DComTalk.com

Rich Noonan · Guest

I'm looking for a device which could act as an ethernet breaker switch
based on the application or loss of power to that device. On one hand
we could accomplish what we need with a low-cost switch and two cables.
However, in our application and environment this would introduce
several new problems. What I really need is a two port fast ethernet
repeater that I could cut power to and break the link. Of course, such
a thing doesn't exist (or does it?).

Any suggestions on a part that could fill this role? We could probably
design and build a mechanical switch to break the link, but I'm trying
to find something we could buy before going down that road.

Thanks in advance for your suggestions.

-Rich

Walter Roberson · Guest

In article <1126791575.591318.240920@o13g2000cwo.googlegroups.com>,
Rich Noonan <rnoonan@gmail.com> wrote:
:I'm looking for a device which could act as an ethernet breaker switch
:based on the application or loss of power to that device. On one hand
:we could accomplish what we need with a low-cost switch and two cables.
: However, in our application and environment this would introduce
:several new problems. What I really need is a two port fast ethernet
:repeater that I could cut power to and break the link. Of course, such
:a thing doesn't exist (or does it?).

Your reference to "several new problems" makes it difficult for us to
know what would be acceptable and what would not.

I don't know whether this will help you, but recently I have been
investigating remote-controllable serial ports and remote
controllable power switches. For my particular application, I was
looking at ones that can be reached via ethernet. Especially for
the power switches I also found devices that can be controlled via
serial lines -- up to 2000 feet for one of the devices.

Typical remote control units have 1, 8, or 16 ports, with 4 and 32
port units not too hard to find, and with 2 port units available but
not easy to find.

The choices are differentiated on their remote access method (serial /
ethernet); their security (ssh?); their modularity; their flexibility
(several are uLinux with SDKs); their number of ports; their physical
size; the amperage the power controls handle; and, of course,
upon their cost.

Links to several manufacturers can be found at
http://www.voip-info.org/tiki-print.php?page=Remote+Console+and+Power+Control

In addition to those, I suggest checking out http://www.synaccess-net.com/
(Note: their web page has not yet caught up to several recent security
enhancements.)

I have several other units bookmarked; if you want a bookmark dump,
send me an email note.
--
Entropy is the logarithm of probability -- Boltzmann

Walter Roberson · Guest

In article <dgc27d$n5i$1@panix5.panix.com>, Al Dykes <adykes@panix.com> wrote:
:If you add up the all the extra electrical connections, power sources
:etc for a relay-based system in use on any scale I'd argue that a
:managed hub with a web interface is *much* more reliable and in the
:unlikely case that it does fail it will do so in *much* less
:interesting ways and be trivial to spare and fix.

Ummm... where are you going to find a managed *hub* with a web interface?

There were some managed hubs, but they weren't common at the best of times,
and they predated widespread use of web interfaces. They used serial, or
telnet + snmp, or custom GUIs.

These days, it is apparently difficult to find true hubs; according to
some of the groups I follow, devices being marketed as hubs have often
been switched over by the manufacturer to actually be switches under the
cover.
--
Any sufficiently old bug becomes a feature.

Rich Noonan · Guest

Thanks for the replies. Maybe I should describe the actual
application.

We have a machine with multiple PC based operator interfaces. During
machine servicing it is sometimes necessary to disable control of the
machine at all but one interface PC. We are looking for a way to
disable control that is a) not software based (we already have numerous
s/w interlocks built in to the system and b) doesn't involve removing
power to the PC. What we'd like to do is simply disable the data line
to the PC.

Two easily implemented methods are currently under consideration.

1) Use a managed ethernet switch at the center of the control system
and disable individual switch ports as needed. The problems are that
switch silicon
(arguably or provably) has a higher failure rate than say a
mechanical/electrical switch and to the user/safety inspector it still
looks and feels like a software switch.

2) Install a low-density ethernet (branch) switch between each PC and
the central (core) ethernet switch with an electrical relay controlling
power delivery to those branch switches. The relay states can then be
selected with a traditional mechanical knob switch. The problems here
are that we just greatly increase the number of potential failure
points in the control system and introduced a whole bunch more unused
ethernet ports which we are trying to limit for other safety reasons
(e.g. factory staff plugging in computers without having to follow
procedure).

My imaginary solution is a fairly simple, two port device that would
sit inline on the PC ethernet link and when it lost power would break
the link. Do you know of such a device?

Thanks again-
Rich

PS I'm using the groups.google interface for the first time. Apologies
if it does anything rude like top post or full-quote.

Al Dykes · Guest

In article <1126794928.866211.26990@z14g2000cwz.googlegroups.com>,
Rich Noonan <rnoonan@gmail.com> wrote:

Al Dykes · Guest

In article <1126791575.591318.240920@o13g2000cwo.googlegroups.com>,
Rich Noonan <rnoonan@gmail.com> wrote:

Walter Roberson · Guest

In article <1126794928.866211.26990@z14g2000cwz.googlegroups.com>,
Rich Noonan <rnoonan@gmail.com> wrote:
:My imaginary solution is a fairly simple, two port device that would
:sit inline on the PC ethernet link and when it lost power would break
:the link. Do you know of such a device?

There are many "broadband routers" on the market, a number of which
can be set to turn off NAT.
--
"No one has the right to destroy another person's belief by
demanding empirical evidence." -- Ann Landers

Walter Roberson · Guest

In article <dgc8ds$t1e$1@panix5.panix.com>, Al Dykes <adykes@panix.com> wrote:
:Somehow I think we are playing in different leagues. CISCO, HP and a
:pile of other venders make managed hubs and switches. The HP procurve
:stuff (except for the bottom end) has built-in WEB management
:interface. I'm not familiar with CISCO.

I found the Procurve Hubs on HP's site, and you are right that they
do have a Web interface. On the other hand, checking around to stores,
I'm finding them available as Used, or "1 in stock" (at wildly different
prices.)

:Look for an HP Procurve box on ebay

If you have to buy it on eBay now, then it it isn't a good choice
for devices intended to be easily replaceable.
--
Daylight is a trademark of OSRAM SYLVANIA INC.

Frank Stutzman · Guest

A long time ago, at a trade show far away...

There was a company called LanHopper that made managable (via console or
out of band SNMP) mechanical ethernet "switches". I put switches in
quotes as they were not what we consider switches today. The were full of
relays. You could hook one port to any other port, or just physically
disconnect a port.

Sounds like what the OP is looking for. Have no idea where they went or
what happened to them.

--
Frank Stutzman
Bonanza N494B "Hula Girl"
Hood River, OR

Walter Roberson · Guest

In article <1126811324.599443.75360@f14g2000cwb.googlegroups.com>,
Rich Noonan <rnoonan@gmail.com> wrote:

Al Dykes · Guest

In article <dgc65v$j6n$1@canopus.cc.umanitoba.ca>,
Walter Roberson <roberson@ibd.nrc-cnrc.gc.ca> wrote:

Al Dykes · Guest

In article <dgc99i$n7s$1@canopus.cc.umanitoba.ca>,
Walter Roberson <roberson@ibd.nrc-cnrc.gc.ca> wrote:

William P. N. Smith · Guest

"Rich Noonan" <rnoonan@gmail.com> wrote:

Rich Noonan · Guest

William P. N. Smith wrote:

Jack Masters · Guest

William P. N. Smith wrote:

	DComTalk.com Forum Index -> Ethernet	All times are GMT Goto page 1, 2, 3 Next
Page 1 of 3