On Sun, Aug 10, 2014 at 7:50 PM, Colin Reese wrote:
> I cannot comment on their durability in weather, although I can testify
> that they are well-constructed.
>
> What are the conditions that are causing failure? Moisture, immersion,
> dripping water?
The USB dongle itself is safely housed insid
Thanks,
Is that known to be more reliable?
I'd have to find somewhere in europe to buy it. Shipping from that
website is >100$.
Pedro
On Sun, Aug 10, 2014 at 7:40 PM, Colin Reese wrote:
> http://www.ibuttonlink.com/products/linkusb
>
> On 8/10/2014 11:39, Pedro Côrte-
Hi everyone,
I've now had 3 or 4 DS9490R adapters fail after bad weather. They seem
to be particularly susceptible. After I swap them out the whole
network is fine though. Does anyone have any ideas on how to protect
them or what good quality replacement USB 1wire masters there are?
Cheers,
Pedr
On Fri, Jan 3, 2014 at 1:02 AM, Colin Reese wrote:
> Good info. I'm not sure why it should choke on that. There are examples of
> sqlite for huge implementations, but I'd probably journal a log long before
> it got to that size.
If I remember correctly it didn't choke on the writing itself. It'
On Thu, Jan 2, 2014 at 11:57 PM, Colin Reese wrote:
> Looks like I wrote essentially the same thing, except using python except
> ruby and sqlite instead in MySQL.
I started with sqlite and had to change it for a real database when it
became too slow after I had a few million readings in the dat
On Thu, Jan 2, 2014 at 10:31 PM, Colin Reese wrote:
> My current approach is to separate client/database and server/database
> interactions. A client can still force a read if they wish. It's essentially
> a cache, but I use it generally for all io, not just owfs. Logging just
> automagically w
On Thu, Jan 2, 2014 at 7:49 PM, Colin Reese wrote:
> The more intelligent way would be to bring it up and push a notification to
> the server telling it that it is online and ready. Server receives
> notification, reads (or not), and replies that it is done with the node. Node
> goes back to sl
On Thu, Jan 2, 2014 at 7:12 PM, Colin Reese wrote:
> From what I've read, the series 2 are capable of mesh networking and low
> power mode, provided the unit is not a router or coordinator. I believe
> anything in a mesh that is not an end node cannot go down to a low-power
> state, by virtue o
On Fri, Dec 20, 2013 at 7:58 PM, Colin Reese wrote:
> Hello all,
>
> I've seen fragments here and there about wireless 1wire networks, but
> nothing that sticks out as developed and/or inexpensive. Anybody have
> success with IO over wireless, be it zigbee, sdr, or otherwise, to
> something like a
On Wed, Feb 27, 2013 at 1:31 AM, Paul Alfille wrote:
> What is your setup? There have been issues with recent kernel implementation
> of the w1 module. Otherwise owserver should be a well behaved program.
Blacklisting the w1 module might be a good idea.
> owserver uses standard tcp messaging, th
Today when debugging why a server had become unresponsive for a few
days I figured out it had had a "general protection fault" in the
kernel. It was caused when running owserver. My guess is that owserver
isn't really related and that I probably have bad memory or a kernel
bug but I was wondering i
Paul, any suggestion about how to get this working? What baud rate
changes if any does owfs have to do to get the 2480B to initialize?
Cheers,
Pedro
On Fri, Aug 31, 2012 at 1:40 PM, Pedro Côrte-Real wrote:
> So I've finally gotten the DS2480B wired up to a xbee board and have
> bee
On Wed, Aug 15, 2012 at 1:45 AM, Pedro Côrte-Real wrote:
> Hi Paul, thanks for looking into this. I've gotten the modules in the
> mail today and did some initial setup. I still have the DS2480B on
> order but the simple case of doing a point-to-point wireless
> replacement of a
On Wed, Aug 15, 2012 at 2:15 AM, Alex Shepherd wrote:
> I think from memory the XBee's have a temperature sensor on-board and in
> their API Driver firmware you can enable it and it just works over ZigBee
> out-of-the-box. You can also add external analog type sensors and
> reconfigure it to do th
On Wed, Aug 15, 2012 at 12:21 AM, Paul Alfille wrote:
> Hi Pedro,
Hi Paul, thanks for looking into this. I've gotten the modules in the
mail today and did some initial setup. I still have the DS2480B on
order but the simple case of doing a point-to-point wireless
replacement of a serial line seem
Hi everyone,
I've been investigating ways to create wireless sensor networks. I
read through "Building Wireless Sensor Networks: with ZigBee, XBee,
Arduino, and Processing" and although it's a bit of a Xbee marketing
book it's pretty interesting and straightforward. The projects
described there al
On Fri, Jul 15, 2011 at 8:55 PM, Pedro Côrte-Real wrote:
> On Fri, Jul 15, 2011 at 7:55 PM, Paul Alfille wrote:
>> That's great.
