DSC Not Working [SOLVED! And could be a prob for you one day.]

[lboucher]

Hi All

Just got done working with 14 DSC's.
Went through and programmed them all, and did the channel assignment.
But for some reason, 1 just won't work right.
When I try to use the Smart STring Utility, I hit transmit, and for a brief moment, the LEDs look like they are starting the first blink of the confirmatoin blinky, then everything goes out.
I have to unplug the RJ45 and hit transmit again, to get the same flash, otherwise they stay dark.

I have confirmed the 2 chips work right, by swapping to another controller.
Interestingly, I did confirm the chip is taking the channel assignment correctly, it works fine in another controller.

I'm kinda out of thoughts.
Chips work right.
Its obvious the LEDs can get power, however just for a brief second, then go dead.

I doubt its the mosfets, If one were bad, it would only effect one color. 3 cannot be bad.
Its almost like the current draw is somehow resetting the microcontroller.

Any thoughts appreciated.

[tbone321]

I don't use dumb strings but isn't that what it's supposed to do.  IIRC, after they flash did you power down the DSC and move the jumper from program to run and power it back up again?

[lboucher]

After programming the channel its supposed to keep flashing until it is powered off. Mine just briefly flash once for half a second.

Then I unplug, move jumper, replug in and try to use them, but it doesn't work.

[drlucas]

Yeah, they should keep flashing....have you tried a different string of nodes at all since you hit the bad DSC? Perhaps there is a short somewhere in the string. try that string now on a different controller and see if the problem follows the string. Only other crazy thought is to restart computer...maybe another application is also sending e1.31 multicast traffic and is stuck running. I sometimes jump from xlights to test and there is conflicting traffic on the network because I don't hit stop before I play with the smartstring utility. Likely rebooting won't help, nor will moving to a new string of nodes, but might eliminate something silly like a short in the wires.

[RJ]

Are you sure the pore supply can handle the load. Once the ssc or dsc is programmed and begins to blink no outside data will have any effect. It locks itself into blink and ignoring anything. It sounds ax if when you lights come on at the end the supply can handle the load and the dsc is resetting. Try just a few nodes to test this.

RJ

[lboucher]

I oné hundred percent agree the controller is resetting.
Power supply is resetting rock solid 25 amp ham radio beast i inherited, and this is the only string on it.
Samé string and supply work with my dožen or o other dsc.
Starting to wonder if i have a bad connection on the power on the ethernet jack...

[jnealand]

I'm wondering if the wires from the DSC to the string or 4 pin connectors could be wrong.  Some strings seem to use different color codes and at least I get confused by what the correct wiring should be.

[rm357]

My next step would be to go through looking for cold solder joints... You've already eliminated most of the possibilities.

You might have a marginal voltage regulator that is dropping out...

[lboucher]

[SOLVED]

Finally, figured this out.

So next thing I do is try the test software.
That produces a kinda odd result, in that green and blue would seem to flash, like the test pattern, but very low brightness.
Red was not seen at all.

So I start into it, and first thing I do is solder in a secondary 12V hookup directly to the board, so I know thats not an issue.
Then I start probing (with a mega old oscilloscope I inherited and had never used.) to look at the voltage regulators.
They looked good, but I thought I saw something on the 3.3 Volt a couple times, but I think that was just my hand slipping.

At some point during this I noticed, that they stopped flashing the test pattern. (Only 1 slip with the probe that cause a spark, but I didn't think I actually hurt anything)
So i swap out the chip with another with test code, same result, and the original chip still worked.

So at this point the 3.3 V still looked suspicious, so I ripped it off, and wired a direct 3.3 volts.
Nothing....
Now that I am hooked up to a regulated supply, I can see the current draw from the micro.
I note that it goes in a disinct 4 beat patter, with the fourth pulse being a 20% ish higher than the others.
This causes me to really wonder about the MOSFETS.
So, I did more probing, but didn't really seem to see anything.
1 of them seem to have a slightly odder input voltage transition, like high then brief dip low then back to high. Instead of just a half second high.
So at this point, I just say the heck with it and cut off the MOSFETS one at a time....
 Nothin... (Off course nothing would work, but at no point did any color work)..
So at this point, I just start applying voltage to the point at where the wires leave the board and go to the strand.
For the life of me I cannot get the lights to light.

So i cut off the 4 wire white socket so I can easily probe it.
AND WOULDN'T YA KNOW

The manufactured white socket had the 4 wires messed up.
As you look at the keyed end with the nob up, the wires were mirrored horizontally.

Now how in the world does having them mis-wired like that cause blue and green to just barely flash.
Only thing I can figure is the caps on the small boards in each led bulb were somehow storing just enough current to flash the LEDs a bit? If you have a circuit diagram that could explain to me how this mis-wire results in a slight flash of the green/blue LEDs I would love to see the explanation.

Long night of being Obsessive Compulsive
Included for fun is a pic of my workbench when all was said and done, and I stepped back and said, what am i doing.
I know I know, the bench is a mess. I let it go way downhill 1.5 years ago when we were trying to move, and I just haven't had the time since to get it back the way I want it.

[tbone321]

That scope is rather modern compared to my old Textronics which was "state of the art" in its day.   

[jnealand]

Look at reply #6 above.  This issue has come before with the connectors.  Not a lot, but it has come up before.

[lboucher]

Yes i did note that reply. Wish i had paid more attention in retrospect.
When i first saw it i thought you were referring to color codes of different types of equipment.
Like some have RGBY and others RBlackBlueY.
Never would have thoughy same item, from same supplier in same order woild be wrong.

Still clueless how those leds ever lit at all, even if it was momentary and dim.