Sign in
Search 68hc12
Subscribe to 68hc12
68hc12 by Keywords
68HC1 | 812A4 | 9S12DP256 | Bootloader | CodeWarrior | D60A | Debugger | DP256 | ECT | EEPROM | EVB | Flash | HC1 | HCS12 | I2C | IAR | ICC1 | Interrupts | LCD | M68KIT912DP256 | MC9S12DP256 | MC9S12DP256B | Metrowerks | Motor | MSCAN | Multilink | PLL | Quadrature | SDI | SPI | Transceiver | XFC
Ads
Discussion Groups
See Also
Discussion Groups | 68HC12 | OT: Lightning protection
Join our technical discussions about Freescale Microcontrollers: M68HC12. (Freescale Semiconductor is a Subsidiary of Motorola).
OT: Lightning protection - garyolmstead - Nov 12 16:22:18 2006
I have an RS-485 network that will connect up to 100 separate points
that are all outdoors. The topic of lightning protection has come up,
but it is still sort of a question mark for us.
I have an idea that MOVs aren't adequate, but I'm not really sure what
would be. If you have solved this problem, what worke for you?
Gary Olmstead
Toucan Technology
Ventura CA
(You need to be a member of 68hc12 -- send a blank email to 68hc12-subscribe@yahoogroups.com )
that are all outdoors. The topic of lightning protection has come up,
but it is still sort of a question mark for us.
I have an idea that MOVs aren't adequate, but I'm not really sure what
would be. If you have solved this problem, what worke for you?
Gary Olmstead
Toucan Technology
Ventura CA
(You need to be a member of 68hc12 -- send a blank email to 68hc12-subscribe@yahoogroups.com )
Re: OT: Lightning protection - Donald E Haselwood - Nov 12 20:05:51 2006
On Sunday 12 November 2006 16:21, garyolmstead wrote:
> I have an RS-485 network that will connect up to 100 separate points
> that are all outdoors. The topic of lightning protection has come up,
> but it is still sort of a question mark for us.
>
> I have an idea that MOVs aren't adequate, but I'm not really sure what
> would be. If you have solved this problem, what worke for you?
>
> Gary Olmstead
> Toucan Technology
> Ventura CA
Gary,
I have an rs485 network that is outdoors, and in FL (Central FL is near the
top in the world for lightning-days per year). It has been in operation for
5 years with no problem. The number of points on my network is only 7, and
it spans about 300 feet.
For wiring I used cat5 (cheap) for +12, gnd, rs485 a|b, and #14 (also very
cheap at the time) for gnd/gnd and 24v AC. I ran a bare #8 gnd wire along
side the cat5 & #14, in hopes that ground currents from lightning would
travel along it and not punch through the insulation on the other wires.
(Some years ago the power company did some experimenting with underground
power transmission and found lightning currents punched holes in the
insulation.)
For the units, (HC11 based), the rs485 lines come in to a gas-sparkgap type of
protector (90v, both sides to ground), then via a small resistor (5-10 ohms),
to a TVS protectors (about 6-7v), and thence to a SN75176 (half duplex).
MOV's are high capacitance so they are not appropiate for the data lines.
The idea was that most of the energy in a surge would be absorbed by the
sparkgap protectors, and the TVS protectors (via the resistors) would keep
the voltage to SN75176 low enough for it to survive.
The +12 goes through a small resistor, 33 ohms, to a 15v MOV, and on to the
7805 regulator.
There has been no direct cloud-ground lightning hits in the last 5 years, but
there has been a number nearby, e.g. 1000 feet away, and hundreds of
cloud-cloud overhead. I don't know if the protection scheme is overkill, but
it has not failed.
My only failures has been due to ants getting into the units. Exposed
conductors with a voltage across them seem to draw ants. They die and
corrode away the traces on the pc board. They are highly conductive, too. I
had an outdoor electrical box on a circuit with a 20a breaker that would flip
when the humidity got high. I opened it up and there were a "million" dead
ants across the exposed conductors. I would have never thought they could be
conductive enough to trip a 20a breaker. So far ant poison in the box with
the electronics has been successful.
Regards,
Donald E Haselwood
(You need to be a member of 68hc12 -- send a blank email to 68hc12-subscribe@yahoogroups.com )
> I have an RS-485 network that will connect up to 100 separate points
> that are all outdoors. The topic of lightning protection has come up,
> but it is still sort of a question mark for us.
>
> I have an idea that MOVs aren't adequate, but I'm not really sure what
> would be. If you have solved this problem, what worke for you?
>
> Gary Olmstead
> Toucan Technology
> Ventura CA
Gary,
I have an rs485 network that is outdoors, and in FL (Central FL is near the
top in the world for lightning-days per year). It has been in operation for
5 years with no problem. The number of points on my network is only 7, and
it spans about 300 feet.
For wiring I used cat5 (cheap) for +12, gnd, rs485 a|b, and #14 (also very
cheap at the time) for gnd/gnd and 24v AC. I ran a bare #8 gnd wire along
side the cat5 & #14, in hopes that ground currents from lightning would
travel along it and not punch through the insulation on the other wires.
(Some years ago the power company did some experimenting with underground
power transmission and found lightning currents punched holes in the
insulation.)
