]
jwelch wrote: [quote="supernull":25ageei2]I know I'm not alone here, but I've been using T-568A & B pinouts on 100-300' runs for over 12 years and never had a problem with them passing data. Would you be able to provide any reputable documentation to back up your claim?
You are 100% correct, a reference can be found at http://www.duxcw.com/faq/network/diff568ab.htm.
The confusion was on my part, for lo those many years I had been making up my Cat 5 cables just wiring the pairs straight across. Never had a problem. And real easy to verify on the test set.
Then I was helping out at the convention center and made up a couple of 250+ cables and they would not pass traffic. Whipped out my cable test set and they checked fine, exactly as I had wired them.
Fortunately the guy I was helping out is smarter than I am, he looked at the color codes and told me to swap the pairs, which is pretty much what T-568a/b does. And they worked fine.
Looking into it afterwords it appears to be a cross talk problem, without the swap-over the near end TX cross talk is enough to overload the near end RX signal coming down the copper pipe.
Since the amount of cross talk is constant, they work fine for shorter cables (up to around 100') because the RX signal is still higher than the X-Talk, past that point cable attenuation has its way and the RX signal drops to the point that the X-Talk wipes you out.
This lesson was really driven home, when my pal told me what the problem was I bet him 50 bucks that he was wrong. <sigh>
JayR <poorer but wiser in Vegas>
The moral here might be that just because you have been doing something a certain way for a long time does not mean you are doing it right.
Unfortunately the link you provided does not substantiate your claim. It is a generic T568A/B wiring guide web page.
For clarity sake, you initially wired the “convention center” cables for 568B on both ends and it did not work. Once you used 568A on one end and 568B on the other (eg crossover) it worked fine? If so then it had nothing to do with the length of the cable.
jwelch wrote: Unfortunately the link you provided does not substantiate your claim. It is a generic T568A/B wiring guide web page.
For clarity sake, you initially wired the "convention center" cables for 568B on both ends and it did not work. Once you used 568A on one end and 568B on the other (eg crossover) it worked fine? If so then it had nothing to do with the length of the cable.
Life is pretty short, and at my age not much of it left, certainly not enough left to pick nits, I agreed with you that the standard ways as shown in the reference link I provided are the right way to go, and that wiring pin to pin, as I used to do, is not the right way to go.
I have no idea what more I can do to make you happy, and frankly don't intend to attempt to. Please do not respond to this, I may have croaked by the time a response would get here.
jayr
You appear to be taking this personally, and quite honestly there’s no need get all bent out of shape over this discussion. Everyone is hear to learn something. I am not doubting that a 250’ crossover worked for your particular application. However, when you come in and state that your way (eg. a crossover) is the [u:1ypl0msl]only[/u:1ypl0msl] way data will pass on 100’+ cat5/5e cable runs you should expect to be questioned about it. You also provide no written data to support your claim; in fact you’d be hard pressed to find anything at all.
supernull wrote: Oddly we did some troubleshooting. its not a cable problem as far as length goes it must be some kind of ingress signal or something. We used exact same outdoor cable we usually use and made a 260ft patch cable (exact length of runs to aps) and that works perfectly. We thought it might be the patch panel so we punched it down to that with no success. We removed surge suppressors, that didnt help. The only thing that makes sense here is some odd termination problem to RJ45 or some kind of cable interference causing this. I will probably end up whining to motorola.
So you made a 260' cable and tested it on the ground and timing worked? What else is on that tower? I've seen this issue occur when shielded cable was not grounded properly enough. The induced interference can kill the timing pulse but still allow data to pass through it. Are you showing any ethernet errors on the statistics page of your AP's?
jmaysww wrote: So you made a 260' cable and tested it on the ground and timing worked? What else is on that tower? I've seen this issue occur when shielded cable was not grounded properly enough. The induced interference can kill the timing pulse but still allow data to pass through it. Are you showing any ethernet errors on the statistics page of your AP's?
Jmaysww,
Grounding the shield on the Ethernet run? I have tried to find an authoritative source to tell me how to go about doing the grounding without any luck.
I believe that at all costs you do not want to ground the shield at both ends of the run, this will create a potential ground loop that will allow external sources to induce currents in the shield, which will then get into your cable. Worst case would be a lighting strike which could have some really amusing results.
We currently use UV proof exterior grade cat 5 cable, can dig out the part number if that is required. There is no provision for bringing the shield out through the connector, so I break out the shield at one end folding it back and then treating it with shrink wrap to create a pig tail.
I then use the shield lead as a pig tail, and run a (very) short jumper cable to take it to the earthing lug on the surge suppressor.
This was the standard method of treating the overall shield for 124 ohm balanced cable runs in the old days.
If there is a better way to handle grounding the o/a shielded cable I would sure like to know about it
jayr
Sounds like the sync by power port is blown. I see AP’s on eBay all the time advertised with that problem.
Alright, not even motorola knows whats wrong and they admitted that. We ended up buying some ‘Cell Tower Cat5’ which is shield tape and braid AND REALLY EXPENSIVE. We have it grounded on the CMM side of course. As previously mentioned everything works perfectly on the ground. We plan on doing a labor intensive check to see where it fails by slowly moving the ap up the tower (Motorola Suggestion).
Thanks,
Joe
I’d be looking at what VLAN1 had asked. Is there an active transmitter at this site (ie…FM or AM radio transmitter) if there is then that is probably your problem.
If things work on the ground but not on the tower then the tower hows something on it that is causing some EMI.