Flickering Pixels after an ornament (all of a sudden) - Troubleshooting Steps

[xdandev91x]

Hi all,

First year running a show, so go easy haha.

So show has been running successfully for about 2 weeks - a few hiccups but all resolved through basic user error (i.e. matrix inverted, wire runs too long and needed a data booster etc.).

Tonight's show I noticed that in a line of 3 ornaments, ornament 2 and 3 have started flickering intermittently during the show. There is 2.5m of cable between ornament 1 and 2. It is like the data is semi-corrupted - ornament 2 is a singing tree and besides the white flickering, the tree still 'sings' in between flickers.

So basic troubleshooting undertaken indicates that there is an issue with the output of ornament 1:

• Changed ports on controller - same issue. All other outputs working well. Controller seems fine.
• Power issue? I checked voltage - 12.4V at ornament input - and there is literally a power injection T right before the ornament. Voltage / Power Injection is fine.
• Power Supply supplies power to other lines with no issue. PSU is fine.
• All pixels in the prior ornament are working fine. No dead pixels in ornament 1.
• Removing the prior ornament from the line and replacing with two x100 pixel strings resolves the issue. Issue must be prior ornament.

My question starts here. All pixels in ornament 1 are working perfectly so the data corruption presumably is happening between the last pixel in ornament 1 and the first pixel in ornament 2. However, I've run the following tests and now I am not sure what is going on:

• I ran the Falcon in test mode with 'chase', 'RGB colour wash' and a few others modes and there was no flickering for any of the ornaments on the string. I only see flickering during my xlight sequences.
• I used an RGB tester to test ornament 2 and connected a string of 100 pixels after it. Everything worked also - no flickering and the patterns changed as expected.
• I cannot reproduce the data problem - even using the same extension leads and T joins. It only occurs during my multiple xlights sequences.

Any advice here?

My initial thoughts were to cut off the last pixel and xconnect output and replace them... But I am worried there is something else at play and I did not want to introduce an unnecessary cut unless I tried all other options (especially if it is an xlights issue).

 

[Indigogyre]

Have you added a f-amp between the two props to see if that resolves the flickering?
Have you swapped the props 1 and 2 to see if the flickering follows the prop or goes away? For this test I would expect prop 3 to then have the flickering. I would not worry about rendering for this test as the number of pixels doesn't change but the order where they show up is wrong.
Did you just check or did you replace everything with new wires? Sometimes the flickering is not due to voltage but due to issues with the negative.
What's behind the prop in the wall? Could there be some sort of additional electrical interference going on during the show? A bunch of wires, power outlets, etc.

Just some additional thoughts to try and help.
Good Luck.

 

[xdandev91x]

Have you added a f-amp between the two props to see if that resolves the flickering?

I've run out of f-amps but I will pull one from another string to test this!

Have you swapped the props 1 and 2 to see if the flickering follows the prop or goes away? For this test I would expect prop 3 to then have the flickering. I would not worry about rendering for this test as the number of pixels doesn't change but the order where they show up is wrong.

This is an obvious test I should have done! I am going to try and test this now actually.

Did you just check or did you replace everything with new wires? Sometimes the flickering is not due to voltage but due to issues with the negative.

I haven't replaced anything yet. It is weird that an issue with the negative would happen randomly after 2 weeks of success - but not be replicated when doing the falcon tests and pixel tester. I highly doubt this one is the issue.

What's behind the prop in the wall? Could there be some sort of additional electrical interference going on during the show? A bunch of wires, power outlets, etc.

Interesting consideration. Ornament 1 is on the house pillar which has an electrical circuit behind it for the pillar lights. However, there are other ornaments on the pillar which have no issue, the show ran fine for 2 weeks, and the issue appears to be the data transfer between ornament 1 and 2.

Just some additional thoughts to try and help.
Good Luck.

Thank you for your prompt reply though Indigogyre. I'll be running the f-amp and reordering test as soon as possible!

 

[Kartman]

I have had similar intermittent problems and they usually turn out to be a faulty pixel, I am assuming a bad solder joint somewhere. I usually stand out and watch the prop and when the failure occurs I get a clue about where the last working pixel is and I push that one out. Watch it some more and see if guessed it right, then I replace a few pixels either side of it and that seems to fix it.

Have been down the testing path, with a pixel tester, FPP and the controller and all three methods test out OK and then you have a heap of people watch the show and it gets the flickers, I assume that is the law of maximum inconvenience coming into play.

I do like the idea of swapping props over and seeing if the error moves.

 

[xdandev91x]

Okay, so some more troubleshooting steps done today.

