ELK - multiple problems after storm/lightening

shenandoah75

Active Member
We had a storm today and after a lightening strike, my shop overhead door closed. I thought this really odd (and in fact unsafe - given other devices i have on relays) and i'm pretty sure i'd never seen elk activate relay like that before on startup. My wife called a few minutes later to note a sprinkler zone was on.

Long story short so far i've noticed a few things over last few hours (adding as i go):

1) i have two relay boards and one relay on each is stuck on. I've tried to cycle on the elk (which shows the output is off) to no avail. Also cycled the M1G twice, both relays remain on up[on boot up. a) curious as to what would cause it B) are the boards bad / do i have any options (one if fully used, the other might have a free relay)?

on second thought - i guess it could be the M1XOVs' outputs themselves are locked and not the relay. i can try to test that tomorrow...


2) phoneline seized LED is on and i can't access that line (the other line works fine).

seems to be same as: http://www.cocoontech.com/forums/index.php?showtopic=17066
though i haven't troubleshooted as much yet...


3) i no longer have control of the HAI thermostats. I have two. The serial expander seems to be flashing normal (i think) but no temp/setting/status at all seen in elk. Changing the setting at the thermostats is not forcing recognition by Elk either...

4) may or may not be an issue, but one of my motions has an issue (always violated with no motion) I'll test it tomorrow...


thx in advance
-brad
 
Unfortunately it sounds like you had a very close strike. In this situation the magnetic pulse from the lightning actually travels through the air and inducts current into nearby metal objects. The small, long wires used to run between automation components are particularly vulnerable to picking up this stray current.

The relays are likely damaged permanently; their coils melted in the on position or snapped so the conductor is forever stuck in the on position.

It is possible that some of the other functions of your system were also permanently damaged. It happens with this kind of thing. I had a very near strike a few years ago that severely confused some equipment (had to unplug/reset some of it for a while to make it happy) but did indeed kill a few things. Lost a gigabit switch, several ethernet cards and one random door sensor. It was just before i got my main elk panel which i now absolutely adore so I'm fearing the next one :S

Did you loose any other appliances or such in your house? Its pretty common to have sensors, zones and network equipment, even serial stuff like tstats, get damaged by relays are higher current by a decent bit. You may want to check your whole home over pretty good here.
 
If you can determine that the relays are fried and the rest of the board is OK, you can solder in new relays.

http://www.hascorelays.com/electro_klt_series.asp

http://cgi.ebay.com/12V-COIL-SPDT-RELAYS-12A-120-VAC-Hasco-KLT1C12DC12-x4-/180652051410



These are the relays Elk uses. A lot of 4 on ebay for $9 including shipping. Probably worth it to just buy them if you ask me.

edit:
ooops, just noticed the ebay item ended. But, it looks like he had quite a few and they didn't sell. Maybe an email to the seller will get you another chance.
 
It looks like onlinecomponents.com has them in stock, at $1.61 qty 1:

http://www.onlinecomponents.com/hasco-components_klt1c12dc12.html?p=11507275

I will caution that I haven't done business with onlinecomponents.com, so I don't know how they are. Mouser and Digikey probably have substitutes, but it would take a bit of research.
 
Also, I'm not sure which HAI thermostats you have, but some of them seem to be a bit sensitive to lightning. I have three RC1000s, and after my lightning hit last year, two of the three were no longer able to communicate on the bus.
 
Neurorad, unfortunately, no - and reason was i had it in a UPS, but when power went out i couldn't tell unless it was long enough for the UPS to die. I should have used the non-battery backup, but i guess i wasn't thinking and hindsight is 20/20....

-> Update: actually I do have whole house suppression at the meter (i know it's not ideal)

-> Correction - i forgot. I used the leviton panels and used their surge suppressing outlets (blue), so ya it was protected.

Some updates, after seeing this thread: http://www.cocoontech.com/forums/index.php?showtopic=19731

Some Updates:

1) i realized that HAI support had moved to a proprietary firmware. I updated the M1XSP to that version and set the jumpers accordingly (only one that had been different was baud), but still seems to be no recognition...

2) The Relays are ok... the RED LED was lit so i suspected it's the M1XOV. Sure enough, i remove the ribbon connector and the relay goes off indicating that the M1XOV is the one with an output locked on...

3) Even disconnecting the TLINE connector, the M1 still seizes phoneline upon reboot... (diagnostic menu showed 0 volts before disconnecting). I should also note the phone line is voip via At&t uverse which has it's own UPS/protection on the gateway.

