Bad dumb nodes

[davron12]

I am also seeing additional failures.  The count continues to rise on a daily basis.  Still haven't had the opportunity to check out the new strings, but looks like I am going to have a bunch of splicing to do this weekend

Your count is still rising? That's not promising. You started out with 5-6 bad nodes per string. What are you up to now?

I'm wondering if it's related to the DSC since that's new this year. Red is the most delicate color (requiring the lowest voltage drop). Is it possible that the DSC is somehow creating an inductor-like surge that's making the voltage > 12v. I don't have a way to check this. Just doesn't make sense how I had some of my lights running for 5+ hours without a problem and then within minutes of connecting them to the DSC the reds started blowing.

[JonB256]

It is more likely that the current limiting resistor value used for the Red LED is low. If enough people complain to Ray, he will get the manufacturer to bump up the resistance a little.

If you are using an adjustable power supply rather than a non-adjustable ATX supply, you could dial it down to 11.5 volts. The SSC or DSC wouldn't care (they have their own voltage regulators).

[davron12]

That was one of my initial thoughts too, but wouldn't it have burned out while having it plugged directly into the power supply for hours? The reds only seem to burn out after being plugged into the DSC then into the same power supply. I do have my power supplies set exactly to 12.00. I'll try turning them down to 11.5 and see if any more burn out.

[chrisatpsu]

do you have a link to the ones you bought?

[kjam22]

I bought the ones listed in the wiki, 100 node count.  I am seeing almost an 8 percent failure rate - mostly red.  I am going to swap out the bad nodes this weekend and see if I continue to get failures.

[JonB256]

I have two kinds of Dumb RGB in the display.

I have flex strips that I got from China, not from Ray. They are cut up and mounted around 2 window frames. They are connected (this year, anyway) to a 16 channel DC controller. Zero failures on these.

Then I have four strings of RGB dumb pixels mounted on a 4 layer mega-star. These were from Ray (indirectly, though - bought from someone who bought from Ray). There are 200 pixels total, and other than one string not having the same "white" appearance when All On, there are no failures. This is also on a DC controllers (not DSC). This one is using a power supply that was originally sold as 24vdc but easily adjusted down to 12vdc. If it fails at some point due to not being at its design voltage, I'll power from a SS Hub and ATX power.

My failures have only been in the Metal Square style SStrings. I bought 280 (14 strings of 20) and have had to cut/splice 8 pixels.

[davron12]

Mine are the same ones from the wiki too.

JonB256 - Do you happen to have a DSC and be willing to try potentially burning some nodes? I have no more dumb strings left that don't have bad nodes to try testing with a different 4 channel DC controller.

BTW, I turned the voltage down to 11.5, and it looks like I've still lost a another couple nodes. Can't tell for sure since I haven't had a chance to cover up all the bad ones yet, but it looks like there's a couple more out. So don't rely 100% on this observation.

Can someone with a scope check the output from a DSC plugged into a set of dumb strings and see if there's any spikes or other anomalies or is anyone using the DSC with dumb strings and not having any problems?

[JonB256]

I have a scope but no DSCs.  Seems like that is about the only thing from DLA that I don't have.

Well, no SSRs or stuff that pre-dates the LE version 5.

If you have a DSC (kit or built) I'd be willing to test. Now that my show is pretty much in auto, I'll have a chance.

[tbone321]

Since all the DSC seems to be doing is switching on a mosfet, I really doubt that it's the DSC.  When this is happening, are you switching the LED's on and off a lot?  I wonder if there is an issue with the nods themselves?

[davron12]

I'm actually not switching them on and off at all. I'm leaving them at around 60%-80% brightness white, but since it's pwm, they're flashing on and off thousands of times per second. Look at it from this perspective and you'll have to agree it's the dsc - plugged directly into a 12v power supply for hours without any problems, then plugged into a DSC using that same 12v power supply and some reds (and one or two blues) flicker then burn out within minutes. That really sounds like the DSC, or the DSC in conjunction with the strings. I do realize there's nothing special about the DSC that could cause this to happen or that makes it different than any other pwm controller, but it's hard to ignore what's happening and blame it on bad nodes. If JonB256 had this problem with a different controller, then I'd blame the design of the strings.

JonB256, pm me your address and I'll send you a dsc to test. Thanks in advance for your help.

BTW, I'm up to around 9 bulbs per string out now. I just checked the voltage coming out with a multimeter and I'm getting 10.68v with my power supplies set to 11.5v using a 2ft cat5e cord. One might think that it's too low of a voltage, but if I plug it back into a power supply directly, the bad nodes don't come back on. And too low of a voltage wouldn't burn out a resistor and an led. I'm utterly confused now.

[kjam22]

Any one get anything figured out with the dumb nodes?  I spent hours on Friday cutting and soldering bad nodes just to have many of the spliced nodes fail.  For some reason they seem to mostly burn out near the beginning of the string.

[rm357]

Are you sure you have 12 volt dumb strings?

All of the LEDs are really below about 2.3v and use a resistor to control the current. 12 volt ones have segments that are usually in groups of 3 LEDs.

As far as the cycling burning out the LEDs, not a chance. Over-current and heat are the two things that will destroy an LED. If you are getting a voltage spike or static electricity, that could cause the over-current, but at the rate it sounds like you are having issues, I would almost be willing to bet that you have 5 or 6 volt nodes...

Try just connecting to a 5 volt supply and see if you get a good white...

[kjam22]

I bought the nodes that are linked through the wiki, the should be 12v.  I am having failures ith two different power supplies as well.  One is a pc power supply the other is from ray.  The failure rate seems to be the same between the 2

[davron12]

rm357 - I also bought the ones linked through the wiki. Please note that I had no problems with the strings plugged directly into the power supply. Problems only (immediately) happened when plugged into the DSC, plugged into a hub, plugged into the same power supply.

kjam22 - don't bother soldering. Get some black vinyl tubing from the plumbing or pond section at lowes/home depot. I don't remember the size I got, but get one that fits snug over the lights. Then, put a piece of gorilla tape on one end. You're making black cap that you can use to cover up the bad nodes. Use a heat gun or hair drier to heat that up and they slip easily over a bad node. It's the quickest easiest way I've found. I'm not going to bother fixing my strings until we find out the problem.

I send a dsc to JonB256 last week. He's going to do some testing with his scope to see if anything crazy is going on.

[JonB256]

I send a dsc to JonB256 last week. He's going to do some testing with his scope to see if anything crazy is going on.

and I got it but haven't started testing yet. I have some rectangular Dumb Strings and some Dumb Flex Strips plus some 12vdc halogen lights for comparison.