For the units, (HC11 based), the rs485 lines come in to a gas-sparkgap type of
protector (90v, both sides to ground), then via a small resistor (5-10 ohms),
to a TVS protectors (about 6-7v), and thence to a SN75176 (half duplex).
MOV's are high capacitance so they are not appropiate for the data lines.
The idea was that most of the energy in a surge would be absorbed by the
sparkgap protectors, and the TVS protectors (via the resistors) would keep
the voltage to SN75176 low enough for it to survive.
The +12 goes through a small resistor, 33 ohms, to a 15v MOV, and on to the
7805 regulator.
There has been no direct cloud-ground lightning hits in the last 5 years, but
there has been a number nearby, e.g. 1000 feet away, and hundreds of
cloud-cloud overhead. I don't know if the protection scheme is overkill, but
it has not failed.
My only failures has been due to ants getting into the units. Exposed
conductors with a voltage across them seem to draw ants. They die and
corrode away the traces on the pc board. They are highly conductive, too. I
had an outdoor electrical box on a circuit with a 20a breaker that would flip
when the humidity got high. I opened it up and there were a "million" dead
ants across the exposed conductors. I would have never thought they could be
conductive enough to trip a 20a breaker. So far ant poison in the box with
the electronics has been successful.
Regards,
Donald E Haselwood
(You need to be a member of 68hc12 -- send a blank email to 68hc12-subscribe@yahoogroups.com )
Re: OT: Lightning protection - theobee00 - Nov 13 9:52:16 2006
--- In 6...@yahoogroups.com, "garyolmstead" wrote:
> I have an RS-485 network that will connect up to 100 separate points
> that are all outdoors. The topic of lightning protection has come up,
> but it is still sort of a question mark for us.
>
> I have an idea that MOVs aren't adequate, but I'm not really sure what
> would be. If you have solved this problem, what worke for you?
I wrote a paper for an IT course spurred on by building plans for my house, it's on top of a hill with RJ45 to a shed 30m further downhill.
Protection boils down to limiting any voltage differences below damage levels.
It helps to think along the lines of connecting –all- incoming cables to a central point, the nominal building ground, (give or take the protectors voltage) together with preventing this ground reaching too high a level relative to the next grounded point.
Say you sustain a hit on a building, the first step is to divert these impulse currents to Earth by means of lightning arrestors and a good earth plane made of deep stakes, bonding to water pipes, concrete mesh etc.
Essentially this is designed to limit potential differences across the building and preventing the whole shebangs voltage rising too far above true ground potential.
Even so count on the building as a whole reaching ten kV plus above true ground, depending on the Earth system and the magnitude of the impulse it can be even more. Here in Sydney we get some real lulus, strikes can reach 200kAmps.
The effect is that despite a good Earth system any incoming cables remaining at real ground potential will now have ten kV plus voltage differentials with respect to the internals of the building.
That means everything, telephone, RJ45, mains, all of them at 10kV relative to the building itself despite the Earths.
So the next bit is intended to keep –all- incoming cabling at the same potential as the building itself, in effect bringing their voltage up to 10kV if they like it or not.
Mains is easily fixed with heavy duty MOVs bonded to ground, that brings the effective mains pulse down to a kV or so, well within the ratings of insulation meant for mains, note that these MOVs often can sustain only limited strikes and then need to be replaced.
Telephone and RJ45 same thing, where they enter the building you connect diverters to the buildings central ground, these are readily available made specific for the purpose.
Of course they have lower protecting voltages, also different technologies are used, gas discharge tubes are popular (essentially a spark gap in inert gas).
The argument holds for all sorts of scenarios, a hit on outside cabling like overhead power lines will also be diverted to the buildings Earth by the MOVs again elevating the ground potential.
Note that any data cabling must always be protected from direct hits but induced voltages from any conductors carrying the strike current can be substantial even from overhead cables 5 meters or more away.
Next you need to consider potential differences between buildings, say my shed where the RJ45 comes in also has a good ground, made up of the same ¾" water pipe supplying the house, the mains supply cabling and its local Earth stake and an in ground concrete pool that has its mesh bonded to the sheds Earth.
That again will limit the possible voltage rise and hence the current that gets carried in the connecting cabling, RJ 45 cables will handle the expected 10kv differential along their length during a direct strike without evaporating especially if run along power and water pipes.
If you only want to consider induced voltages and remote strikes life becomes easier.
The same argument applies though, you must limit voltage differentials, bring any cabling into a building at a single point, incoming mains earth/neural and its stake are usually the central bonding and reference point, any diverters should bond as directly to there as you can make it before entering the building as a group.
Mains diverters are optional in that case too, if no direct strike close by is involved you simply assume the mains Earth remaining close to Active.
This would hopefully keep potentials between incoming cabling at sensible levels relative to each other and ground and thus protect your machines, but the cables will function as primary lightning arrestors in a hit.
We had cases recorded where a farms telephone cable exploded out of the ground for tens of meters, they convert to plasma, in such a case your puny RJ45 arrestors will bite the dust big time.
