SS don’t flash when programmed

[Jimm]

I have two 128 node strands one works great and the other does not.  When I program the controller using RJs Smart String Utility the lights stay on and don’t flash.  When I try to use the string with Vixen it acts erratic.  If I turn on one light at a time it works ok but when I turn on many at a time colors change when they are not supposed to and then some times it locks up.  Once it locks up (some on and some off) I have to unplug it before it will do anything again.  I swapped out the controller with the working one and the problem stayed with the nodes so the controller is ok.  Any Ideas?
Thanks,
Jim   

[rrowan]

I have two 128 node strands one works great and the other does not.  When I program the controller using RJs Smart String Utility the lights stay on and don’t flash.  When I try to use the string with Vixen it acts erratic.  If I turn on one light at a time it works ok but when I turn on many at a time colors change when they are not supposed to and then some times it locks up.  Once it locks up (some on and some off) I have to unplug it before it will do anything again.  I swapped out the controller with the working one and the problem stayed with the nodes so the controller is ok.  Any Ideas?
Thanks,
Jim

I had the same problem

You have one or more defective node(s) causing this issue

I went thru and tried to take heat measurements, voltage measurements. Only lit sections of the string in quarter sections to narrow down the problem area. What I found was one node would always flicker even if it was suppose to off. I remove that node and then the string worked right along with the lights flashing when programming with the SS Utility.

Hope that helps

Rick R.

[Jimm]

Hmm it looks like it might be the first and maybe second node.  I will try removing one at a time and see what happens.
Thanks for the help Rick.
Jim

[Jimm]

Well it wasn't the first or second node.  Now I have no idea which one it could be, several are blinking when they are not supposed to.  I guess that I will start isolating it down by cutting the string in half and see which way the problem goes but this could be messy and may take 6 cuts.  I will try Ray and see if he can help.  Anyone have any ideas? 
Thanks,
Jim

[rm357]

Rather than cut, use a piece of wire to extend the data wire from the controller and put a clothing pin on the end. You can make a connection by pushing the pin through the insulation on the data wire at various points in the string... the small hole will most likely close on its own after you are done. Those cuts would take a lot of time to repair...

RM

[Jimm]

Rather than cut, use a piece of wire to extend the data wire from the controller and put a clothing pin on the end. You can make a connection by pushing the pin through the insulation on the data wire at various points in the string... the small hole will most likely close on its own after you are done. Those cuts would take a lot of time to repair...

RM

I got very happy when I saw this post, thought what a great idea but it did not work    The data from the previous node is likely conflicting with the data on the pin.  Any more ideas?  I guess it's back to cutting but I only need to cut the data wire and not the power for the troubleshooting.
Thanks,
Jim     

[tng5737]

what we need is a little device like those little xmas light testers which sense the current in the wire - only this device would pick up the data sig and turn on a little led indicator.  Of course I never got those darn xmas light testers to work reliably anyway!

[chrisatpsu]

what we need is a little device like those little xmas light testers which sense the current in the wire - only this device would pick up the data sig and turn on a little led indicator.

wouldn't that be the node itself?   

[tbone321]

I believe that he is looking for something that could scan the wire for the signal along multiple nodes.  The problem with that is if one of the nodes either pulling the signal line high or shorting it to ground, it wouldn't work.

[tng5737]

wouldn't that be the node itself?   

The more I thought about the tester the more correct this statement became.
So I took a 5" piece of 1/2 pvc pipe and two caps.  In one cap I drilled a hole and mounted a spare node with some epoxy.
I took some of Ray's hookup wire and solered it to the node.   Next I separated the blue wire off of the cable.  The other end of the red
and green wires were connected to a std pc 5 pin molex connector (like they use to connect to the old pata drives.)
I took the free end of the blue wire data line) and soldered it to a pin which was inserted and glued into a small hole drilled in the other cap.
The red and green wire exit the pvc tube from a slot in the side of the pvc.  the caps are placed on the pvc pipe and voila a tester is born.
To use it I simply plug the molex connector into the hub 12v connector and simply insert the pin probe into the data line of the string.
It is a simple way to probe the data line for signal.  I use RJ's config utility as a source of data but could also just use a SSC with the test firmware.
It is not the end-all be-all of testers but I find it useful.

[chrisatpsu]

glad i could help.


even a blind squirrel finds an acorn once in a while   

[tng5737]

I went thru most of my ss devices (strings, rectangles and flex strips) last night abt 40 'strings' in total of varying lengths of nodes.  I had a failure rate of a 12%.   Failure meaning in this case - inability to flash after config.   Almost always it was the first one or two nodes. In one case I went thru 4 bad nodes in the begining before they flashed.  Surprisingly I had a couple of rectangles bad, also the first rectangle in each case.   I did NOT observe any flickering of any of the bad nodes - all the nodes were solid white!   I still have to go thru my squares - I am only hoping that the ones I used in the coroflakes pass the flash test since the flakes are sealed with siicone.
I must say that I was getting pretty bummed out with all of this.  However, I am hoping that once LSP gets its act together with the pixelnet plugin AND I weed out the defective nodes that the situation will improve.   
I would like to thank RJ and crew for all of this.  I have played with alternative methods of driving these nodes - mostly e1.31 - which is nice- BUT, RJ's solution with the SSCs is the ONLY one which allows an economical solution to to getting the controllers next to the strings.  To do this with the other devices costs much more.  The big problem with these strings is that they will not tolerate much of a distance between the controller  and the first node.   In fairness, many of these other solutions are oriented towards drvings megatrees where all the strings are close to a central controller.

[RJ]

I'm glad you are making good head way. I was worried for a little bit. I have been extremely lucky as we have not had any nodes like that. I had a few that did not have the color correct that I replaced when we first got them. and I had a few get damaged by water. But none stopped the string from working.

It is interesting that it has seem to be the first nodes.

RJ

[tng5737]

My theory is that perhaps the damage is done at the factory when testing.  the damage is in some fashion which prevents the rapid on/off . I was able to simulate your flashing in Vixen and got the same results as your config.  I also have some otehr bad nodes as you have mentioned (off or missing color)  however I am not worried about them right now.  I mark the defective nodes with tape and on a subsequent pass I will remove them and retest.  The really interesting part is that you won't find this failure with the TH2010 tester - possibly because it does NOT flash the nodes like the config does.
I'll work thru this - it will just take time and patience!
thanks again