owfs and owhttpd

On Thu, May 14, 2009 at 11:41 AM, Rob Fugina <rob.fugina@...> wrote:

> I hope this isn't too off-topic for the group.  If it is, please
> direct me somewhere else.
>
> I need to get some 1-wire weather instruments mounted away from my
> house, on a post about 75 ft away.  This worries me because of how the
> system basically becomes an antenna.  Nearby lightning strikes would
> wreak havok back at the bus master.  AAG seems to have recognized this
> problem, as the bus master they ship with their weather station is
> optically isolated.  But I don't know if their bus master is supported


It probably is, but hasn't been tested. It depends on how good the OS
support is for that particular serial->usb convertor.


>
> by owfs...  It's based on a DS2480(B?) chip, which probably means
> their USB device is really a USB-to-RS232 converter in the first
> place...
>
> So I see several possible solutions:
>
> 1)  The risk is negligible, and it's not worth worrying about.  Does
> anybody do something like what I'm describing without problems that
> can testify about it?
>
> 2)  AAG has it right, is the only game in town, and I need to bite the
> bullet, give up real USB, and buy AAG's gear.  Will one owserver
> instance talk to both USB and RS232 bus masters and aggregate the
> results?


One owserver instance will aggregate any number of bus masters, of any kind,
including other owserver instances, local and reote. Magic.


>
>
> 3)  Maybe it's not difficult to optically isolate a 1-wire bus.  Does
> anybody have a circuit diagram?
>
> 4)  Last but not least, but probably the most complicated:  Are there
> 1-wire-to-wireless bridge chips/circuits around somewhere?  That would
> prevent me having to bury cable, too, although I'd have to worry about
> how to power the station...
>

Any of the supported routers (OpenWRT project) can host owserver. You home
machine can then treat the remote machine as an owserver instance.

Am Donnerstag, 14. Mai 2009 schrieb Rob Fugina:
> I need to get some 1-wire weather instruments mounted away from my
> house, on a post about 75 ft away.  This worries me because of how the
> system basically becomes an antenna.  Nearby lightning strikes would
> wreak havok back at the bus master.
>
A lightning striking a pole only ~25m away would shift earth level very much 
and damage most of your computer equipment anyway. It makes no sense just to 
protect one input -- You'll have to protect all inputs, especially the mains. 

Where exactly do you live? A single house or an urban area? Are thunderstorms 
frequent? How dry is the soil at your area? These are important questions if 
you ask how to protect your equipment from lightning.

Kind regards

	Jan
-- 
You cannot achieve the impossible without attempting the absurd.

On Thu, May 14, 2009 at 14:33, Jan Kandziora <jjj@...> wrote:
> Am Donnerstag, 14. Mai 2009 schrieb Rob Fugina:
>> I need to get some 1-wire weather instruments mounted away from my
>> house, on a post about 75 ft away.  This worries me because of how the
>> system basically becomes an antenna.  Nearby lightning strikes would
>> wreak havok back at the bus master.
>>
> A lightning striking a pole only ~25m away would shift earth level very much
> and damage most of your computer equipment anyway. It makes no sense just to
> protect one input -- You'll have to protect all inputs, especially the mains.
>
> Where exactly do you live? A single house or an urban area? Are thunderstorms
> frequent? How dry is the soil at your area? These are important questions if
> you ask how to protect your equipment from lightning.

Ok, you're right, and I'm not really trying to protect myself from a
direct strike.  But having a 75m antenna sticking out into my back
yard does seem to be asking for nearby lightning, or overhead
lightning, to induce significant current in that wire -- or am I
wrong?  I live in the far suburbs -- very few trees, 1-acre lots,
houses 50m to 150m apart.  The soil's pretty wet right now, and I
would say it never gets all that terribly dry.

In answer to Paul's suggestion, I hadn't really considered running a
full-on WiFi node on that pole, but I'm sure that's a possibility,
too.  When I said "wireless" I was thinking of something at a much
lower level -- at the 1-wire level, in particular...  But I've been
thinking about this more now, and I've got an NSLU2 hiding somewhere
in my house, and I have a pair of 10base-F transceivers that I got at
a hamfest a while back.  Now if I could just power the NSLU2 and a
transceiver and bury some fiber, that'd 100 times more reliable than
wireless, and perfectly isolated.  Now I wonder what it would take to
power them...