>> In looking through the code to figure out your difficulty, I found a way to
>> optimise the code and reduce wear on the DS2438. It will b
On Sat, Mar 10, 2012 at 6:04 PM, Pedro Côrte-Real wrote:
> 2012/2/28 Eric Vickery :
> I guess I'll have to add some temp sensors to the daisy chain port of
> each of these, or should I worry about those getting hot from the
> board as well?
I just tried this and it seems the 1
2012/2/28 Eric Vickery :
> It is the voltage regulator heating up the board, it usually stays at a
> consistent amount so you can
> subtract the "offset" and get the correct temperature. Also the DS2438 has
> self heating issues so I
> would not use the temperature from the DS2438.
>We offer the
On Fri, Jul 15, 2011 at 12:40 PM, Paul Alfille wrote:
> From the list, also look at the
> T8A http://www.embeddeddatasystems.com/T8A--8-Channel-12-bit-Analog-1-Wire-Card_p_35.html
> although I can't say I've tested it and it calls for "analog support".
I'm still looking into this. The T8A sounded
On Fri, Jul 15, 2011 at 8:30 PM, Mick Sulley wrote:
> That's it, it now says USB is enabled.
>
> I still cannot see my 1-wire network, but I guess that is a different
> problem, I will investigate further.
Did you try "-u all" and running as root? Sometimes your user won't
have enough permissions
On Fri, Jul 15, 2011 at 7:55 PM, Paul Alfille wrote:
> That's great.
> In looking through the code to figure out your difficulty, I found a way to
> optimise the code and reduce wear on the DS2438. It will be reflected in the
> next release.
> Briefly, the VAD/VDD switch is taken from the scratchp
On Fri, Jul 15, 2011 at 6:46 PM, Paul Alfille wrote:
> Not libusb-1.0, we use libusb-0.14
> Looking at the Ubuntu packages, it's
> libusb-dev
> I've toyed with moving the libusb-1.0 but it's a fair amount of work.
Ah right. I had both installed and didn't notice.
Pedro
-
On Thu, Jul 14, 2011 at 5:17 PM, Paul Alfille wrote:
> I'm having trouble recreating the error.
> I tried 8000 consecutive readings on an AAG humidity sensor. No stray
> readings.
> I'll look through your logs next.
I seem to have fixed the problem. I had a 12V power brick connected
directly to a
On Fri, Jul 15, 2011 at 6:01 PM, Mick Sulley wrote:
> I do have libusb-0.1-4 and libusb-1.0-0 installed
>
> Any ideas? I am running Ubuntu 11.04 and owfs 2.8p12
Since you're compiling you need the dev package. Install
libusb-1.0-0-dev and you should be fine. You probably know this but
you need t
On Fri, Jul 15, 2011 at 12:40 PM, Paul Alfille wrote:
> See http://owfs.org/index.php?page=voltage-measurement
> You'll need an offset, but many of the "current" sensors measure voltage in
> that range.
> With a 1.25V offset, the DS2438 would be a great choice. It will measure
> +-0.25V with .25mV
On Fri, May 27, 2011 at 11:55 AM, Pedro Côrte-Real wrote:
> The way the system works is it reads temperatures in the panel
> (outside) and on the top and bottom of the water tank. So for those I
> would need to use the built in sensors as it would be hard to access
> the spots to
On Thu, Jul 14, 2011 at 5:17 PM, Paul Alfille wrote:
> I'm having trouble recreating the error.
> I tried 8000 consecutive readings on an AAG humidity sensor. No stray
> readings.
> I'll look through your logs next.
Are you reading USB-dongle -> sensor directly? My setup is a USB
dongle connected
On Wed, Jul 13, 2011 at 11:45 AM, Paul Alfille wrote:
> Honeywell has updated and simplified their datasheets. I use the formula
> straight from the datasheet:
> http://sensing.honeywell.com/index.cfm/ci_id/142534/la_id/1/document/1/re_id/0
> Voltage output (1st order curve fit) VOUT=(VSUPPLY)(0.0
On Fri, Jul 8, 2011 at 11:34 PM, Pedro Côrte-Real wrote:
> On Fri, Jul 8, 2011 at 9:54 PM, Paul Alfille wrote:
>> Thanks for the report -- I found the error with DS2409 aux/main directories.