If any of the buildings involved are likely to be hit at all and have no lightning arrestors and good ground planes consider optical fibre, nowadays fairly cheap and its lightning proof …
My house has not sustained a direct hit so far, they are rare after all, but a strike about a hundred meters down the road resulted in a loud crack from the spark gaps, the machines did not even reset.
I live in hope…
Cheers,
(You need to be a member of 68hc12 -- send a blank email to 68hc12-subscribe@yahoogroups.com )
> I have an RS-485 network that will connect up to 100 separate points
> that are all outdoors. The topic of lightning protection has come up,
> but it is still sort of a question mark for us.
>
> I have an idea that MOVs aren't adequate, but I'm not really sure what
> would be. If you have solved this problem, what worke for you?
I wrote a paper for an IT course spurred on by building plans for my house, it's on top of a hill with RJ45 to a shed 30m further downhill.
Protection boils down to limiting any voltage differences below damage levels.
It helps to think along the lines of connecting –all- incoming cables to a central point, the nominal building ground, (give or take the protectors voltage) together with preventing this ground reaching too high a level relative to the next grounded point.
Say you sustain a hit on a building, the first step is to divert these impulse currents to Earth by means of lightning arrestors and a good earth plane made of deep stakes, bonding to water pipes, concrete mesh etc.
Essentially this is designed to limit potential differences across the building and preventing the whole shebangs voltage rising too far above true ground potential.
Even so count on the building as a whole reaching ten kV plus above true ground, depending on the Earth system and the magnitude of the impulse it can be even more. Here in Sydney we get some real lulus, strikes can reach 200kAmps.
The effect is that despite a good Earth system any incoming cables remaining at real ground potential will now have ten kV plus voltage differentials with respect to the internals of the building.
That means everything, telephone, RJ45, mains, all of them at 10kV relative to the building itself despite the Earths.
So the next bit is intended to keep –all- incoming cabling at the same potential as the building itself, in effect bringing their voltage up to 10kV if they like it or not.
Mains is easily fixed with heavy duty MOVs bonded to ground, that brings the effective mains pulse down to a kV or so, well within the ratings of insulation meant for mains, note that these MOVs often can sustain only limited strikes and then need to be replaced.
Telephone and RJ45 same thing, where they enter the building you connect diverters to the buildings central ground, these are readily available made specific for the purpose.
Of course they have lower protecting voltages, also different technologies are used, gas discharge tubes are popular (essentially a spark gap in inert gas).
The argument holds for all sorts of scenarios, a hit on outside cabling like overhead power lines will also be diverted to the buildings Earth by the MOVs again elevating the ground potential.
Note that any data cabling must always be protected from direct hits but induced voltages from any conductors carrying the strike current can be substantial even from overhead cables 5 meters or more away.
Next you need to consider potential differences between buildings, say my shed where the RJ45 comes in also has a good ground, made up of the same ¾" water pipe supplying the house, the mains supply cabling and its local Earth stake and an in ground concrete pool that has its mesh bonded to the sheds Earth.
That again will limit the possible voltage rise and hence the current that gets carried in the connecting cabling, RJ 45 cables will handle the expected 10kv differential along their length during a direct strike without evaporating especially if run along power and water pipes.
If you only want to consider induced voltages and remote strikes life becomes easier.
The same argument applies though, you must limit voltage differentials, bring any cabling into a building at a single point, incoming mains earth/neural and its stake are usually the central bonding and reference point, any diverters should bond as directly to there as you can make it before entering the building as a group.
Mains diverters are optional in that case too, if no direct strike close by is involved you simply assume the mains Earth remaining close to Active.
This would hopefully keep potentials between incoming cabling at sensible levels relative to each other and ground and thus protect your machines, but the cables will function as primary lightning arrestors in a hit.
We had cases recorded where a farms telephone cable exploded out of the ground for tens of meters, they convert to plasma, in such a case your puny RJ45 arrestors will bite the dust big time.
If any of the buildings involved are likely to be hit at all and have no lightning arrestors and good ground planes consider optical fibre, nowadays fairly cheap and its lightning proof …
My house has not sustained a direct hit so far, they are rare after all, but a strike about a hundred meters down the road resulted in a loud crack from the spark gaps, the machines did not even reset.
I live in hope…
Cheers,
(You need to be a member of 68hc12 -- send a blank email to 68hc12-subscribe@yahoogroups.com )
Re: OT: Lightning protection - eqh2 - Nov 13 10:02:38 2006
I have no expertise in this area but I would start by visiting B&B
(www.bb-elec.com) and searching on "lightning" if I were you. I have
no affiliation with them but have used their RS-485 converters for
years.
Good Luck,
Dan
--- In 6...@yahoogroups.com, "garyolmstead"
wrote:
>
> I have an RS-485 network that will connect up to 100 separate
points
> that are all outdoors. The topic of lightning protection has come
up,
> but it is still sort of a question mark for us.
>
> I have an idea that MOVs aren't adequate, but I'm not really sure
what
> would be. If you have solved this problem, what worke for you?
>
> Gary Olmstead
> Toucan Technology
> Ventura CA
>
(You need to be a member of 68hc12 -- send a blank email to 68hc12-subscribe@yahoogroups.com )
(www.bb-elec.com) and searching on "lightning" if I were you. I have
no affiliation with them but have used their RS-485 converters for
years.
