Lightening

Installed a system a few years back for good friends in FNQ. Two dimmers, two relays, four shutter controllers and a probably 20 DLTs. Last night during a huge electrical storm they had a hit they estimated lobbed about 100 metres away. Followed by a number of others close by.

After the hit nearby the system stopped working. Just spent an hour on the phone going through a checklist.

One relay and the 4 shutter controllers have no C-Bus LED light. No clock present. (2 dimmers and 1 relay display C-Bus LED OK)
All the DLT's are powered and display but are not functioning or controlling any load.
Toolkit only sees the CNI on the the Ethernet side and when scanning the C-Bus network sees no units.
The two units that I enabled C-Bus clocks on are still showing C-Bus Ok. (The CNI and one of the dimmer packs)
All the loads can be controlled at a local level on the output units.

There was the best part of 300m of C-Bus cable installed in this large house and the most likely cause of the issue up there. A huge C-Bus antenna and the ionisation in the air of the nearby strike would I suspect would have injected significant spike into the C-Bus side of the system.

Trying to fault find from 3800km away is a challenge.

Anyone's guess how many volts appeared on the C-Bus cable at the height of the lightening hit nearby but suspect it would be with nearly 300 metres of cable significant. I suspect and hope that the ESD held up and protected some, however I fear the induced voltage could have fried a good number of units if not all on the C-Bus buss side?

Anyone else had lightening near hit experiences with systems?

Regards,

Fleetz

 

A couple years ago, a house with ~150 units across two networks had a very close lightning strike. The lightning strike randomly killed a handful of units but the vast majority was fine. Damaged units were a CTC, one of the high current dimmers that look like a car amplifier, a bus coupler and a shutter relay. All of the items stopped responding to the C-Bus network. I don't know whether it was just the transformer going open circuit or whether more serious damage had occurred.

Thankfully the rest of the network survived and doesn't appear to have been damaged by whatever killed those units.

 

I've wondered about this, won't something like below help even though it's for Ethernet? Protection is on all 8 legs and the voltage ratings are above the Cbus spec.

https://www.dehn-international.com/... Data Networks and Ethernet Applications.html

 

Thanks Tasman for the feedback.

The units that died in your situation were you able to still scan the C-Bus network and locate the units that were toast?

How were you able to determine that your CTC, high current dimmer, shutter relay and bus coupler has expired?

Hopefully in this situation there is only a unit or two causing the bus not to respond. The 5 units that are not displaying C-Bus green LED may be not receiving the clock from up stream devices or they maybe the source of the problem?

The hope is a unit (or maybe two) is corrupting or killing the C-Bus bus which is why TK can not see any on the network.

Really need to start isolating the system into smaller sections.

Fleetz

 

I have a question.
Is diagnostic utility more likely to be able to communicate with the units on a damaged network than Toolkit CGate?

 

Cheers Ashleigh,

Agree with all your comments.

The reason I was suspecting radiated or induced maybe the major problem is that I do not believe there is any other appliances or telcom that were effected by the strike. Of course that might just mean those devices have more robust ESD protection on the mains/earth and or signal paths.

Would have thought that if there was an EPR (earth potential rise) issue that it would be likely modem, fridge, TV's, audio gear etc etc would have had issues too. But as you said it will go where it dam where wants.

There is a Ness M1 system installed which has a C-Bus interface. The M1 survived OK but the C-Bus interface like all other C-Bus devices are not detected on a scan network. Your doubt about being an ionisation issue does hold up here as the C-Bus interface is connected both to C-Bus bus as well as Comm 0 of the Ness M1.

One thing is for sure an open mind needs to be in place when fault finding is being done as you said you can't rule any path in or out.

Agreed require a skilled person on site to resolve this issue and the quest to find such a person is underway. It ain't likely to be a 5 minute and $5 fix!

Good point on the insurance possible cover...worth checking.

Agree also on the difficulties in protecting against strikes. When I built up in FNQ I installed a Clispal surge and lightening suite consisting of a gas arrestor, delay and MOV. I was located just 50 metres from the sub division's sub power distribution transformer. EPR was a major concern in the event of local lightening strikes.

Hope your well.

Regards,

Ian

 

Spare a thought for this guy......

 

Always someone worse off than yourself! Can't believe he got up twice...

 

It's a fake!

 

Thanks again Ashleigh,

I have been pondering the informative reply and trying to wrap my head around the status of the system. Seems that from memory there are 5 units which are hooked up in series 4x shutter relays and 8 ch relay (c-Bus series) that all have no C-Bus green LED.

