Lightening protection ?

Can someone suggest the most reliable way for lightening protecting a three
wire rs485 setup.

Boards on the structure(steel) will be connected via 6 or 8 core twisted
shielded.
Three wires for rs485,then +24VDC , pwr gnd and a few spare.
Option for separate power cable if required.

I was planning on using transorbs or similar to protect all the lines at
each board location
with the transorbs connected to local ground.

Five to eight daq boards per rs485 bus, with two to five buses per gateway.
Two gateways at the moment.
Server to the gateways is approx 100 - 200 feet(Ethernet).
Then each rs485 bus can be 50 - 200 feet.

Trying to keep the cables runs reasonably short.

Some of the people I'm working with want to use zigbee or 802.11 but
still need power plus wireless and steel structures don't mix well.

I'd much prefer to stick with an industrial solution that will work
for the base system, then if there is time and budget
throw in a few wireless nodes.

Any tips or tricks or things to avoid ?

Thank you

Alex

On Thu, 5 Apr 2007 11:54:10 +1000, "Alex Gibson" <>
wrote:

>Can someone suggest the most reliable way for lightening protecting a three
>wire rs485 setup.
>
>Boards on the structure(steel) will be connected via 6 or 8 core twisted
>shielded.
>Three wires for rs485,then +24VDC , pwr gnd and a few spare.
>Option for separate power cable if required.

Unless the steel structure is an antenna tower :-), I would simply use
galvanic isolation on both the serial data and the power supply side.

The RS-485 works fine with a single twisted pair without separate
signal ground with termination resistors at both ends of the bus, so
there is not going to be any significant differential mode voltages on
the line and the isolation can handle 500-2500 V common mode voltage.

One other pair could be dedicated for the +24V and 0 V and use
isolated switched mode power supplies at each device. Ground the cable
shield at a single point to avoid loop currents.

With transmitting antenna towers, you must be aware of the lightning
rod current, which generates a huge magnetic field around the
conductor and also about RF fields with possibly wavelengths of the
same order of magnitude as the cable length.

Paul

"Paul Keinanen" <> wrote in message
news:...
> On Thu, 5 Apr 2007 11:54:10 +1000, "Alex Gibson" <>
> wrote:
>
>>Can someone suggest the most reliable way for lightening protecting a
>>three
>>wire rs485 setup.
>>
>>Boards on the structure(steel) will be connected via 6 or 8 core twisted
>>shielded.
>>Three wires for rs485,then +24VDC , pwr gnd and a few spare.
>>Option for separate power cable if required.
>
> Unless the steel structure is an antenna tower :-), I would simply use
> galvanic isolation on both the serial data and the power supply side.

Old steel bridge over salt water.

> The RS-485 works fine with a single twisted pair without separate
> signal ground with termination resistors at both ends of the bus, so
> there is not going to be any significant differential mode voltages on
> the line and the isolation can handle 500-2500 V common mode voltage.

The bridge structure is the highest point for quite a distance(few km's)
and has a center lift section.
The electric motors for the lift section are the originals from the 1940's
but may have been rewound in the 70's.

The AC power in the area is very noisy due to
a lot of industrial and mining equipment.
Bridge has 240VAC for lighting and the lift motors.

> One other pair could be dedicated for the +24V and 0 V and use
> isolated switched mode power supplies at each device. Ground the cable
> shield at a single point to avoid loop currents.
>
> With transmitting antenna towers, you must be aware of the lightning
> rod current, which generates a huge magnetic field around the
> conductor and also about RF fields with possibly wavelengths of the
> same order of magnitude as the cable length.
>
> Paul

Thank you

Alex

Alex Gibson wrote:

> Can someone suggest the most reliable way for lightening protecting a
> three wire rs485 setup.
>
> Boards on the structure(steel) will be connected via 6 or 8 core twisted
> shielded.
> Three wires for rs485,then +24VDC , pwr gnd and a few spare.
> Option for separate power cable if required.
>
> I was planning on using transorbs or similar to protect all the lines at
> each board location
> with the transorbs connected to local ground.
>
> Five to eight daq boards per rs485 bus, with two to five buses per
> gateway. Two gateways at the moment.
> Server to the gateways is approx 100 - 200 feet(Ethernet).
> Then each rs485 bus can be 50 - 200 feet.
>
> Trying to keep the cables runs reasonably short.
>
> Some of the people I'm working with want to use zigbee or 802.11 but
> still need power plus wireless and steel structures don't mix well.
>
> I'd much prefer to stick with an industrial solution that will work
> for the base system, then if there is time and budget
> throw in a few wireless nodes.
>
> Any tips or tricks or things to avoid ?

Like Paul Keinanen has already stated in response, you need to look at
Galvanic Isolation of the RS485 bus and the power supplies to each board.
The drivers connected to the bus may have to be considered sacrificial for
the worst of strikes unless you can really bolster up the interface
somewhat.

I took the RS485 style bus for a rail project and operated the switching
levels at 52V. I had to develop drivers and receivers for this special
situation but we were faced with getting signals through old train
autocouplers and resisting extreme transients from the rail stock traction
power.