Good Luck,
Dan
--- In 6...@yahoogroups.com, "garyolmstead"
wrote:
>
> I have an RS-485 network that will connect up to 100 separate
points
> that are all outdoors. The topic of lightning protection has come
up,
> but it is still sort of a question mark for us.
>
> I have an idea that MOVs aren't adequate, but I'm not really sure
what
> would be. If you have solved this problem, what worke for you?
>
> Gary Olmstead
> Toucan Technology
> Ventura CA
>
(You need to be a member of 68hc12 -- send a blank email to 68hc12-subscribe@yahoogroups.com )
Re: OT: Lightning protection - garyolmstead - Nov 13 12:01:55 2006
--- In 6...@yahoogroups.com, Donald E Haselwood
wrote:
>
> For wiring I used cat5 (cheap) for +12, gnd, rs485 a|b, and #14
(also very
> cheap at the time) for gnd/gnd and 24v AC. I ran a bare #8 gnd
wire along
>
> For the units, (HC11 based), the rs485 lines come in to a gas-
sparkgap type of
> protector (90v, both sides to ground), then via a small resistor
(5-10 ohms),
> to a TVS protectors (about 6-7v), and thence to a SN75176 (half
duplex).
>
Sounds pretty bullet proof. I don't have a good way to test it, so
it's probably better to go for a bit of overkill rather than
redesign it later.
> My only failures has been due to ants getting into the units.
Hummm, I hadn't considered that part.
Thanks.
Gary Olmstead
Toucan Technology
Ventura CA
(You need to be a member of 68hc12 -- send a blank email to 68hc12-subscribe@yahoogroups.com )
wrote:
>
> For wiring I used cat5 (cheap) for +12, gnd, rs485 a|b, and #14
(also very
> cheap at the time) for gnd/gnd and 24v AC. I ran a bare #8 gnd
wire along
>
> For the units, (HC11 based), the rs485 lines come in to a gas-
sparkgap type of
> protector (90v, both sides to ground), then via a small resistor
(5-10 ohms),
> to a TVS protectors (about 6-7v), and thence to a SN75176 (half
duplex).
>
Sounds pretty bullet proof. I don't have a good way to test it, so
it's probably better to go for a bit of overkill rather than
redesign it later.
> My only failures has been due to ants getting into the units.
Hummm, I hadn't considered that part.
Thanks.
Gary Olmstead
Toucan Technology
Ventura CA
(You need to be a member of 68hc12 -- send a blank email to 68hc12-subscribe@yahoogroups.com )
Re: OT: Lightning protection - garyolmstead - Nov 13 12:05:50 2006
--- In 6...@yahoogroups.com, "theobee00" wrote:
>
> Protection boils down to limiting any voltage differences below
damage levels.
> ....
> If you only want to consider induced voltages and remote strikes
life becomes easier.
> ...
> If any of the buildings involved are likely to be hit at all and
have no lightning arrestors and good ground planes consider optical
fibre, nowadays fairly cheap and its lightning proof …
>
Yes, the trouble is that this exercise is based on an almost off-
hand remark in the requirements document. So, I have no idea how
much or what kinds of risk there is. And here in Southern
California, lightning is rare enough to make the front page of the
newspaper. I do know that every place that connects to the RS485 is
a box mounted on a building, and that each building will have its
own AC mains power.
I like the fiber optic cable idea; but I suspect my client will be
happier with cat-5 cable, gas discharge tubes and TVSs.
Thanks.
Gary Olmstead
Toucan Technology
Ventura CA
(You need to be a member of 68hc12 -- send a blank email to 68hc12-subscribe@yahoogroups.com )
>
> Protection boils down to limiting any voltage differences below
damage levels.
> ....
> If you only want to consider induced voltages and remote strikes
life becomes easier.
> ...
> If any of the buildings involved are likely to be hit at all and
have no lightning arrestors and good ground planes consider optical
fibre, nowadays fairly cheap and its lightning proof …
>
Yes, the trouble is that this exercise is based on an almost off-
hand remark in the requirements document. So, I have no idea how
much or what kinds of risk there is. And here in Southern
California, lightning is rare enough to make the front page of the
newspaper. I do know that every place that connects to the RS485 is
a box mounted on a building, and that each building will have its
own AC mains power.
I like the fiber optic cable idea; but I suspect my client will be
happier with cat-5 cable, gas discharge tubes and TVSs.
Thanks.
Gary Olmstead
Toucan Technology
Ventura CA
(You need to be a member of 68hc12 -- send a blank email to 68hc12-subscribe@yahoogroups.com )
Re: OT: Lightning protection - garyolmstead - Nov 13 12:25:15 2006
--- In 6...@yahoogroups.com, Donald E Haselwood
wrote:
>
> For wiring I used cat5 (cheap) for +12, gnd, rs485 a|b, and #14
Did you do anything with the unused wires in the cat5 cable?
Gary Olmstead
Toucan Technology
Ventura CA
(You need to be a member of 68hc12 -- send a blank email to 68hc12-subscribe@yahoogroups.com )
wrote:
>
> For wiring I used cat5 (cheap) for +12, gnd, rs485 a|b, and #14
Did you do anything with the unused wires in the cat5 cable?
