Cold Weather Problems?

[toddm1919]

I live in Chicago and last night was our 1st major sub-zero (-7 F) night. This morning when the lights are supposed to be in a steady on state, two of my 6 controllers are behaving erattically. One has certain lights that won't go on and others that are on that are not supposed to be. The other contoller loses communication even with the hardware utility. (When I send an on command, the lights turn on but then immediately go out.) I have tried resetting the boxes and rebooting the computer.

Everything was working fine last night, but this morning everything was going haywire. I was wondering if the cold weather could cause problems with the controllers. That seems to be the only difference between last night and this morning.

Todd

[Guest wbottomley]

Todd... the weather shouldn't make a difference. I know people in Alaska that don't have any trouble.

You might need to check your communication cables to see if the problem is there.

[garyfunk]

I've been runing in below 0 temps with no problems.

[Richard Hamilton]

Todd, are you using all Cat5 or are some of the controllers wireless?

[toddm1919]

I am using Cat5 cable. I was having communications issues with another controller that was randomly losing communcations. (Blinking LED) Dan suggested that I switch the speed of the network. That fixed the problem for a week. Now this.

Todd

[cmoore60]

Try doing a reset on ALL of your controllers.

Chuck

[toddm1919]

Thanks Chuck I will try it tomorrow, when the temp is above zero. It seems that there are a large amount of communication problems going on out there. I hope resetting the boxes fixes this one.

Todd

[ClydeL]

You now, this is the reason I wanted to keep my controllers inside. It was like 5ºF outside yesterday. Crazy... And I finally got the bugs worked out of the system. My biggest problem is to keep from tripping over and un-plugging the extension cords. Other than that, It is great! I wish you the best.

[basis21b]

My LOR controllers are out in the elements and have been running in some pretty severe temperature ranges. On Dec 14 at 1:00PM it was +37ºF and raining. By the time the display was turned off at 10:00PM it was snowing and the temperature was -3ºF. On Dec 15th the temp was -17ºF and on the 20th it was -15ºF, during the time the show was running.

The only problem I have had this year was because the Easy Light Linkers (ELL) were having issues with communication. This was due to the fact that I moved the show computer to another room. The problem was solved by changing the location of the ELL.

[radioguy1007]

Todd:

I'm a few miles north of you and made it through the weekend snows and cold with only a couple minor mini light issues. Everything on the ground is buried in white, but still lights up. All of my controllers are remotely located to keep extensions as short as possible. I also have numerous ELL links, and on Friday night I noticed the controllers nearest my south neighbor were acting up. On Saturday all was good - but I had noticed that my neighbor had guests from out of town since late on Thursday and Saturday they were not there. Well, their grandkids visiting had some radio controlled toys that were on the same 900 MHz frequencies as the ELL's. I am glad it was a simple "fix".

Mike

[toddm1919]

Thanks for all the responses. Resetting the boxes did not fix the problem so here is what I tried.

First I switched from my dedicated computer to a laptop, but had the same issue. Next I switched out the network cable from the USB485B Adapter to box 1..same issue. Finally I figured I would concentrate on the one box that was having the most trouble so I disconnected all of the controller boxes and directly linked to the box 3, one that was acting up.

Using the hardware utility, I turned on each of the 16 channels one at a time with no issues. I then turned on channel 1, then 1 and 3, ( I skipped 2 because I could not see it from inside) then 1,3, and 4, and so on. I made it up to 7 without incident. I then went on to 8, it blinked on then all channels went out. I went out and checked the light on the box and it was blinking. I hit refresh in the hardware utility and went throught the same senario with the same result. I then tried other combinations of channels but the most I chould get on was 10 channels.

Assuming a communication problem because of the blinking light, I went to work and got another USB485B adapter that I use there (I am a school teacher and have a 16 channel show in my classroom.) Anyway, brought it home and tried the same test with the new adapter. I was able to get 15 channels to go, but on the 16th, it blinked on, then all of the channels went out.

Knowing that I don't have all 16 channels on at once, I reconnected in my original configuration and ran my steady on sequence (the one that I originally had problems with). It worked fine. Feeling good so far, I ran one of my shows that has 8 individual musical sequences and about 16 minutes total length. It was working flawlessly for 7 and on the 8th controller 3 studdered and recovered, but controller 1 went dark. I disabled shows edited my schedule so my steady on sequence would run for the rest of the night, saved and re-enbled shows. It has been running for 2 hours without a problem. I am running out of things to try, but it still seems like there is a communications issue.

Any other suggestions?

[-klb-]

Any compact fluorescent lamps? Any 2008 CDI LED's? Is the cord running to the controller large enough gauge to cover the load? To me, it mostly sounds like too much noise being generated by the loads, and knocking out communication, but there are other possibilities, like too much shift in neutral potential, or too much voltage drop, creating more offset than the RS-485 chips can handle...