>
> That's great, thanks! Is there anything else I should try to debug the
> hu
On Fri, Jul 8, 2011 at 9:54 PM, Paul Alfille wrote:
> Thanks for the report -- I found the error with DS2409 aux/main directories.
That's great, thanks! Is there anything else I should try to debug the
humidity stuff? I've also just gotten a broken value for the pressure
sensor so I suspect there
On Fri, Jul 8, 2011 at 3:09 PM, Pedro Côrte-Real wrote:
> I spoke too soon. Seems like stdout was just not being flushed. It's
> now outputting readings. I should have one of the strange ones in a
> bit.
After running it for a while I got a few wrong values and put them in
a spread
On Fri, Jul 8, 2011 at 2:46 PM, Pedro Côrte-Real wrote:
> I did exactly this and got no output.
I spoke too soon. Seems like stdout was just not being flushed. It's
now outputting readings. I should have one of the strange ones in a
bit
On Fri, Jul 8, 2011 at 2:46 PM, Pedro Côrte-Real wrote:
> have just gotten another 85ºC reading from one of the DS18S20 in one of the
> Humidity/Temp hobby board sensors.
That 85C reading was exactly on startup. It seems that what is
happening is that sometimes the first reading from the
>> Possibly putting in a line line:
>> printf( "%g,%g,%g,%g\n", VAD, VDD, T, OWQ_F(owq) ) ;
>>
>> before the return statement
>> File: owfs/module/owlib/src/c/ow_2438.c
>> Subroutine: FS_HUMID_4000
>> Line 372
>>
>> // temperature compensation
>> temperature_compensation = 1.0546 - 0.00216
On Fri, Jul 8, 2011 at 11:41 AM, Paul Alfille wrote:
> Hi Pedro,
>
> I looked at the HobbyBoards site and I think the humidity sensor is HIH-4000
> based.
> http://www.hobby-boards.com/catalog/product_info.php?cPath=24&products_id=54
>
> I presume you are reading the HIH4000/humidity property.
I
Hi,
I have had a persistent bug reading from owserver that it will
sometimes give me a very wild read (e.g., 135% humidity) from a sensor
and then go right back to normal values. I've even built an outlier
removal function into my saal package. I've had that happen today even
more. My first though
Hi everyone,
I've been using owfs for over three years now to log atmospheric
information and pestering you with questions. During that time I have
been building some software around it that some of you may be
interested in. Time to give something back.
saal is an abstraction layer around sensors
Hi,
Does anyone have experience with hobby-boards.com? Are their sensors,
particularly the humidity/temp ones, of good quality? I was
considering buying a bunch of those but wanted to know if there were
any known pitfalls?
Thanks,
Pedro
--
me
>
> /Par
>
>
> 26 maj 2011 kl. 20:14 skrev Pedro Côrte-Real :
>
>> Thanks for all the suggestions. Since this needs to be connected to
>> the solar installation controller primarily and the 1wire is just a
>> reading mostly to satisfy curiosity, I'll have
Thanks for all the suggestions. Since this needs to be connected to
the solar installation controller primarily and the 1wire is just a
reading mostly to satisfy curiosity, I'll have to figure out how to
make both work. I can potentially open up the controller itself and
see if I can tap into it's
I have a water heating solar panel installation that has several
platinum resistance thermometers. Has anyone read these using 1wire
parts? They are connected to the control box of the installation.
Would it even be safe to connect another reader to it. I'm assuming
that any resistance meters would
On Sat, Jul 3, 2010 at 10:27 AM, Pedro Côrte-Real wrote:
> I can also try to pinpoint the release between p11 and p38 where this started
> to happen.
The p11 configure script hung mid-way through so I downloaded p24
which seems to work fine.
Hi,
I was trying to upgrade my owfs installation to 2.7p38 from 2.7p11 and
ran into a crash in owserver at startup. Here's a trace:
$ sudo gdb --args /opt/owfs/bin/owserver --foreground -u all
GNU gdb (GDB) 7.1-ubuntu
Copyright (C) 2010 Free Software Foundation, Inc.
License GPLv3+: GNU GPL versi
On Thu, Jan 15, 2009 at 2:06 AM, Paul Alfille wrote:
> Ok, so the current problem is that you get many "reconnects" in the log?
I've been having a similar problem with my network. It seems to have
happened after I changed where the server was and reduced the total
length. There could be something
On Sat, Jan 17, 2009 at 7:19 PM, Pedro Côrte-Real wrote:
> Anyone know how it is done?