• I moved ornament 1 away from the pillar wall to check for electrical interference. Issue persisted. Not electrical interference.
• Swapped the prop order and anything after what was originally ornament 1 has an issue. Not a controller issue - local to pixels in ornament 1.
• I placed pixels (without the 1.5m xConnect extension wire) between ornament 1 and ornament 2. This resolved the issue.
  • If I re-add the 1.5m xConnect extension wires between ornament 1 and ornament 2, the issue comes back.
  • If I replace the 1.5m xConnect extension wire with another 1.5m xConnect extension wire, the issue persists. It is not the extension cable.
  • If I add an F-amp before the 1.5m xConnect extension wire between ornament 1 and 2, the issue is resolved.
• Conclusion: The data signal is dying from travelling the 1.5m distance between ornament 1 and 2.

Things were fine for 2 weeks and now all of a sudden the 1.5m distance between ornament 1 and ornament 2 is too much? Is this a usual occurrence to have pixel data transmission strength die all of a sudden?

I don't want to waste an F-amp for such a small distance, plus if the last pixel on ornament 1 was previously sending this distance fine, is it on its way out and I should just replace it?
Do I replace the last 3 or so pixels in case the data signal weakened before reaching that last pixel?

 

[xdandev91x]

I have had similar intermittent problems and they usually turn out to be a faulty pixel, I am assuming a bad solder joint somewhere. I usually stand out and watch the prop and when the failure occurs I get a clue about where the last working pixel is and I push that one out. Watch it some more and see if guessed it right, then I replace a few pixels either side of it and that seems to fix it.

Yeah I did the same but it didn't help given that I think the culprit in this scenario is a pixel which is working (lighting up perfectly) but all of a sudden had a decline in sending data.

Have been down the testing path, with a pixel tester, FPP and the controller and all three methods test out OK and then you have a heap of people watch the show and it gets the flickers, I assume that is the law of maximum inconvenience coming into play.

Haha exactly! My blood was boiling when I had someone say "hey is that tree meant to do that?" and then I put it in test mode and everything works fine.

I do like the idea of swapping props over and seeing if the error moves.

Yeah this was a good call. Thanks @Indigogyre! I am now confident that it is an issue with the last or last few pixels in ornament 1.

Just not sure what the best option here is. Replace those pixels or just leave the F-amp. I hate the idea of having to leave an F-amp there permanently though, for such a small distance.

 

[Indigogyre]

From my perspective, which is the lazy one, I would leave the f-amp there for this season and fix the prop after the show. I just hate messing with things once the show is up and running. The more you fiddle the more things can go wrong. Personally a $5 f-amp costs less than an hour of my time trying to fiddle and stress over getting a prop fixed and back up and working. If the prop was totally dead I would fix it.

 

[Gilrock]

When troubleshooting always be open to the fact that something you ruled out might actually be the problem. When signal integrity is borderline it doesn't mean a pixel went bad all because it now reacts different than yesterday. I had a run of icicles that gave me problems for a few years. It would run fine and then sometimes start flickering and showing noise. First thing I tried was adding power injection in-between the two strings. Thought the problem was fixed cause it seemed fine the rest of the night. Next night it comes back. So I add a null pixel halfway down the line that feeds it. Again thought the problem was gone. Sometime later problem shows itself but its not all the time. Kinda seemed like it flickered more in parts where things went white so I still figured it was a power issue. I lowered the brightness a couple times and it seemed to improve but I never got rid of all the flickering. This went on for over 2 years of shows putting up with it. Then one day I read a post where a guy talked about changing the resistor packs in his Falcon. I ordered a bag and changed out all of mine and haven't seen the problem since. So when you have a sketchy data signal a whole lot of things can improve it and make it seem better.

 

[xdandev91x]

From my perspective, which is the lazy one, I would leave the f-amp there for this season and fix the prop after the show. I just hate messing with things once the show is up and running. The more you fiddle the more things can go wrong. Personally a $5 f-amp costs less than an hour of my time trying to fiddle and stress over getting a prop fixed and back up and working. If the prop was totally dead I would fix it.

A great point. Literally doing this. Thank you.

@Gilrock, wow fair enough. I'm definitely not comfortable to unsolder resistors from the falcon Controller though.

 

[merryoncherry]

It really depends on the pixel as to whether it'll send 1.5m, and the type of pixel that is receiving matters too.

A controller is designed to hit any pixel 5-10m away. A pixel is only required to hit the next pixel of the same brand 20-30cm away.

 

[Gilrock]

A great point. Literally doing this. Thank you.

@Gilrock, wow fair enough. I'm definitely not comfortable to unsolder resistors from the falcon Controller though.

Forgot to checkin on this forum since I posted that. I'm not sure if its changed for new board versions but the resistors on my Falcon plugged into a connector built to receive resistor packs. Changing them out was less than a minute. He could have changed resistor values by now so it may not apply.

 

[Skymaster]

I believe the Falcon V4s and V5s ship with 33R resistors for combined power/data runs, rather than the 100R/220R resistors they had previously.
They are soldered directly, only the latch chips are socketed now.

 

[xdandev91x]

Anything requiring soldering of the boards is beyond me. I don't have the confidence that I wouldnt screw it up.
That said, I've stocked up on f-amps for this year

I've posted another q on the forum in an attempt to scale up the show. Bunch of new questions haha.