Still have an issue with V3

[lonewolf41]

OK, I have one string of lights that is still giving me the rainbows even with the V3 mod.  I have tried 2 different V1 controllers modified to V3.  I used a 330 ohm resistor on these since that is what they originally had.  Both controllers work fine on another string, so I don't think it is the controllers or mod.  It almost looks like the first node or 2 might be the issue, but am not sure.  I am using xlights test function to test these.

One thing that I have done to see what happens, is I cut the data wire between the first and second node so the rest of the string (84 nodes total) still get power, just no data.  When I re-hook the data wire up when the string is supposed to be on red, the rest of the string turns red.  I disconnect the data wire and the rest of the string stays red and the first node continues cycling through RGBW as it should.  It behaves the same with green and blue; however, when I get the rest of the string on white and then disconnect the data wire, the first node begins to turn all sorts of rainbow colors as it tries to cycle through the RGBW colors. 

Obviously, white draws the most current and it seemed like the string would always start acting up right on or after it turned white.  Does this behavior seem to indicate that the first node is the problem node since it acts up with the data wire disconnected from the rest of the string?  Just want to get some feedback before I start cutting up this string.

Thanks,
-Keith

[tbone321]

Without hacking the string up to badly, I would remove the first few nodes and see what happens.  Remember that each node removes its channel and then regenerates the data string starting at the next channel.  If one of the nodes is bad and not regenerating the string properly or leaking some of the original data along with that it is trying to do, then you will have completely unpredictable issues. 

[lonewolf41]

That is why I thought that the first node being isolated from the rest of the string (data line cut to next node) but still acting up would at least indicate that it has an issue since it is passing data to an empty wire.  My logic sounded good in my head, just wondering if anyone more proficient in circuits, etc. might see if my logic is valid.

Thanks,
-Keith

[tbone321]

If you have the data wire disconnected, then feed the data directly to the second node bypassing the first and see what happens.  That will more than confirm a first node issue.

[lonewolf41]

OK, I jumped the data line over the first node with a jumper wire and the rest of the string seems solid now...acting like it should.  The first node...still getting data from the wire...is still acting up.  Guess that pretty well proves it out.  First node bad.

Thanks for the help,
-Keith

[tbone321]

No problem. at least that seemed to be an easy fix.

[lonewolf41]

Replaced first node, and string is good now.  Just got really concerned as the V3 mod did not work for this string.  Just happened to be the first node.  Isolating it from the rest of the string at least made it pretty easy to track down.

Thanks again,
-Keith

[tbone321]

As RJ said, the ver 3 mod will not correct a defective node.  All it does is push the signal voltage back to where it should be which will correct for added voltage drops due to dirty boards and boarder line nodes.  I would bet that with the VER 3 mod, you could significantly increase the length of the cable between the controller and the first node as well.

[taybrynn]

Last year, first node bad was very common for me.   

[rrowan]

Kinda of strange in a lot of cases the first node is defective.

Wondering if the factory workers pick the worst one first (I am kidding of course) 

or if there is some power up spike messing up the node.

Rick R.

[injury]

maybe it's something like when they plug in the node at the factory to test their controller is already running? Wasn't it assumed pre v3 mod that the reason we were seeing more bad nodes via SSC was that controller Ray used for testing sent a stronger voltage?

[Dennis Cherry]

With all the BAD fist nodes, it seems static electricity (ESD) is killing them. This is a common failure if you do not practice standard ESD safety procedures.

You are not allowed to view links. Register or Login

Some other precautions:
1. Do not work on a carpet floor.
2. Put your RGB Nodes in Anti statics bags until you are ready to connect them to the controller.
3. Do not leave the Nodes laying on carpeting them pick them up.
4. Plastic containers are not static proof unless they say so. Plastic maintains a good charge.

[injury]

Seems if it was ESD I'd get more failures from last nodes than first. Before I ever can use the strand I put the giant plug on it and it's mate on the SSC where as in my case until the lights go outside I haven't bothered to cover the bare cut ends on the other end of the line that rub against the carpet or whatever.

[chrisatpsu]

Personally, with my expierence with the ic of the nodes. They really can take a beating.  ,the only ones I've ever lost were the ones i've left the iron on too long. And even with that, I've only lost two chips. It almost seems like the issues where the nodes are acting up seem to just need touched up with solder. When I went though the qa testing of my smart string project. I had a handful of my nodes that were behaving like bad smart string nodes. Some didn't turn on, some acted flaky, or even affected other nodes in the string. After correcting the soldering of the ic, they now work correctly.

[dpitts]

This year I have been very fortunate. I bought 2500+ TM1809 nodes from Ray. I ran my 1600 node mega-tree earlier this summer for two days and not one was bad. I hope to have same luck with remaining nodes.