Rob

Rob, 

I've been looking around for running 1-Wire over Zigbee but these thoughts are in a very early stage and I dont even have the devices yet. Maybe I can report more in a few weeks, but in terms of costs and easy usage I'd currently say it's most likely better to pick up Pauls suggestion and simply connect a 1-wire busmaster to some Openwrt-Box..
Nethertheless, I'd be interested to opto-isolate the 1-wire bus from the master for other (regulatory) reasons ?

regards Michael

Hi,

Rob Fugina:
> Ok, you're right, and I'm not really trying to protect myself from a
> direct strike.  But having a 75m antenna sticking out into my back
> yard does seem to be asking for nearby lightning, or overhead
> lightning, to induce significant current in that wire -- or am I
> wrong?

No, that's a real problem. I wouldn't do that without insulation
of some kind.

My own remote-sensing devices are NSLU2s; I simply throw a 240V cable at
them and use Ethernet-over-mains adapters for connectivity.

The setup is pretty reliable; we get no power failures to speak of, so I
don't even have to worry about restarting the slugs after a failure.

-- 
Matthias Urlichs   |   {M:U} IT Design @ m-u-it.de   |  smurf@...
Disclaimer: The quote was selected randomly. Really. | http://smurf.noris.de
v4sw7$Yhw6+8ln7ma7u7L!wl7DUi2e6t3TMWb8HAGen6g3a4s6Mr1p-3/-6 hackerkey.com
 - -
A wise ruler ought never to keep faith when by doing so it would be against
his interests.
					-- Niccolo Machiavelli

Am Donnerstag, 14. Mai 2009 schrieb Rob Fugina:
>
> Ok, you're right, and I'm not really trying to protect myself from a
> direct strike.  But having a 75m antenna sticking out into my back
> yard
>
75m band, I think? If it is really 75m high, someone should already have done 
deep thinking about lightning...


> does seem to be asking for nearby lightning, or overhead 
> lightning, to induce significant current in that wire -- or am I
> wrong?
>
No, you are pretty right. But it's just the way a direct lightning strike will 
damage any electronic equiment in a range of ~50m, regardless of protective 
circuits. It just runs over any surface and grasps any kind of conductor.
Optocouplers are no match for lightning -- it just made it about 2kms down to 
earth, so 2mm would be no problem, right?.

The only chance to protect against direct strikes is proper earthing. Wet soil 
and rings of metal buried deep into the ground, if possible connected to 
ground water. That way, the "range of destruction" can be shrinked to about 
~10m around the pole.


So I don't think the optocouplers in the AAG equipment are for lightning 
protection. They are for ground separation -- otherwise there would be a 
noise-inducing ground loop between the PC, which *always* connects signal 
ground to protective earth (major design flaw) and an auxiliary-powered 
sensor outside.

Kind regards

	Jan
-- 
Williams and Holland's Law:
	If enough data is collected, anything may
	be proven by statistical methods.

On Thu, May 14, 2009 at 16:20, Jan Kandziora <jjj@...> wrote:
> 75m band, I think? If it is really 75m high, someone should already have done
> deep thinking about lightning...

Oh, it's not 75m high...  I got the units screwed up going back and
forth.  I originally said 75 ft, then someone said 25m, and I came
back with 75m.  Regardless, it's not up, it's out.  It'll be about 4m
up.

> So I don't think the optocouplers in the AAG equipment are for lightning
> protection. They are for ground separation -- otherwise there would be a
> noise-inducing ground loop between the PC, which *always* connects signal
> ground to protective earth (major design flaw) and an auxiliary-powered
> sensor outside.

As I said, it's not direct lightning I think I can protect myself from
this way, but nearby lightning.  Also, as you've mentioned now, ground
loops.  Ug...  I knew I was trying to do the right thing, just
apparently for the wrong reasons...

Rob

Am Donnerstag, 14. Mai 2009 schrieb Rob Fugina:
>
> It'll be about 4m up.
>
A 4m pole made of wood will without much doubt *never* been hit by 
lightning... It's much more likely a 20m high tree nearby is hit (which alone 
is very unlikely)

If you use a metal pole, a halved, tin foiled ball on the top of the pole 
should be enough to avoid any corona discharge and thus, lightning strikes. 

Kind regards

	Jan
-- 
"The C Programming Language -- A language which combines the
flexibility of assembly language with the power of assembly language."

Am Donnerstag, 14. Mai 2009 schrieb Paul Alfille:
> I found an obscure bug in the device cache code. Can you test?
>
I tried it, but no change.

Kind regards

	Jan
-- 
TkKasse will not run on any kind of MS-Windows. It can be adapted to
do so but I won't do it, as most Microsoft customers are expecting
software to be broken, and TkKasse isn't. I can't break it just
for the sake of compatibility to these people's mindset.