The two units I have clocks enabled on are showing solid green C-Bus LEDs indicating they are happy along with 3 other output unit up stream of the 5 not displaying the C-Bus OK LEDs.

I assume that each output unit ie dimmers, relays, shutter relays etc C-Bus I/O ports have an active buffered loop? If my assumption is correct and the unit is not passing through C-Bus comms traffic then any down stream device will not respond on the network.

In this installation the first thing the CNI sees is the M1 C-Bus interface which then connected to the 4 x shutter relays and then onto the relay unit that all are not displaying C-BUs Ok.

Very possible that the M1 interface and the 5 units mentioned above are possibly toast. Could be that the CNI on the C-Bus side is toast too. Either scenario are plausible reasons that TK can not see any device on the network.

This leads to your theory that the issue may have come in on the mains side not withstanding there is no other appliance down in the house from the strike. Gets back to every strike is different and goes where it dam well wants to.

Be interested in finding out if the output units with two C-Bus I/O RJ-45 have active buffered comms or are passive just pass through. I would have thought it would have been active to maintain current impedance and performance of the C-Bus network.

Regards,

Ian

 

There is no buffering between any CBus units. The 2 RJ45 connections are connected in parallel inside.The CBus is a single multidrop bus.
Any device can take out the whole network. (It's always seemed a bit of a limitation to me ).

I have a shutter relay and fan control sitting on my bench that when connected to the Cbus kills every unit on it. When viewed on an oscilloscope the dc voltage is fine but the cbus comms signal is suppressed to half it's proper level.

This is why I am a fan of star connected networks with a central distribution point rather than just ad-hoc wiring. When you have a fault like this you can easily segment the network and quickly find the malfunctioning unit.

 

I would be hesitant to say that the units which have the clock and the C-Bus indicator are OK. If they are generating the clock internally then the unit itself has a clock signal and so you have the C-Bus light, I would say that if you were to scope on the RJ45 sockets of those units you actually wouldn't have any C-Bus signal coming out, and that the isolation/protection between the In/Out and the C-Bus circuitry has been blown open.

I would say that both units have enabled their clocks and cant see anything else. As Ashley has said, it may just be one or two units that are pulling the whole bus down, but until a process of elimination can happen you really cant do that much personally.

J

 

They stopped responding to the C-Bus network entirely so it was easy to work out what wasn't working. Interrupting power to them didn't restore communications. The rest of the network was still functional though.

The home owner was happy to replace the dead modules so there wasn't any attempt at repairing the modules. The CTC was kept for spares but the rest was thrown away.

Yes. You need to split your network into smaller portions whilst testing it to track down what parts are faulty.

 

I lost three relay units last year after a strike nearby. One was in a stable block about 50m from the house, one in the pool filter room located under the house and another in a meter box about 30m from the main house.

No units in the house were affected. The strike occurred about midnight and we only knew something was wrong because two 400w flood lights came on so despite the units being 'dead' (had to replace) the channel with the floods zapped to 'ON' and couldn't be turned off as the unit was not responsive.

Not sure if that gives any clues. Expensive business when it happens...insurance covered the cost but the excess was $1,000.....

 

Thanks guys appreciate the informative feedback.

Knowing how the units process and or pass the bus through is very helpful. The fact that one unit can bring the entire C-Bus network down is also useful to know.

Segmenting the system down and having a few good C-Bus units in the toolbox to verify what does and doesn't work is definitely the go. It could be one or it could be many brake the system down into smaller systems/units and work up from there.

Thanks again for you collective input.

Regards,

Fleetz

 

Hi Fleetz,

Just spotted this post and thought my recent situation would help.
I had almost exactly the same issue as you have detailed, Lighting strike to the property pole with the tranny on it. All of the relays showed a unit light. Two out of ten had C-bus light. Various Dlt's had power but no control. After isolating a few legs of the system some areas began functioning as normal. I continued to reconnect legs until I found the cable that was killing the bus. The failure was a single 8channel relay.
I have had similar problems with C-bus in the past with almost identical solutions. It seems the rise in potential is passed on until the system finds one or two units to bully.

Brent

 

Thanks for the feedback Brent. Certainly sounds similar. Just depends how much energy the system has had to absorb and how many units as you eloquently put it were bullied!

Haven't got any feedback yet from my friends up in FNQ. They were going to contact the local Point One folks up there.

Will feedback what the final outcome is when known.

Cheers,

Fleetz