Gary Olmstead
Toucan Technology
Ventura CA
(You need to be a member of 68hc12 -- send a blank email to 68hc12-subscribe@yahoogroups.com )
Re: Re: OT: Lightning protection - Steve Russell - Nov 13 14:36:44 2006
At 08:53 AM 11/13/2006, garyolmstead wrote:
>I do know that every place that connects to the RS485 is
>a box mounted on a building, and that each building will have its
>own AC mains power.
Even with no lightning, you may have a problem with different ground
voltages. Depending on what the ground situation is ground at the far end
may not be within the RS 485 differential spec.
I've seen several volts AC between neutral and neutral in the same
building. I've also seen neutral and equipment ground interchanged on outlets.
I think that at least optically coupled RS485 is a good idea.
Fiber optics is really safe.
The 24 VAC should probably be isolated from local electronics ground.
At 04:54 PM 11/12/2006, Donald E Haselwood wrote:
>My only failures has been due to ants getting into the units. Exposed
>conductors with a voltage across them seem to draw ants. They die and
>corrode away the traces on the pc board. They are highly conductive, too.
Coating the boards with silicone goop would deal with the ants, and also
other side effects of high humidity. It makes repair more painful, but not
impossible. There were many different formulations available years ago
when I had occasion to deal with it.
Steve Russell
Nohau Emulators
(You need to be a member of 68hc12 -- send a blank email to 68hc12-subscribe@yahoogroups.com )
>I do know that every place that connects to the RS485 is
>a box mounted on a building, and that each building will have its
>own AC mains power.
Even with no lightning, you may have a problem with different ground
voltages. Depending on what the ground situation is ground at the far end
may not be within the RS 485 differential spec.
I've seen several volts AC between neutral and neutral in the same
building. I've also seen neutral and equipment ground interchanged on outlets.
I think that at least optically coupled RS485 is a good idea.
Fiber optics is really safe.
The 24 VAC should probably be isolated from local electronics ground.
At 04:54 PM 11/12/2006, Donald E Haselwood wrote:
>My only failures has been due to ants getting into the units. Exposed
>conductors with a voltage across them seem to draw ants. They die and
>corrode away the traces on the pc board. They are highly conductive, too.
Coating the boards with silicone goop would deal with the ants, and also
other side effects of high humidity. It makes repair more painful, but not
impossible. There were many different formulations available years ago
when I had occasion to deal with it.
Steve Russell
Nohau Emulators
(You need to be a member of 68hc12 -- send a blank email to 68hc12-subscribe@yahoogroups.com )
Re: OT: Lightning protection - garyolmstead - Nov 13 17:31:33 2006
--- In 6...@yahoogroups.com, Steve Russell wrote:
> Even with no lightning, you may have a problem with different ground
> voltages.
I probably should have mentioned that we are using the Maxim MAX3535
isolated RS-485 transceiver for that exact reason. The problem is
that the client appears to believe that this also confers lightning
protection, which of course is not true.
But here is a question for Donald and Theo: what did you do about the
PC traces outboard of the protection network? How heavy/wide are they?
Gary Olmstead
Toucan Technology
Ventura CA
(You need to be a member of 68hc12 -- send a blank email to 68hc12-subscribe@yahoogroups.com )
> Even with no lightning, you may have a problem with different ground
> voltages.
I probably should have mentioned that we are using the Maxim MAX3535
isolated RS-485 transceiver for that exact reason. The problem is
that the client appears to believe that this also confers lightning
protection, which of course is not true.
But here is a question for Donald and Theo: what did you do about the
PC traces outboard of the protection network? How heavy/wide are they?
Gary Olmstead
Toucan Technology
Ventura CA
(You need to be a member of 68hc12 -- send a blank email to 68hc12-subscribe@yahoogroups.com )
Re: Re: OT: Lightning protection - Donald E Haselwood - Nov 13 20:49:32 2006
On Monday 13 November 2006 12:02, garyolmstead wrote:
> --- In 6...@yahoogroups.com, Donald E Haselwood wrote:
> > For wiring I used cat5 (cheap) for +12, gnd, rs485 a|b, and #14
>
> Did you do anything with the unused wires in the cat5 cable?
>
> Gary Olmstead
> Toucan Technology
> Ventura CA
No, the unused cat5 wires are not used. I had thought that maybe they would
be useful if a pair "went bad," but that has not happened. The wires are
buried about 6" in sandy soil. I also had toyed with using the spares for
ethernet (today hacking the Linux code for the Linksys WRT54GS makes a good
way to communicate to the far reaches, greatly reducing the lightning
protection problem to just the power to the unit).
One problem I did have was a splice in the cat5 at one station was under water
for a week after a lot rain when one of the hurricanes came close. The
black, vinyl paint that I had used to coat the splices didn't completely
cover the +12 and where it was bare, electrolysis completely ate it away,
thus opening up the +12 to the that unit.
Don
(You need to be a member of 68hc12 -- send a blank email to 68hc12-subscribe@yahoogroups.com )
> --- In 6...@yahoogroups.com, Donald E Haselwood
> > For wiring I used cat5 (cheap) for +12, gnd, rs485 a|b, and #14
>
> Did you do anything with the unused wires in the cat5 cable?
>
> Gary Olmstead
> Toucan Technology
> Ventura CA
No, the unused cat5 wires are not used. I had thought that maybe they would
be useful if a pair "went bad," but that has not happened. The wires are
buried about 6" in sandy soil. I also had toyed with using the spares for
ethernet (today hacking the Linux code for the Linksys WRT54GS makes a good
way to communicate to the far reaches, greatly reducing the lightning
protection problem to just the power to the unit).
One problem I did have was a splice in the cat5 at one station was under water
for a week after a lot rain when one of the hurricanes came close. The
black, vinyl paint that I had used to coat the splices didn't completely
cover the +12 and where it was bare, electrolysis completely ate it away,
thus opening up the +12 to the that unit.
Don
(You need to be a member of 68hc12 -- send a blank email to 68hc12-subscribe@yahoogroups.com )
Re: Re: OT: Lightning protection - Donald E Haselwood - Nov 13 22:15:31 2006
On Monday 13 November 2006 17:18, garyolmstead wrote:
>
> I probably should have mentioned that we are using the Maxim MAX3535
> isolated RS-485 transceiver for that exact reason. The problem is
> that the client appears to believe that this also confers lightning
> protection, which of course is not true.
>
I considered the isolated type of RS485 drivers, but I agree with you that it
hardly confers lightning protection. With all the protection I put in ahead
of the driver I figured that there was little need for the more expensive
isolated versions.
> But here is a question for Donald and Theo: what did you do about the
> PC traces outboard of the protection network? How heavy/wide are they?
>
The cat5 has a RJ45 plug. The pc board has the jack. The sparkgap protection
is as close to the pins of the jack as practical. The traces were made as
wide as possible (i.e. copper pouring to fill the whole space), and top and
bottom (i.e. 2-sided pc board) was also used. The soldering adds a lot of
conductive mass as well. My SWAG was that the cat5 wire would fail before
the pc board->sparkgap. The extra, #8 ground wire was expected to carry the
heavy current.
BTW, regarding fiber. That would be an fine solution, if it were not for the
expense. My current thinking is to use the Linksys WRT54GS as a wireless
ethernet link-to-rs232. It runs Linux and Linksys makes the whole software
tool chain available, plus there are several www sites with software
modifications that allow it to be used as a bridge, repeater in a mesh
network, etc. The last I checked they were around $70. (The new WRT54G
models are cheaper, have less memory, and run Wind River, rather Linux.) A
header can be easily added which gives access to 2 standard Linux serial
ports, which can send/receive to local HC12s, etc. I currently use one as a
gateway to the RS485 network so it can be accessed from ethernet in my house
as well as providing isolation.
Don
______________________________
Stellaris® MCU Family: New Parts, New Package, New Price.
(You need to be a member of 68hc12 -- send a blank email to 68hc12-subscribe@yahoogroups.com )
>
> I probably should have mentioned that we are using the Maxim MAX3535
> isolated RS-485 transceiver for that exact reason. The problem is
> that the client appears to believe that this also confers lightning
> protection, which of course is not true.
>
I considered the isolated type of RS485 drivers, but I agree with you that it
hardly confers lightning protection. With all the protection I put in ahead
of the driver I figured that there was little need for the more expensive
isolated versions.
> But here is a question for Donald and Theo: what did you do about the
> PC traces outboard of the protection network? How heavy/wide are they?
>
The cat5 has a RJ45 plug. The pc board has the jack. The sparkgap protection
is as close to the pins of the jack as practical. The traces were made as
wide as possible (i.e. copper pouring to fill the whole space), and top and
bottom (i.e. 2-sided pc board) was also used. The soldering adds a lot of
conductive mass as well. My SWAG was that the cat5 wire would fail before
the pc board->sparkgap. The extra, #8 ground wire was expected to carry the
heavy current.
BTW, regarding fiber. That would be an fine solution, if it were not for the
expense. My current thinking is to use the Linksys WRT54GS as a wireless
ethernet link-to-rs232. It runs Linux and Linksys makes the whole software
tool chain available, plus there are several www sites with software
modifications that allow it to be used as a bridge, repeater in a mesh
network, etc. The last I checked they were around $70. (The new WRT54G
models are cheaper, have less memory, and run Wind River, rather Linux.) A
header can be easily added which gives access to 2 standard Linux serial
ports, which can send/receive to local HC12s, etc. I currently use one as a
gateway to the RS485 network so it can be accessed from ethernet in my house
as well as providing isolation.
Don
______________________________
Stellaris® MCU Family: New Parts, New Package, New Price.
(You need to be a member of 68hc12 -- send a blank email to 68hc12-subscribe@yahoogroups.com )
Re: Re: OT: Lightning protection - Frank - Nov 14 0:02:27 2006
You can use a optically/magnetically isolated tx/rx pairs. Have a look at the Analog
Devices iCoupler ADuM1100/ADuM120x etc.
An alternative is the use of normal opto coupler pairs chained in a half duplex loop and driven by a constant current loop current. Each node will have its transistor switched on while in idle mode and when transmitting, it will open the loop.
Since each node will have its own AC power there is no need for power supply isolation.
