Dying USB485??

[Cram72]

Hello everyone. I requested help from LOR, but I was hoping to get my troubleshooting done today so I thought I'd ask for help here. Thanks ahead of time!

The problem:

LOR hardware utility search for controllers will not find all the controllers. 

My system

I use a dedicated desktop PC to run the show and a laptop to use outside for performing light/channel checks. I have seven - 16 channel controllers(#1 - #7). My input cable from the house plugs into controller #1. I have two banks of controllers. Bank one contains controllers #1 - #4 and bank two contains #5 - #7. The banks are connected via a 25 foot cable. Also, the cable from inside house is made up of two 25 foot cables connected together with a coupler. In all, there is 95 feet of cabling which includes the short cables connecting the controllers. All cables are cat 5. I have used this exact set up for that past 3 years and have never experienced an issue.

My findings

(Note: during the checks in which I'm using my laptop, I'm connecting to the controllers with a 3 foot cat 5 cable from the USB485)

From the PC, the controller search only finds the first 5 in sequence, #6 and #7 are not found. Yet I did noticed the green LED's on ALL the controllers are on steady. It is my understanding that if an LED is on steady then it is connected to the LOR network, correct? So, suspecting I had a bad connection or cable, I took my laptop outside and connected to controller #5 (where the cable from #4 plugs into) and ran the controller search again - to my surprise it found #5, #6 and #7. I checked all the connections and found nothing visually wrong. I did many other checks, and tasks including replacing all the short cables, the cable connecting bank 1 to bank 2 , and the cable from the coupler in the house. When I ran the search from the desktop again, LOR was still only able to find #1 - #5. Then I took my laptop and unplugged the connection at the coupler and plugged the cable going to controller #1 directly to the USB485 and again, only #1 -#5 were found. 

Back outside, I stumbled onto something. At one point, I connected to #7( the last open port) and did a search. The system now found controllers #3 - #7. So now that I'm plugged into the last controller, LOR is not seeing controllers #1 and #2. Coincidence? Then for the heck of it, while still plugged into #7, I disconnected the cable coming from the house (at controller #1) and ran the search. LOR now found all 7 controllers. What the heck! Bad jack on #1 controller? Nope! I plugged into controller #1 with the laptop and ran the search, and it found all 7 controllers. 

In the end, after installing all new cables, I discovered that I can communicate with all the controllers and the system works fine, BUT only if I used a 3 foot, or 6 foot cable from the USB485 to #1 controller. Or controller #7 as long as the cable from the house is disconnected at controller #1. But if I swap out the shorter cable at the USB485 with a 25 footer I will lose the last two controllers. I even tried this: Using a 3 foot cable connected at controller #1, the system works fine. I decided to plug in a 25 foot cable (the other end not plugged into anything) to the second available port on the USB 485. Guess what happened? The last two controllers were lost. I tried 3 different 25 foot cables and they all did the same thing.

Is it possible that something is wrong with the USB485? Or the power cable? Maybe they can't supply enough power on the longer runs? Or could it still be a bad controller/jack? It almost seems like the system can't handle the extra cabling, all of the sudden. It has always worked fine.

Any help would be great

Thanks

C Ram

 

[PhilMassey]

In my experience the more likely suspects are the RJ45 jacks themselves,  Cable lengths are fine.

See  this thread

 

Does an actual sequence play on the controllers, because it has been reported that the Hardware Utility does sometimes fail to detect a controller. Try Playing a sequence.

[k6ccc]

OK, that is a strange one.  My first thought is that since your total cable length is only about 2 1/2% of the allowed length for a RS-485 network, cable length SHOULD not be an issue.  However your testing shows that cable length certainly appears to be causing a problem.  It sounds like you have made very detailed tests and documentation.  I'm largely in agreement that at this point the RS-485 adapter is suspect.

My suggestion would be to buy a USB-485-HS adapter.  Even if that is not the problem, it would give you a spare adapter which in my opinion is one of the most important items to have as a spare - with a dead adapter your whole show is dead.  It would also allow you to rule out the adapter very quickly.  I recommend the High Speed version because one of these days you will need one as you add more channels (especially pixels).  You don't need the more expensive Booster version that you currently are using.

I see Phil just responded.  Your testing would largely rule out a RJ-45 connection - I think.

7 minutes ago, PhilMassey said:

Does an actual sequence play on the controllers, because it has been reported that the Hardware Utility does sometimes fail to detect a controller. Try Playing a sequence.

I completely agree with this.

One last suggestion.  Build or buy a 50 or so foot cable and get rid of the coupler.  Can't tell you how many problems have been traces to those.   It's just two more points of failure.

 

[jfuller8400]

I doubt its a problem with the 485.  I've had one fail on me (and burned out two others - my own fault) and if it was a 485 problem it wouldn't communicate with any of the controllers.  I agree with Jim, get rid of the coupler.  That can cause all sorts of problems.  You might also consider adding a terminator off controller 7.  Your cable lengths aren't all that long that you should need one but sometimes they help.