[toddm1919]

No LEDs or compact fluorescents, just minis, c7s and spots (incandescent and halogen). We use 16/3 outdoor extenstions for all of our connections. They are rated for 10-13 Amps depending on length. Is there a way of measuring the other things or is there another adapter that can handle the offset?

Thanks,
Todd

[-klb-]

The voltage offset is pretty easy to measure... Use a digital multi meter to measure between ground and neutral at the controller inlet, with channels turned on to just before it looses communication, and it will give a good idea.. What we don't know is how much offset the controller will handle.. There is an isolated USB-485 adapter, but it is pricey, and it does nothing for the offset between one controller and the next...

Am I understanding correctly that you are feeding both sides of the controller from a single 16/3 cord? Any idea how much length to the controller?

Also, running heavier cord to the controller will be much cheaper than the isolated USB adapter...



- Kevin

[DanR]

Todd,

I am seeing the same type of problem which has occurred since the temperature dropped below zero. I first noticed there was a problem when I turned on all the channels to make my nightly check for burnt out bulbs. Four arches and two floods were repeatedly turning on and then back off - this didn't look good. Controllers red LED was turning off and then back on also. Looked like a communication issue, but I don't know how much traffic is generated by turning on all channels.

And now the part that doesn't make sense: One segment on the 3rd arch seemed to be the channel that was causing the problem. If this segment was turned on with the other others, communications would be lost. This segment worked fine by itself. I wanted to try different combinations of channels, but after minutes of testing, this problem fixed itself (why, how?). I could now turn on all channels without problem.

So now I watched the entire show by staring at the controller LEDs for a loss of communications. There were two parts of sequences causing problems. I would guess there was a lot of communications going on during these parts where 28 channels (4 arches) are fading at a time. Attached is a screen capture of the sequence causing the problem.

Both controllers are mounted in a garage. Maximum current for both controllers is about 10 amps. Both controls were initially on the same power circuit. I later tried putting each controller on its own 20 amp circuit with no change. Power cords are 12/3, about 3 ft in length. I also tried to up the baud rate on the USB adapter to 19200, with no change.

Tomorrow, temperature is supposed to be in the mid 20's, so I am not going to change anything until I can run a test then.

-Dan


Attached files

[George Simmons]

For what it's worth, I've had no issues that appear weather-related as cold at -20 so far with all cords and cables under a foot of snow. The only think I notice is every once in a while a song (not always the same one) will have irregularities in certain parts of the song, my mini-trees for example, but during the rest of the song and the following songs, they'll work fine. Then in another couple of days, I'll notice the same thing on yet another song. I have no clue why, but re-booting my PC seems to straighten it out.

[-klb-]

The PC reboot is an interesting idea. It also occurrs to me that incandescent lights might manage to draw more power when cold, which might have an impact...

[DanR]

I tried the same test that I was having problems with last night. Last night air temperature was -4. Now, air temperature is 23. I am having no problems with the two sequences any more.

I did not reboot the show computer - it is in the same state as last night. Also, I did not unplug the controllers since having the problems. No cables were touched.

So, some possible ideas/questions are:

• 
  
• Our electric power has a lot of noise or is below 110V when very cold out. I wish I would have measured this last night. Is there a cutoff point where input power would affect communications?
  
• Is termination of the RS485 line at the end of the chain not required at the comm speed we are running at? If termination was added, would this help clear up noise issues?

We are supposed to remain in the 20s-30s thru January 1st, so hopefully this problem will not appear again.

-Dan

[-klb-]

Utility power voltages might dip with more heat running.. Even if most of it is not electric, most of it is still fan forced heat.. Plus in severe cold, heat pumps (if there are any up there) switch from being fairly efficient to straight electric heat, so that load goes way up...

Termination handles reflections off the end of the cable, not noise so much... If one wanted to try, it should be a 100 ohm resistor, I believe on the center two pins...

I also wonder if it is necessarily always a com issue, or if maybe there are some controller reboots going on...

[DanCampbell]

Just some thoughts...

Could the line voltage is dropping enough when you kick the load in that the controller's power supply dips too low for it to be stable? At this point I'd try putting small loads, say a single string of mini's on each channel of the controller having the most problems, and see if communications stay up.

If you have a kill-a-watt (very handy) or equivalent, you can plug the controller into it to measure the line voltage and steady on load.

Without that tool I might try plugging a power strip into one side of the controller's extension, plug the lights from the same side into the power strip, and measure the voltage there with the lights on and off. This will at least show you the voltage drop you're seeing up to the controller.

If your steady on sequence ramps the lights up slowly, you're minimizing any surge of turning the lights on. Run a show and there's lots of spikes that the line has to handle. Unless the sequence is for a lullaby.

Hope you can resolve the problem soon.