Hello! David, at your shop who do I thank for picking the T568A standard to support the One-Wire RJ45 ideas? (And of course the USOC RJ25 and the RJ25C codes as well.)
I am busy wiring an adapter for my adapter from DS/Maxim to my board. Essentially I decided to adapt the wiring arrangement between the sensors and the cable out from the adapter. Essentially I have a bundle of adapters who use those connectors to attach to either DB25s or DB9s. If this works it will become part of my continuing project. I'm working as fast as I can with this, believe me. Both of you, (the fellow who chose those, if iits not you) and yourself are entitled to a GMail account. (On me) -- Gregg C Levine [EMAIL PROTECTED] "This signature was once found posting rude messages in English in the Moscow subway." On 8/6/06, David Lissiuk <[EMAIL PROTECTED]> wrote: > Hi > > Let me add my two cents once again here about sharing 1-wire and other > protocols. The bottom line is don't. While it's possible in some cases > to get it to work you increase the chances of problems with your 1-wire > network. And the problems you see will be nearly impossible to > troubleshoot and may show up as early part failures, unexplained errors > and retries and the like, that seem to change with the every change of > the weather or day. CAT cabling is cheap when considered against the > troubles you may cause. > > The typical 1-wire slave device is specified with a VIH (guaranteed high > state) of 2.2 volts and a VIL (guaranteed low state) max of .8 volts. > This is the transition area in which the logic state is undefined for > the slave. (What the slave sees is really the critical thing in a 1-wire > system and in some ways the hardest thing to control due to cable > effects). > > With an active pull-up being triggered at .95 volts (as in the DS2480B) > there is not much margin for error. And with a cable impedance match of > 100 ohms for the CAT5/6 cable and a typical bus master weak pull-up of 4 > ma it pulls the low to about .4volts typically. So you are left with > about .4 volts working range for a low signal best case. > > Now if you consider the signal cable integrity effects of the 1-wire > signals over a transmission line (which a typical 1-wire net is), the > end effects can be quite startling due to the mis-matched slave > impedances. Simulations and real life measurements have shown that you > can see as much as .5volts overshoot/undershoot also (clamped with > BAT54S at the slave device) which may be near the slaves max limits to > handle in some situations. > > And you have to consider both the near and far end cable effects since > your slaves may be distributed throughout the cables length. The actual > effects depends on sensor placements, loads, slew rates, cable length > and other numerous things but the end effect is that it can push the > working range of the signal very hard depending on the system. Add in > the effects of hubs which lower the noise floor and you can see you > really very little if any at all to work with (depending of network > design). So ANY noise source (be it from crosstalk from an unregulated > supply ripple or signals like Ethernet can easily push the 1-wire over > the edge. And any of these reasons can show up as unexplained and > interment errors on the 1-wire network. > > I especially don't like unregulated supplies in the cable in particular, > since they can cause problems due to coupling to power line transients > and suffer from wide voltage ranges depending on load, and the quality > and wide tolerances and differences between manufactures of the > transformer and basic filtering in the supplies (Among other reasons). > Much better to use a regulated supply as they only cost a dollar or two > more over an unregulated supply and you generally only need one. And I > also prefer linear regulation (IMO) since they are cleaner and typically > better regulated than switching supplies. But either form of regulated > supply should work. Heavy use of BAT54's is also recommended to help > clamp signals at the sensors to the supply rails. > > The 1-wire protocol is surprising in that it can work due the relaxed > nature of its timing over a considerable range. But don't push it. Keep > to linear networks, use a good bus master (like the LINK), use lots of > protection throughout the network (BAT54s typically), and don't mix > other signals with your 1-wire cable. (Or at worst use good clean > regulated supplies in the cable if you must like the 1WRJ45 standard > allows). Don't mix multiple 1-wire networks into the cable and don't > double back the 1-wire signal through the same cable. Use the Dallas > RJ12 or 1WRJ45 wiring standard (both available on www.1wire.org for your > wiring standards). If you follow these simple rules things should work > fairly well and be fairly reliable. > > Note people have made all sorts of networks work, but they tend to be > unreliable or limited in range etc., and not at all worth the troubles > they can cause. Cat5E cable is cheap compared to your time spent trying > to find a problem. > > If you need to run another protocol, run it in its own cable at least > from the 1-wire network. > > Hope this helps, > Cheer > David Lissiuk > > > > > > > > > ------------------------------------------------------------------------- > Take Surveys. Earn Cash. Influence the Future of IT > Join SourceForge.net's Techsay panel and you'll get the chance to share your > opinions on IT & business topics through brief surveys -- and earn cash > http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV > _______________________________________________ > Owfs-developers mailing list > Owfs-developers@lists.sourceforge.net > https://lists.sourceforge.net/lists/listinfo/owfs-developers > ------------------------------------------------------------------------- Using Tomcat but need to do more? Need to support web services, security? Get stuff done quickly with pre-integrated technology to make your job easier Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642 _______________________________________________ Owfs-developers mailing list Owfs-developers@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/owfs-developers