Gogetit!!
----- Original Message -----
From: garyolmstead
To: 6...@yahoogroups.com
Sent: Monday, November 13, 2006 8:53 AM
Subject: [68HC12] Re: OT: Lightning protection
--- In 6...@yahoogroups.com, "theobee00" wrote:
>
> Protection boils down to limiting any voltage differences below
damage levels.
> ....
> If you only want to consider induced voltages and remote strikes
life becomes easier.
> ...
> If any of the buildings involved are likely to be hit at all and
have no lightning arrestors and good ground planes consider optical
fibre, nowadays fairly cheap and its lightning proof .
>
Yes, the trouble is that this exercise is based on an almost off-
hand remark in the requirements document. So, I have no idea how
much or what kinds of risk there is. And here in Southern
California, lightning is rare enough to make the front page of the
newspaper. I do know that every place that connects to the RS485 is
a box mounted on a building, and that each building will have its
own AC mains power.
I like the fiber optic cable idea; but I suspect my client will be
happier with cat-5 cable, gas discharge tubes and TVSs.
Thanks.
Gary Olmstead
Toucan Technology
Ventura CA
[Non-text portions of this message have been removed]
(You need to be a member of 68hc12 -- send a blank email to 68hc12-subscribe@yahoogroups.com )
An alternative is the use of normal opto coupler pairs chained in a half duplex loop and driven by a constant current loop current. Each node will have its transistor switched on while in idle mode and when transmitting, it will open the loop.
Since each node will have its own AC power there is no need for power supply isolation.
Gogetit!!
----- Original Message -----
From: garyolmstead
To: 6...@yahoogroups.com
Sent: Monday, November 13, 2006 8:53 AM
Subject: [68HC12] Re: OT: Lightning protection
--- In 6...@yahoogroups.com, "theobee00"
>
> Protection boils down to limiting any voltage differences below
damage levels.
> ....
> If you only want to consider induced voltages and remote strikes
life becomes easier.
> ...
> If any of the buildings involved are likely to be hit at all and
have no lightning arrestors and good ground planes consider optical
fibre, nowadays fairly cheap and its lightning proof .
>
Yes, the trouble is that this exercise is based on an almost off-
hand remark in the requirements document. So, I have no idea how
much or what kinds of risk there is. And here in Southern
California, lightning is rare enough to make the front page of the
newspaper. I do know that every place that connects to the RS485 is
a box mounted on a building, and that each building will have its
own AC mains power.
I like the fiber optic cable idea; but I suspect my client will be
happier with cat-5 cable, gas discharge tubes and TVSs.
Thanks.
Gary Olmstead
Toucan Technology
Ventura CA
[Non-text portions of this message have been removed]
(You need to be a member of 68hc12 -- send a blank email to 68hc12-subscribe@yahoogroups.com )
Re: OT: Lightning protection - garyolmstead - Nov 14 0:27:45 2006
--- In 6...@yahoogroups.com, "eqh2" wrote:
>
> I have no expertise in this area but I would start by visiting B&B
> (www.bb-elec.com) and searching on "lightning" if I were you.
There are a number of places that sell that sort of thing. If I were
building one, or a few, systems that would definitely be the way to
go. But for our volumes, it's just too expensive.
Thanks for the suggestion.
Gary Olmstead
Toucan Technology
Ventura CA
(You need to be a member of 68hc12 -- send a blank email to 68hc12-subscribe@yahoogroups.com )
>
> I have no expertise in this area but I would start by visiting B&B
> (www.bb-elec.com) and searching on "lightning" if I were you.
There are a number of places that sell that sort of thing. If I were
building one, or a few, systems that would definitely be the way to
go. But for our volumes, it's just too expensive.
Thanks for the suggestion.
Gary Olmstead
Toucan Technology
Ventura CA
(You need to be a member of 68hc12 -- send a blank email to 68hc12-subscribe@yahoogroups.com )
Re: OT: Lightning protection - theobee00 - Nov 14 2:32:19 2006
--- In 6...@yahoogroups.com, "garyolmstead" wrote:
>
> --- In 6...@yahoogroups.com, "theobee00" wrote:
> >
> > Protection boils down to limiting any voltage differences below
> damage levels.
> > ....
> > If you only want to consider induced voltages and remote strikes
> life becomes easier.
> > ...
> > If any of the buildings involved are likely to be hit at all and
> have no lightning arrestors and good ground planes consider optical
> fibre, nowadays fairly cheap and its lightning proof …
> > Yes, the trouble is that this exercise is based on an almost off-
> hand remark in the requirements document. So, I have no idea how
> much or what kinds of risk there is. And here in Southern
> California, lightning is rare enough to make the front page of the
> newspaper. I do know that every place that connects to the RS485 is
> a box mounted on a building, and that each building will have its
> own AC mains power.
>
> I like the fiber optic cable idea; but I suspect my client will be
> happier with cat-5 cable, gas discharge tubes and TVSs.
If the client is made aware of the risks of not having a full scale protection system and is happy with the chance of a direct strike taking the whole system down, no worries, you are of the hook, be sure to make them very aware of the risks.
In that case you only have to worry about the mains distribution and potential mains fault currents between buildings, a few hundred volts max, standard protection will do the job, you simpy rely on the mains cabling to limit differential voltages.
It is only the high impulse curents from direct strikes that are a real worry.
It remains a good idea to have a central bonding point at each building for mains earths, pipes and your incoming cable if not already done that way.
My personal view is that if the client can't confirm good grounds on each building, go for fiber.
Anything else and you play roulette, as we all know clients still will blame us if the ball lands on 00.
Cheers,
Theo
(You need to be a member of 68hc12 -- send a blank email to 68hc12-subscribe@yahoogroups.com )
>
> --- In 6...@yahoogroups.com, "theobee00"
> >
> > Protection boils down to limiting any voltage differences below
> damage levels.
> > ....
> > If you only want to consider induced voltages and remote strikes
> life becomes easier.
> > ...
> > If any of the buildings involved are likely to be hit at all and
> have no lightning arrestors and good ground planes consider optical
> fibre, nowadays fairly cheap and its lightning proof …
> > Yes, the trouble is that this exercise is based on an almost off-
> hand remark in the requirements document. So, I have no idea how
> much or what kinds of risk there is. And here in Southern
> California, lightning is rare enough to make the front page of the
> newspaper. I do know that every place that connects to the RS485 is
> a box mounted on a building, and that each building will have its
> own AC mains power.
>
> I like the fiber optic cable idea; but I suspect my client will be
> happier with cat-5 cable, gas discharge tubes and TVSs.
If the client is made aware of the risks of not having a full scale protection system and is happy with the chance of a direct strike taking the whole system down, no worries, you are of the hook, be sure to make them very aware of the risks.
In that case you only have to worry about the mains distribution and potential mains fault currents between buildings, a few hundred volts max, standard protection will do the job, you simpy rely on the mains cabling to limit differential voltages.
It is only the high impulse curents from direct strikes that are a real worry.
It remains a good idea to have a central bonding point at each building for mains earths, pipes and your incoming cable if not already done that way.
My personal view is that if the client can't confirm good grounds on each building, go for fiber.
Anything else and you play roulette, as we all know clients still will blame us if the ball lands on 00.
Cheers,
Theo
(You need to be a member of 68hc12 -- send a blank email to 68hc12-subscribe@yahoogroups.com )
RE: Re: OT: Lightning protection - "Redd, Emmett R" - Nov 17 11:40:16 2006
--- In 6...@yahoogroups.com , Donald E
Haselwood wrote:
>
> One problem I did have was a splice in the cat5 at one station was
under water
> for a week after a lot rain when one of the hurricanes came close.
The
> black, vinyl paint that I had used to coat the splices didn't
completely
> cover the +12 and where it was bare, electrolysis completely ate it
away,
> thus opening up the +12 to the that unit.
>
> Don
>
Out in the country, some of us have submersible pumps in our wells and
the splices to
them are immersed all the time. We use two kinds of electrical tape.
The inner layers
are a soft, thick, self-adhesive tape that seals very well against water
intrusion. The
outer layers are made from regular PVC electrical tape for abrasion
resistance.
While 3m.com lists many tapes/systems for corrosion resistance/immersion
tolerance,
I could not find exactly what I have used (theirs have a backing strip;
mine does not) so
I cannot give any good references. A visit to your local hardware store
should be
sufficient. Or go with one of the backed 3M products.
Emmett Redd Ph.D. mailto:E...@missouristate.edu
Professor (417)836-5221
Department of Physics, Astronomy, and Materials Science
Missouri State University Fax (417)836-6226
901 SOUTH NATIONAL Dept (417)836-5131
SPRINGFIELD, MO 65897 USA
A bad day doing research is better than a good day doing something else.
[Non-text portions of this message have been removed]
(You need to be a member of 68hc12 -- send a blank email to 68hc12-subscribe@yahoogroups.com )
Haselwood
>
> One problem I did have was a splice in the cat5 at one station was
under water
> for a week after a lot rain when one of the hurricanes came close.
The
> black, vinyl paint that I had used to coat the splices didn't
completely
> cover the +12 and where it was bare, electrolysis completely ate it
away,
> thus opening up the +12 to the that unit.
>
> Don
>
Out in the country, some of us have submersible pumps in our wells and
the splices to
them are immersed all the time. We use two kinds of electrical tape.
The inner layers
are a soft, thick, self-adhesive tape that seals very well against water
intrusion. The
outer layers are made from regular PVC electrical tape for abrasion
resistance.
While 3m.com lists many tapes/systems for corrosion resistance/immersion
tolerance,
I could not find exactly what I have used (theirs have a backing strip;
mine does not) so
I cannot give any good references. A visit to your local hardware store
should be
sufficient. Or go with one of the backed 3M products.
Emmett Redd Ph.D. mailto:E...@missouristate.edu
Professor (417)836-5221
Department of Physics, Astronomy, and Materials Science
Missouri State University Fax (417)836-6226
901 SOUTH NATIONAL Dept (417)836-5131
SPRINGFIELD, MO 65897 USA
A bad day doing research is better than a good day doing something else.
[Non-text portions of this message have been removed]
(You need to be a member of 68hc12 -- send a blank email to 68hc12-subscribe@yahoogroups.com )