Any time you spend modeling the interfaces with SPICE will be well worth the
effort. You need to explore the ways in which you can control the energy
seen by your circuits. Spark Gaps, resistors, inductors, capacitors and
transorbs, may feature in some mix. Ensuring full galvanic isolation and
limiting the level of transient seen by such isolation barriers will
protect the inner (and usually more expensive) circuitry. Some links that
may provide useful information for you:-

<http://ieeexplore.ieee.org/Xplore/login.jsp?url=/iel5/61/30625/01413412.pdf?arnumber=1413412>
<http://www.thomasnet.com/products/suppressors-transient-voltage-surge-81400103-1.html>
<http://www.flomerics.com/events/event_details.jsp?event_id=2682>
<http://www.slopeindicator.com/pdf/papers/new-victoria-dam-case-study.pdf>
<http://www.arema.org/eseries/scriptcontent/custom/e_arema/library/2000_Conference_Proceedings/00055.pdf>

--
************************************************** ******************
Paul E. Bennett ....................<email://>
Forth based HIDECS Consultancy .....<http://www.amleth.demon.co.uk/>
Mob: +44 (0)7811-639972
Tel: +44 (0)1235-811095
Going Forth Safely ..... EBA. www.electric-boat-association.org.uk..
************************************************** ******************

On Apr 5, 11:54 am, "Alex Gibson" <n...@alxx.org> wrote:
> Can someone suggest the most reliable way for lightening protecting a three
> wire rs485 setup.
>
Keep the cables and equipment in the shade or indoors will help with
this.
Unless you mean lightning in which case this will offer little
protection.

David

"Paul E. Bennett" <> wrote in message
news:ev2d15$ikn$1$...

>
> Like Paul Keinanen has already stated in response, you need to look at
> Galvanic Isolation of the RS485 bus and the power supplies to each board.
> The drivers connected to the bus may have to be considered sacrificial for
> the worst of strikes unless you can really bolster up the interface
> somewhat.
>
> I took the RS485 style bus for a rail project and operated the switching
> levels at 52V. I had to develop drivers and receivers for this special
> situation but we were faced with getting signals through old train
> autocouplers and resisting extreme transients from the rail stock traction
> power.
>
> Any time you spend modeling the interfaces with SPICE will be well worth
> the
> effort. You need to explore the ways in which you can control the energy
> seen by your circuits. Spark Gaps, resistors, inductors, capacitors and
> transorbs, may feature in some mix. Ensuring full galvanic isolation and
> limiting the level of transient seen by such isolation barriers will
> protect the inner (and usually more expensive) circuitry. Some links that
> may provide useful information for you:-
>
> <http://ieeexplore.ieee.org/Xplore/login.jsp?url=/iel5/61/30625/01413412.pdf?arnumber=1413412>
> <http://www.thomasnet.com/products/suppressors-transient-voltage-surge-81400103-1.html>
> <http://www.flomerics.com/events/event_details.jsp?event_id=2682>
> <http://www.slopeindicator.com/pdf/papers/new-victoria-dam-case-study.pdf>
> <http://www.arema.org/eseries/scriptcontent/custom/e_arema/library/2000_Conference_Proceedings/00055.pdf>


Thank you, just what I'm after

Alex

On Apr 5, 3:58 am, "Alex Gibson" <n...@alxx.org> wrote:
>>>Can someone suggest the most reliable way for lightening protecting a
>>>three wire rs485 setup.
> ...
> The bridge structure is the highest point for quite a distance(few km's)
> and has a center lift section.
> The electric motors for the lift section are the originals from the 1940's
> but may have been rewound in the 70's.

You have the exact same problem that your telco has when connected
to every other building in town and with overhead wires all over
town. Use this application note figure to understand the principles:
http://www.erico.com/public/library/...es/tncr002.pdf

Every wire (overhead or underground) connects to same earth ground
before entering the building. No way around that necessary reality.
Every wire in every cable connects to that ground either directly or
via a protector. Transients earthed before entering a building do not
overwhelm protection (ie galvanic isolation) already inside all
electronics.

Yes you need galvanic isolation for the trivial transients. But that
internal protection is easily overwhelmed without single point
earthing.

Same is how the telco protects a $multi-million computer and
operates during every thunderstorm without damage.

Consider, for example, this product that has that necessary earthing
wire:
http://www.tripplite.com/products/pr...?productID=151
Each wire then connects 'less than 10 feet' to a single point earth
ground if its earthing wire is properly connected.

In that above application note, the tower (ie other building) and
building are separate structures. Each is earthed. Still, wire must
be in contact with the other structures earthing before entering that
structure.

Introduced is how damage is created and how damage is eliminated
using the phone system as a similar example. See this one long post
in the newsgroup comp.sys.ibm.pc.hardware.storage entitled "Phone
line surge protection?" on 1 Apr 2007 in:
http://tinyurl.com/3abu4q

this is who we begin to understand why your telco need not shutdown
and disconnect phone service during every storm. Why 911 emergency
operators don't remove headsets and leave the room. This is how
protection was installed even 70 years ago.