4) Haven't checked the motion yet...


Also worth noting: i do have the databus running between my house and outbuilding. It's via CAT6 buried underground in 2.5" conduit. Have one M1XOV house-side with one of the impacted outputs and the M1XOVR outbuilding-side with the other. Both thermostats and the M1XSP are house-side as is the M1G. A M1XIN is outbuilding-side for the motion sensor in question. Seems like it was selective in what it hit, but unfortunately 3 cards, the phone line side of the M1 and either the HAIs or the MX1XP as well... Digging through posts trying to figure out how i can debug the HAIs... (probably the wife's favorite feature next to lighting control).


Also here is the log around the time in question:

23,Sat 8/13/2011,17:14,,1161 = EXPANSION MODULE RESTORE,Output Exp. 3
24,Sat 8/13/2011,17:13,,1367 = SYSTEM START UP,Input Exp. 4
25,Sat 8/13/2011,17:13,,1141 = EXPANSION MODULE TROUBLE,Output Exp. 3
26,Sat 8/13/2011,17:13,1,1143 = TELEPHONE LINE FAULT TROUBLE,
27,Sat 8/13/2011,17:13,,1367 = SYSTEM START UP,Keypad 2
28,Sat 8/13/2011,17:13,1,1357 = EVENT LOG, 80% FULL,
29,Sat 8/13/2011,17:13,,1367 = SYSTEM START UP,Control
30,Sat 8/13/2011,17:13,,1367 = SYSTEM START UP,Control
31,Sat 8/13/2011,17:13,,1367 = SYSTEM START UP,Control


thx
-brad
 
Are ANY components plugged into ANY other 110 outlets, e.g. In the outbuilding?

My plan for the future is exactly as you have it - whole-house surge protection at the meter, plus the Leviton surge protected outlet at the enclosure.
 
Yep, but all are surge protected:
(1) 24VAC wall wart powering the M1
(1) M1XEP power supply
(2) 24VAC load side for all the relays on the M1RB (one wart in out building, one in the house).

which leaves the Rj45 jack on the M1, the HAIs, and the Leviton RZCOP (connected to another M1XSP) as only other possible powerline entry points (if i'm thinking that through right). All my networked devices should be on UPSs - the RZCOP was not at the time, but it's still working fine.
 
What are the chances you have more than 1 ground and/or ground wires, among all the components? Is there 1 central ground that all buildings share? Or is there more than one 'service entrance'?

I don't pretend to know too much about surges, but if there are multiple grounds and they are at different potentials, I think it's a setup for a surge to propagate using your cables.
 
What are the chances you have more than 1 ground and/or ground wires, among all the components? Is there 1 central ground that all buildings share? Or is there more than one 'service entrance'?

I don't pretend to know too much about surges, but if there are multiple grounds and they are at different potentials, I think it's a setup for a surge to propagate using your cables.

Having multiple grounds will hurt you when a potential is created in the earth around your house. But a potential can be created by lightening in the wires that are already in your house, and that is very hard to block. The longer the wire and the less shielding it has, the higher the potential for potential. (clever use of words, ehh) :lol:
 
Ya the main panel and the subpanel at the outbuilding have there own ground bars (though they are grounded together as well back in the main panel). However, neutral is not bonded to ground in the subpanel, only the main...

Still hoping Spanky chimes in with a miracle (esp on the HVAC) :)

Next step is to reconfigure the other M1XSP (lighting) to HAI and see if it works. If not, i guess it's two new thermostats?

-brad
 
Ya the main panel and the subpanel at the outbuilding have there own ground bars (though they are grounded together as well back in the main panel). However, neutral is not bonded to ground in the subpanel, only the main...

Still hoping Spanky chimes in with a miracle (esp on the HVAC) :)

Next step is to reconfigure the other M1XSP (lighting) to HAI and see if it works. If not, i guess it's two new thermostats?

-brad

I don't have a phd in lightening physics, but I am pretty sure you don't want to grounding rods bonded together. You will get current from one to the other and possibly through expensive electronic gizmos on the way. If you have two grounding rods, you should have the electrical systems attached to each isolated. The only time you can have more than one grounding round is when you have tons of them completeling surrounding a structure very close together effectively behaving as one giant rod.
 
I bet there is some way to physically disconnect the LV cables between buildings - something like an 'opto-isolator' to limit the propagation of surges.

Edit - would be great to hear some specific recommendations from some people.
 
Back
Top