Brian - I have exactly the same problem(s), using the Linux version of MH.
I really suspect it has to do with the way the 1-wire code distributed with
MH deals with the COM ports. It appears to be abstracted from 'generic'
Windoze code, some of which is flat unnecessary when using the Linux
drivers, and, I suspect that the Dallas code bypasses Windoze as much as
I just haven't had the time to play with things enough to test out this
theory (and, my 1-wire sensors are all still in boxes after a broken water
pipe flooded the lower half of our house). I'll probably be getting to that
layer of boxes sometime in the next few weeks - my plan is to model
something after the Dallas linux or java code and see what happens. I have a
whole bunch of temps I'd like to monitor (freezer, HVAC return, HVAC inlet,
two 'outlying' pipes - for freezing, and several ground temps - for
gardening), but I'd REALLY rather not have to work with 2409s (both the
extra coding, not so big a deal; and the SMT soldering, a BIG deal with my
shaky hands - that's why I'm SOFTWARE Engineer).
I'll keep posting my progress.
> -----Original Message-----
> From: misterhouse-users-admin@...
> [mailto:misterhouse-users-admin@... Behalf Of
> Paulsen, Brian
> Sent: Wednesday, December 27, 2000 10:40 AM
> To: 'misterhouse-users@...'
> Subject: RE: [misterhouse-users] iButton/1-wire questions
> To answer some of my own questions...
> http://www.dalsemi.com/datasheets/pdfs/app132.pdf has a document about how
> to make your 1-wire net work over long distances. I found that the most
> useful tip that they gave was to wire in a 100-ohm resistor to your 1-wire
> network. If you have the problem where you can get signals from
> iButtons/1-wire chips from the end of your net but not from the
> beginning of
> the net, you should put the 100-ohm resistor after the chips that
> are giving
> you difficulty.
> ftp://ftp.dalsemi.com/pub/auto_id/public/ulinuxgnu.zip has source that
> compiles under Linux that will speak to the 1-wire net and read data from
> the 1-wire chips. It also has software that does the temperature
> Now, what is still puzzling to me is that the Dallas supplied software has
> no troubles communicating to the chips on the 1-wire net, yet misterhouse
> seems to not be able to communicate to those chips. If I try to
> list all of
> the chips on the net, misterhouse hangs on me. If I try listing the
> temperature for one chip, it gives me an error.
> Any suggestions? I'm still considering a total rewrite of the
> iButton::Connection package but I would obviously like to avoid
> that if all
> that is needed is a minor fix.
> > -----Original Message-----
> > From: Paulsen, Brian
> > Sent: Tuesday, December 26, 2000 11:13 AM
> > To: 'misterhouse-users@...'
> > Subject: [misterhouse-users] iButton/1-wire questions
> > Having spent Christmas weekend trying to get my 1-wire net up and going,
> > I've got a few questions.
> > 1) mh seems to work real well when I had the iButtons connected
> to a short
> > test network. When I connected the DS9097U to my Cat-3 phone line
> > network,
> > mh hangs when I try listing the iButtons. Is this normal?
> > 2) From looking at the Dallas Semiconductor website, they have
> a few tools
> > that seem extremely useful. For one, they have some cross-platform
> > software
> > (with source!) that will read all of the iButtons/1-wire chips
> and report
> > the status. On the same network that mh hangs on, the Dallas supplied
> > software works. I wonder if some new mods need to be made of the perl
> > modules or if we should do a new port from scratch. I will
> note that the
> > Dallas supplied source code looks extremely different from the perl
> > modules.
> > 3) Dallas Semiconductor also has a number of pdf files about how to deal
> > with large networks and it appears that this involves properly
> > the lines. Has anybody tried this?
> > Brian
> > ________________________________________________________
> > To unsubscribe from this list, go to:
> > http://sourceforge.net/mail/?group_id=1365
> To unsubscribe from this list, go to: