Bad dumb nodes

[davron12]

So, some bad news. JonB256 checked the output coming out of the dsc and it looks clean and unlikely to be causing nodes to burn out. I guess that rules out the dsc causing the problem.

As soon as I get a chance, I'm going to try to cut open one of my bad nodes to check the resistors on it and see if they're too small. If that's the cause, it doesn't make sense that the lights don't burn out when plugged directly into the power supply.

The other idea I was wonder was maybe it's the amount of lights we're using causing some spikes from the power supplies. I have 6x100 dumb strings on one 400 watt switching supply. That should be around 18A flashing on and off at the same exact time (a little differently than the smart nodes which control the pwm themselves so they're probably not all in sync).

Other than that, I'm running out ideas.

[caretaker]

The only way to check that would be to put a meter in line with the power supply and measure the inrush current and voltage on a large amount of strings. 

[davron12]

I had already checked with a meter and my power supply doesn't budge, at least as far as my meter can tell. But that doesn't mean there isn't a quick couple millisecond spike.

Still haven't had a chance to cut open a bad node. I had thrown all my bad ones out that I had already spliced out. The only bad nodes I have are outside being used right now. I'll probably have to wait until I take my display down check one of the bad nodes.

[caretaker]

You would almost need a scope or recording meter to measure because the time span would be so short.

[JonB256]

You would almost need a scope or recording meter to measure because the time span would be so short.

My O-Scope showed no sign of huge inrush currents. I know I didn't have exactly the same power supply as davron, but it is his DSC powering my dumb flexible strip. The waveform is as crisp and clean as I've seen on other DC devices using MOSFETs. Just over 400hz waveform with no odd spikes or harmonics.

[caretaker]

JonB256  I was talking more in the line of measuring all 6 of his strings at once with his transformer to see if there was anything abnormal.  My guess is you would find exactly as you did when you scoped his DSC and the issue is the LED's on the strings themselves. I was more thinking out loud as to narrow one more thing off the list.

[davron12]

I have a feeling if it's not the resistors in the nodes being too low, I'm going to end up buying an o-scope myself to track down the problem.

[davron12]

I was finally able to cut apart one of the bad nodes, and unfortunately I don't think there's anything wrong with it. The resistors used are

Blue - 430 ohms (3.4V)
Green - 430 ohms (3.4V)
Red - 560 ohms (0.8V)

All connections look soldered properly, with no cold solder joints, and nothing looks visually wrong or burnt. The only things that look a little peculiar is the selection of the red resistor which limits the led voltage drop to 0.8V, a little low for red leds (typically 1.8-2.2V). So our guesses about the resistor value being too low is wrong. Using a 560 ohm resistor though would use .224 watts, so 1/3 watt resistors should be used. I can't tell what wattage rating the resistor are, and I'm not sure what happens if that's exceeded. Does it change the resistance or just melt the resistor? The resistor isn't melted, shows no signs of damage, and still reads 559 ohms.

[chrisatpsu]

do the dumb nodes use an 8mm led?

[chrisatpsu]

using those resistors in the leds, you can only run 66 nodes on a dsc

[zwiller]

Wow.  I was hoping to at least get 75-100 on a dsc for some elements.  I wonder if the pixel changed?  Maybe the ssc design can be modded/resistor upgrade to get at least 100?  The MR16 is starting to look better for my plans...  At least it makes sense why there were that many failures. 

[davron12]

Using more nodes than 66 wouldn't cause the others to fry, it would just be trying to use more amps than available and cause them to be dim or the colors not to look right.

Chris, yes they use the same led bulbs as our smart strings. Not that I'm doubting your calculations, but I remember measuring the amps used and it was under 4A. I don't have the exact number on me. I wanna say it was around 3.6-3.8A.

I think I see how you came up with your calculations 20mA*3*66=3.96A. But wouldn't that hold true for our smart strings, so they must not be 20mA for each color. Do you happen to have the operating specs for them? I'm guessing they are different than 20mA which totally changes my calculations.

[chrisatpsu]

the 8mm rgbled specs are for 20mA of current.

and the resistors mentioned before land the current to about 20mA per color.

here are a pic of about 100 of them that i ordered.
You are not allowed to view links. Register or Login

and here's the link on Ray's site for them...
You are not allowed to view links. Register or Login

[davron12]

If that's the case, how are we running 128 of the same leds (different circuitry) in under 4A in our smart strings? I'm not challenging you. I'm just puzzled.

Also, if that's true, wouldn't we be under powering our strings and not burning them out, and wouldn't the people using MR16s have the same problem?

I'm missing something.

[rm357]

The current is not the fixed variable, the voltage is.
Assuming:
12 volt power supply
~1.8 volt for the red LED

Current = (12v - 1.8v)/560 ohms

Or about 18mA