One sequence causing GFCI circuit to blow after rain

[jtomason]

OK, this is a weird one. 

I have 6 AC controllers, 1 AlphaPix, 1 Pixie 4, 1 50W flood, and a CMB24D in my show.  It's been running flawlessly until yesterday.  Two days ago we had a substantial rainfall.  Starting last night, when ONE particular part of one particular sequence plays, the GFCI outlet controlling the outlets that all of the controllers are plugged into trips.  Half the controllers are in one outlet, and the other half is in another outlet on that circuit.   Total max amp draw is 5a, so that's not the issue.  I've isolated it down to one unit (at the far end of the RS-485 chain) and down to a particular snippet of a sequence - specifically, 2:18 into "Let It Go".  Every time I hit that section it trips.  So far I've established that:

 

1. If the RJ-45 is unplugged, the sequence otherwise plays fine for the other controllers. 

2. If I remove from 2:16 to 2:18 for those 16 channels in the SE and play it (with the RJ-45 inserted), it plays fine.

3. Every other sequence plays fine.

4. Using the HW utility and turning all channels on that unit on to steady 100% works fine.  All of the outlets and plugs are dry. 

5. Copying the same sequence data from other channels and pasting it over the existing channel data doesn't fix it.

6. If I plug that half of the controllers into another circuit, everything works fine. 

 

So this one is pretty weird.  Anyone have any ideas as to why it might be happening?

 

 

Edited December 17, 2018 by jtomason

[TheDucks]

3 minutes ago, jtomason said:

OK, this is a weird one. 

I have 6 AC controllers, 1 AlphaPix, 1 Pixie 4, 1 50W flood, and a CMB24D in my show.  It's been running flawlessly until yesterday.  Two days ago we had a substantial rainfall.  Starting last night, when ONE particular part of one particular sequence plays, the GFCI outlet controlling the outlets that all of the controllers are plugged into trips.  Half the controllers are in one outlet, and the other half is in another outlet on that circuit.   Total max amp draw is 5a, so that's not the issue.  I've isolated it down to one unit (at the far end of the RS-485 chain) and down to a particular snippet of a sequence - specifically, 2:17 into "Let It Go".  Every time I hit that section it trips.  So far I've established that:

 

1. If the RJ-45 is unplugged, the sequence otherwise plays fine for the other controllers. 

2. If I remove from 2:16 to 2:18 for those 16 channels in the SE and play it (with the RJ-45 inserted), it plays fine.

3. Every other sequence plays fine.

4. Using the HW utility and turning all channels on that unit on to steady 100% works fine.  All of the outlets and plugs are dry. 

5. Copying the same sequence data from other channels and pasting it over the existing channel data doesn't fix it.

6. If I plug that half of the controllers into another circuit, everything works fine. 

 

So this one is pretty weird.  Anyone have any ideas as to why it might be happening?

 

 

You may be dumping EMI into the (Green) ground due to fast switching in that sequence. The odd thing, is GFCI's look at the precise Balance of Hot to Neutral. Not Ground. so maybe the Triacs insulator is breaking down only under switching.

Try the HU Twinkle on that controller (that does ON/Off), and Shimmer (more of an Up/Dn)

[jtomason]

9 minutes ago, TheDucks said:

You may be dumping EMI into the (Green) ground due to fast switching in that sequence. The odd thing, is GFCI's look at the precise Balance of Hot to Neutral. Not Ground. so maybe the Triacs insulator is breaking down only under switching.

Try the HU Twinkle on that controller (that does ON/Off), and Shimmer (more of an Up/Dn)

I did all three modes (steady/shimmer/twinkle).  No issue.  And those channels get exercised QUITE a bit; so if it was a fast switching problem, it'd be happening a LOT.   

Moreover, this particular part of the sequence is a few slow ramp downs (see the attachment).  Happens right at about 2:18, which is "I stand, and" from the part "here I stand, and here I'll stay" - so right after "stand".   The channels in question are House 1-16.  

 

 

 

Edited December 17, 2018 by jtomason

[TheDucks]

Mostly dimmed generates the most line 'noise'.( I remember have a 1970's dimmer and hearing about it from my dad, a 'Ham' radio operator)  The ON time is relatively short. and you have at least 45 doing it at the same time

[jtomason]

1 hour ago, TheDucks said:

Mostly dimmed generates the most line 'noise'.( I remember have a 1970's dimmer and hearing about it from my dad, a 'Ham' radio operator)  The ON time is relatively short. and you have at least 45 doing it at the same time

Granted, but it has been running in that configuration for weeks now - and all of last season. 

I’m going to replace the GFCI outlet with another and see if that helps - the outlet does look old, and has tripped a lot in the past. 

 

[jtomason]

So, the GFCI outlet was replaced but didn't solve the issue - now it's tripping at different parts of the show.     So, I'm running on two circuits for the time being. 

Anyone have any ideas?

 

[jtomason]

OK..  So... Yeah... I fixed it.

It had nothing to do with the sequence, and everything to do with the one extension cord that was filled with water.  

So, today I went through everything again, elevated the connections I forgot, replaced that extension cord with an SPT-1 cord, and tested for 2 hours this afternoon.  All went well, and the show has been running flawlessly all night.  Whew!

Dielectric grease, outlet plug covers, and a ground leakage detector have been ordered and will be here Thursday - just in time to see if anything is amiss after the rain we are expecting tomorrow night. 

 

[TheDucks]

2 minutes ago, jtomason said:

OK..  So... Yeah... I fixed it.

It had nothing to do with the sequence, and everything to do with the one extension cord that was filled with water.  

So, today I went through everything again, elevated the connections I forgot, replaced that extension cord with an SPT-1 cord, and tested for 2 hours this afternoon.  All went well, and the show has been running flawlessly all night.  Whew!

Dielectric grease, outlet plug covers, and a ground leakage detector have been ordered and will be here Thursday - just in time to see if anything is amiss after the rain we are expecting tomorrow night. 

 

TNX for the updated cause.  Sneaky H₂O

[Richard Hamilton]

12 minutes ago, jtomason said:

Dielectric grease

 

Glad you got it working.   Dielectric grease?   That's a new one on me.  For 20 years, I've used little sandwich bags.

[jtomason]

2 minutes ago, Richard Hamilton said:

Glad you got it working.   Dielectric grease?   That's a new one on me.  For 20 years, I've used little sandwich bags.

Yep, apparently it's helpful for keeping moisture from getting in the pass-through side of a light strand's plug.  I plan to put some in and put a cover on, plus keeping them elevated.   I don't mind a little over engineering to save my sanity. 

[~DOC~]

Oh snap I miss understood die electric grease. Just referenced it again and found it to be an insulator. Good post. I was going to use it on my SPT 1 wire and then got into too much of a hurry. It will be done for next year. My fix is already working now.