Author Topic: Different Zeus problem(?)  (Read 3283 times)

Offline JZ

  • Jr. Member
  • **
  • Posts: 21
Different Zeus problem(?)
« on: December 15, 2015, »
I think the issue I am having is a little different than what I read from others posting here.

Using Zeus V1 for 2nd year with no prior issues.  Using 6 outputs with 80 nodes per output.

Tested fine with a couple strings in the garage.  Worked fine during testing once installed in yard.  During the last test before shutting down for the evening, the first string was dark and the rest were green and purple (green node, purple node, green node, etc.).  No response to any pixelnet commands.  Disconnected and brought it inside for troubleshooting.  Reviewed everything posted here but didn't help with my issue much.  I did eventually figure out that the PIC for output 1 had a problem (maybe dead).  Replacing it with a PIC from a spare output channel fixed the issue.  I think the PIC was taking down communication to all of the PICs and they defaulted to green/purple.

Everything work fine for 2 weeks until yesterday.  Dead mega tree.  We had a ton of rain but GFCI was not tripped.  Opened the box and power LED was out. Unplugged AC power and plugged back in now 3 of the 6 strings are dark (strings 2, 5, 6) and other 3 are green/purple.  Based on my history I will try removing those 3 PICs and replacing with spares to see if functionality comes back.

If that is the issue again - what is taking out the PICs?  Is it cycling of the power supply?  Normally my power is on all of the time but somehow this particular Zeus lost power this particular day.  Now that I say that I remember we had quick power outage during the night (clocks were flashing on the morning).

Others have said their Zeus works then on later tries doesn't (sounds like they cycled power but not sure) and they say their PICs are corrupted.  But I think the symptoms I have are slightly different (i.e. strings don't go all dark or all white).

I have not tried reprogramming any of the PICs at this point.

Anyone seen anything simiar?

Thanks!

Offline taybrynn

  • Sr. Member
  • ****
  • Posts: 2042
    • RockinChristmas
Re: Different Zeus problem(?)
« Reply #1 on: December 16, 2015, »
I don't know but it seems that pic corruption happens somehow ... I've seen just about every permutation possible ... from stuck on, to stuck off, to a mixture like your seeing.

Things I have wondered about with the zeus:
1) PIC chips: Do they all need to have the default firmware in them if they are NOT programmed to a certain channel range?
    It doesn't seem to really matter, but I do flash them all just to be safe and just program the channels I want to use
2) CHANNELS used: does it matter if you use say channels 1,2,3,4 ... or is using 6,8,9,11 just as good?   Issues with skipping?
      It doesn't seem to matter
3) PIC question: could a corrupted PIC in say channel 1, affect the operation of other channels, like 2,3,4 ?  Or no relationship?

The wild part is I see them all going out together, never a partial failure.  That makes me think they are linked together somehow and the problem seems to cause all other pics to be affected somehow.
     I'm convinced that either voltage fluctuations or out of bounds pixelnet data can corrupt the pics somehow.  I've had xlights sequences with imported LOR sequencing layer where
     a certain sequence will cause a zeus to get wiped out.


I'm also not sure if power supplies can cause these problems, or bad lights could cause ... not sure at all
« Last Edit: December 16, 2015, by taybrynn »
Scott - Castle Rock, Colorado   [ 2 homes, 100% RGB in 2016; since 2008; over 32k channels of E1.31 ]
You are not allowed to view links. Register or Login

Offline caretaker

  • Sr. Member
  • ****
  • Posts: 1770
Re: Different Zeus problem(?)
« Reply #2 on: December 16, 2015, »
Doing a quick search of the internet I did see that brown out conditions "can" cause a PIC to have memory corruption.
Jeff Squires
You are not allowed to view links. Register or Login

Offline JZ

  • Jr. Member
  • **
  • Posts: 21
Re: Different Zeus problem(?)
« Reply #3 on: December 17, 2015, »
Plugged my Zeus into power one last time before bringing it inside for troubleshooting and only channel 2 was dead the rest were static green/purple.  My first thought was "I bet the channel 2 PIC is bad."

So I pull it out of socket first thing and everything else works fine.  So I replaced it with the PIC from channel 16 and ran the utility set the string numbers, etc.  (When you move a PIC it will keep it's old string data so you have to run the utility again.)

So yes, 1 bad PIC can lock up all the channels.  I suspect because they all share the same data bus from the 485 IC.  I don't use PICs much anywhere else so don't know all of their details and haven't looked up info, pinout on this particular chip.

At this point I'm thinking the power supply is part of the issue - brownout, startup or otherwise.  Typically these little 8 or 16 bit micros, whether PIC, Atmel or other, are used in embedded systems that will have some local power supply regulation.  I wonder if having a local regulator on board (that took power from the 12v supply and regulated down to the 5v) instead of running from the 5v of the ATX power supply directly wouldn't avert this issue.  Unfortunately this board doesn't have the "handles" to make that an easy mod.

It might be interesting to know exactly what model of power supply people are using that haven't had any issues.

My power supply is a Logisys 480W, PS480D2 that I think I bought from MCM Electronics a few year's ago.

Offline t.jo13

  • Coop Manager
  • Sr. Member
  • *
  • Posts: 626
Re: Different Zeus problem(?)
« Reply #4 on: December 17, 2015, »
I use corsair 750  and they work great

Joe

Offline taybrynn

  • Sr. Member
  • ****
  • Posts: 2042
    • RockinChristmas
Re: Different Zeus problem(?)
« Reply #5 on: December 18, 2015, »
The cheap powersupply brands are not a good idea, generally speaking.  I've regretted it every time. 

THe brands I have found work best for there are the Rosewill, Corsair, Antec, XFX.   Single Rail, 80 plus rated.

I can confirm that Sentey is BAD NEWS ... I have some 750w sentey and they have a "feature" which turns them off when there is no load ... and they absolutely stink for Christmas light use.     They also have another nice feature where if you lose power, they wait a full 10 minutes befeore than will restart.  Horribly painful for in the field, let me tell you.
Scott - Castle Rock, Colorado   [ 2 homes, 100% RGB in 2016; since 2008; over 32k channels of E1.31 ]
You are not allowed to view links. Register or Login

Offline dowdybrown

  • Sr. Member
  • ****
  • Posts: 358
    • Gleannloch Christmas
Re: Different Zeus problem(?)
« Reply #6 on: December 19, 2015, »
I have 7 Sentey power supplies - no issues so far. Before deploying them I tested them at 50% load for extended periods in the battery box, monitoring temperature and fan noise and was very pleased with the results. I have not experienced a shut off at no load.

Matt
Matt Brown
You are not allowed to view links. Register or Login