I also agree with Phil.  The HU sometimes doesn't see all of the controllers but a sequence can run fine.  

You could also try a reset on Controllers 6 & 7.  I had a similar type of issue with a CMB24D this weekend - the light on the board was on solid (indicating that it was communicating with the pc) but the HU didn't see it and it wasn't running any sequences.  Did a factory reset on it and it returned to normal operation, so who knows what was wrong.

 

[Cram72]

Thanks for the replies. 

 

@PhilMassey Yes I did run a sequence and the lights only worked on  #1 - #5. Thanks for the thread! Do you know why the HU will not detect all the controllers. Remember, HU was able to detect all of them, but only under certain condtions.

@k6ccc I have had issues with the jacks before - had to replace one already. That is great advice about having a backup USB adapter, I'm ordering one right now.

I have a couple more ideas to try. I have other unassigned controllers that are not in use right now. (for expansion in the future, or as backups). I am going to take one of these controllers and start swapping out each of the controllers in use and then run the HU search after each swap. What have I got to loose at this point, except time....  I'll check back later after I'm done with this task. 

Thanks again!

C RAM

[Cram72]

@jfuller8400 Never thought about resetting the controllers. I look into that - thanks!

 

C RAM

[jfuller8400]

Just out of curiosity, what type of controllers are 5 & 6 (CTB16's, CMB24D's,...).  I'm asking as I just got back in from running a check on everything for tonight's show and that same CMB24D was acting up again.  Similar problem in that the light was solid on the card but the HU couldn't see it and it wouldn't run any sequences.  So I did a factory reset only this time it didn't make any difference.  So then I swapped out cables and it still didn't make a difference.  I finally reset the unit id (via the dip switches on the card) and it started to respond.  So it's almost like it lost it's unit id for some reason.

Depending on the type of controllers you have and how you set their unit id's, it may be another thing for you to check.

 

[Cram72]

@jfuller8400 All my controllers are CTB16's.

I reset and reassigned controllers #6 and #7, and nothing changed. I also decided to reset #5 and afterwards I noticed something odd. After reassigning it the same ID, and reconnecting all cables in the system I ran the controller search. This time only 4 controllers were detected, #1 -#4. I double and triple checked everything - nothing bad found. Thinking I messed something up when resetting/reassigning I plugged my laptop directly to #5 and it detected it as #5. ARRG! But then I moved the cable to the other jack and HU would NOT detect it. I tried the search several times before switching the cable to the other jack and then HU detected it. I went back and forth between the jacks and each time the same jack was not working. How did one jack go dead by resetting the controller? I know both jacks were working prior to resetting because of all the checks I had done earlier in day(read my original post for details). In all of those checks there was nothing to indicate a possible bad jack on controller #5. I will admit that at the very beginning I did suspect a bad jack in #5 because both controllers past it were not being detected. But after various checks and observing the results I thought a bad jack was not a possibility anymore.

So I decided to take a different unassigned controller, and connect it between controller #4 and #6 ,and guess what? All the controllers are now being detected, and the system is working fine. I even changed out the new cables I had put int with the old ones and everything still works fine.

So, it appears that a bad jack in controller #5 may have been the culprit all along. That or there may be an issue with the board? I'm still trying to understand what resetting the controller did to affect the one jack. I am 100% absolutely positive of my findings earlier in the day. I take notes as I troubleshoot things so that I have references later. If one of those jacks was completely bad from the beginning, then HU wouldn't have been able to detect all 7 controllers, which a various times during my checks, it did. When I have more time I will do a more comprehensive check of the board. I'm usually pretty good at troubleshooting things, but this one threw me for a loop. I hope my troubles are over for the season.

Thanks everyone!

C RAM

 

 

[Thebug]

Glad you found it!

[k6ccc]

55 minutes ago, Cram72 said:

I'm still trying to understand what resetting the controller did to affect the one jack.

Resetting it, did nothing to affect the jack.  Most likely in the process of the reset, the cable in that jack got moved just enough that a flaky jack became a bad jack.  If you carefully study the PC board for the controller, there is nothing except copper PC board traces between the two RJ-45 jacks and the phone jack (for controllers that have a phone jack).  There is no electronics that can cause one jack or the other to fail.  Of course the controllers RS-485 transceiver chip taps onto that connection between the two RJ-45 jacks and the phone jack, but it does not interrupt the signal in any way.  BTW, pins 1 & 2 of the RJ-45 jacks also pass through the PC board with nothing but copper PC board traces in between.

 

[TheDucks]

If you need an emergency 'fix' when 1 jack is still good

https://www.newegg.com/Product/Product.aspx?Item=9SIA6HD4091563&cm_re=rj45_splitter-_-9SIA6HD4091563-_-Product

and a very short (~1') CAT5  This is a T, not a more commonly found splitter