DiyLightAnimation
Hardware => Lynx Smart String => Topic started by: Pete on August 01, 2011,
-
I am burning in all my SS's and I am finding failing nodes. One string in particular surfaced immediately with problems and leaves me scratching my head. I am looking for suggestions on how to locate the problem nodes.
First I have changed out all the 330 ohm resistors with 150 ohm resistors and I am using the test firmware.
The string is a 128 node string. When power is applied the SS starts out red, green, blue and white and after one or two cycles the string begins to act erratically. As the nodes cycle through the 4 colors a random number of nodes throughout the string turn red, green, blue, orange, yellow, purple and all colors in between at the same time; the nodes are not in color groups but individual nodes - no pattern - all random. As the test cycle continues the entire string will all turn blue, green and white sometimes locking at white for a few cycles. Then it will start cycling again but the color results of all the nodes are random.
I am just going to keep power applied and see what happens.
I think what I am seeing is caused thermally in various nodes. Any ideas on the best way to attack this problem?
Pete
-
Pete, a couple of quick ideas if you don't want to go through the hassle of sending the string back to Ray. One check the nodes out and see if you have any that seem warmer than the rest (a pixel that is using more power than the rest will have a higher temp). Two if you see a node act erratically usually the node right before that one is bad. Three (this is a pain but it is how I fixed one of troubled strings) cut the string in thirds and try the first third and if that works correctly try the second and so on. When you have 128 nodes it is real hard to track down the bad nodes, making it smaller helps in pin pointing the bad nodes. Another tip is I would put tape or a small wire tie and bad nodes where one color is dead or if the node acted weird so I could go and deal with them one at a time.
-
Since the whole string is acting weird, check the first node. I have had few strings with the first node bad. If the first node is sending bad data then the whole string is gonna act weird.
Try lightly twisting the first node back and forth before you cut it out and see if it will act right or make producible problems. I have found a few bad ones that way.
Brian
-
I would also check to make sure that you are supplying enough power to the string. Low voltage can also cause erratic operations. If the supply you are using can't supply enough current then there will be a significant voltage drop on the string. I would put a volt meter on the other end of the string and see where you are.
-
Thanks for your suggestions.
Voltage is good - using a 40 amp source and other 128 node strings have no problems.
I started the heavy handed method (twisting, bending and pulling on the wires) on each node and found the first three nodes all would change how the string operated, but only when they started to get warm. I cut them out and applied power to the rest and found two more nodes acting the same way. Cut those out and the rest of the string(s) (now in three pieces) work just fine.
All in all 5 nodes; 3 at the front and two more in different parts of the string were causing the problems.
Very time consuming to locate especially on the larger strings.
Thanks for your help!!
Pete
-
quality is inconsistent
I spliced my first 128 nodes about 50 places, its gotten better since then
-
Pete,
Did the issues you reported at the top of this thread go away after you found and replaced some bad nodes?
I'm seeing very similar results with a bunch of 128 node strings I just received. The strange part is that if I drive the strings with the TH2010 tester from Ray, they behave perfectly, so there is some variation between the tester and the SSC. And, yes, I have already done the resistor change on the SSC. Still wonder if some of these strings need even a lower value for that resistor? The random colored portions of the string always seem to occur near the end.
-Mike
-
I have put this string on the back burner and have not worked on it since my first post. I decided to attack simpler problems with the smaller strings first and then tackle this one. When I used the tester from Ray the string worked with no problems, in fact most of my strings would work with the tester. The problems with single nodes and strings not working did not appear until I used the SSC.
-
This may seem too obvious but check to see if the pigtail on the output end does not have any wires shorted together. That can cause a string to act really strange. It's easy to short them together and not notice it. I started putting tape on all the pigtail wires until I started cutting and splicing them. I got in a hurry to test the strings and thought I had separated the pigtails but missed one. It will not only cause strange behavior but it also caused me to loose a PIC. Ouch.
-
This may seem too obvious but check to see if the pigtail on the output end does not have any wires shorted together. That can cause a string to act really strange. It's easy to short them together and not notice it. I started putting tape on all the pigtail wires until I started cutting and splicing them. I got in a hurry to test the strings and thought I had separated the pigtails but missed one. It will not only cause strange behavior but it also caused me to loose a PIC. Ouch.
Nope, no shorting on the end of the wires. I always snip off the end so there is no bare wire extending beyond the insulation. Also, it would not explain why the string behaves perfectly with Ray's controller but exhibit issues as soon as it's moved to an SS controller.
-Mike
-
Someone had this and it was a pixel causing it in the string. once they located it and removed it the rest of the string was fine. You know it will be a node at or prior to the first one acting up. But then it becomes hard to locate.
RJ
-
Is it allways the same SSC controller or have you tried different SSCs on the string? If not then I suggest you reflash the PIC just to be sure. Sometimes you think it is sprogrammed but for lack of a better term it slps a bit. It will seem to work but the data out might not be correct.
You might also try injecting power into both ends of the string. That is hooking the same power leads to both ends of the string.
-
Tried multiple SSC's as well as injecting power from both ends, still get the issue. I'm going to start cutting up the string(s) to see if I can track down a bad node.
The one thing that really gnaws at me is that if simply switch the connector over to Ray's TH2010 controller and have it running a similar pattern as the SSC test, it behaves 100% correctly. Not one node misbehaves. Switch back to the SSC you get random colors near the end of the string. Clearly there is some difference in how the two controllers (TH2010 and SSC) are driving the pixel strings. I don't know if it's a voltage/current difference or a data/timing difference, but something's different.
-Mike
-
I would start with the old 50/50 rule. Split one of the strings into 2 equal lengths. If one string works fine the problem exsist in the other string. Cut the bad string in half and repeat. Using that method you won't have to resplice as many connections. I'm just saying...
-
I would Not do the 50/50 thing.
Cut after the last node that is correct and test remaining string - it is usually either a bad input or bad output on the node.
A bad out on the last node of a string won't hurt anything.
Bad inputs have to be removed.
RM
-
The biggest problem I'm having with this approach is that there is no consistent spot where the nodes start to display incorrectly. Each time through the red-green-blue-white sequence of the test firmware, will yield slightly different spots where the colors are incorrect. So if I pick a spot that looks like it's bad, next time through it might be bad ahead of that or after that.
I'm really starting to get a bad feeling about the batch of strings I just received from Ray.
-Mike
-
Hi Mike
Have you tried setting the "program" jumper on so the lights go white only? You might see where the trouble maker is
Rick R.
-
Hi Mike
Have you tried setting the "program" jumper on so the lights go white only? You might see where the trouble maker is
Rick R.
Yeah, I've been down that road. Actually all white is usually the only time all the pixels are correct. The random colors only show up when it's changing from red to green to blue. The beginning of the string will be fine. Somewhere from half to near the end it will show the random color issue (but vary from change to change).
-Mike
-
Start with the first bad node - it may work with some colors and not others.
As long as there is a bad node before it, the behavior of the other nodes behind it are influenced by the bad one earlier in the string.
For example, if node 5 is bad, node 7 acting up intermittently does not tell you anything except to confirm that an earlier node is bad.
RM