OpenWRT itself has a build system that now includes a Makefile that
builds owfs. I've created a patch to update it to 2.7p13 and to allow
building packages without having fuse (needed for 2.4 kerne
Can someone point me to instructions on how to build owfs packages for
openwrt? I want to build it for 8.09RC1. I see that the project itself
has this Makefile:
https://dev.openwrt.org/browser/packages/utils/owfs/Makefile
Looking back at the mailing list Christian Magnusson seems to be
creating p
I use a DS9490 connected to my server to run a small network of
sensors. It tends to work fine but last Saturday when I rebooted the
server the owserver did not come up properly. After restarting it a
few times I came to the conclusion the adapter seemed dead, not
registering on the USB bus at all.
On Dec 30, 2007 8:35 PM, Pedro Côrte-Real <[EMAIL PROTECTED]> wrote:
> On Dec 30, 2007 8:29 PM, Pedro Côrte-Real <[EMAIL PROTECTED]> wrote:
> > I just checked the type attribute in owfs itself. What else do I need
> > to look at to confirm this? I'l
On Dec 30, 2007 8:29 PM, Pedro Côrte-Real <[EMAIL PROTECTED]> wrote:
> I just checked the type attribute in owfs itself. What else do I need
> to look at to confirm this? I'll try to look up the info in the
> manufacturer's datasheets. (we bought some prepackaged sensors)
On Dec 30, 2007 5:41 PM, Paul Alfille <[EMAIL PROTECTED]> wrote:
> Are you sure the humidity sensor is HIH-3610?
I just checked the type attribute in owfs itself. What else do I need
to look at to confirm this? I'll try to look up the info in the
manufacturer's datasheets. (we bought some prepacka
On Dec 30, 2007 12:05 PM, Paul Alfille <[EMAIL PROTECTED]> wrote:
> I'll test against known humidity. In theory, the device could give numbers >
> 100%. Does it seem otherwise accurate?
I've only started keeping log of values today but we've had them um on
a webpage for a few months and we've seen
On Dec 30, 2007 1:18 AM, Pedro Côrte-Real <[EMAIL PROTECTED]> wrote:
> - The first one is that owserver when in --fake mode doesn't implement
> "/uncached/". I'm using it for testing and noticed that. This is
> happening in the latest version from CVS.
Scratch
Hi everyone,
I think I've found two small bugs in owfs.
- The first one is that owserver when in --fake mode doesn't implement
"/uncached/". I'm using it for testing and noticed that. This is
happening in the latest version from CVS.
- The second one is that my humidity sensor sometimes reports
On 8/23/07, Pedro Côrte-Real <[EMAIL PROTECTED]> wrote:
> On 8/23/07, Pedro Côrte-Real <[EMAIL PROTECTED]> wrote:
> > > Try OWTAP to inspect.
> > >
> > > There is an extensive writeup in the web site.
> > > http://www.owfs.org/index.php?page=owta
On 8/23/07, Pedro Côrte-Real <[EMAIL PROTECTED]> wrote:
> > Try OWTAP to inspect.
> >
> > There is an extensive writeup in the web site.
> > http://www.owfs.org/index.php?page=owtap
> >
> > It looks like the second message is a PING message, payload = -1
&
On 8/23/07, Alfille, Paul H.,M.D. <[EMAIL PROTECTED]> wrote:
> In short, the first 24 bytes are a header that include information like error
> status, message type, version, data length, data format (temperature scale,
> etc)
Yeah, the header, this I get. What I meant was that in the uncached
rea
On 8/12/07, Pedro Côrte-Real <[EMAIL PROTECTED]> wrote:
> I'm not sure. owread outputs the right thing but so does my ruby code
> if I do a puts to a terminal since printf ignores those characters.
> I'll have to strace owread.
I've straced owread and it's rea
On 8/12/07, Paul Alfille <[EMAIL PROTECTED]> wrote:
> I presume that when you try one of the other programs (owread, for example)
> you get the data you expect?
> I.e. the question is about the ruby code, not the sensor?
I'm not sure. owread outputs the right thing but so does my ruby code
if I do
On 8/12/07, Matthias Urlichs <[EMAIL PROTECTED]> wrote:
> * check that you're actually using the offset+payload_length fields
> instead of blindly using the whole data portion of the message ;-)
I'm truncating using the length I assumed the offset didn't actually
do anything. My Ruby code is tra
On 8/12/07, Chris Malton <[EMAIL PROTECTED]> wrote:
> You may be seeing the Ruby Binding printing out the protocol header.
>
> The owfs header looks something around this.
It can't really be that because it only happens when reading uncached
and the code does one 24 byte read for the header and th
I'm trying to read the presure from a sensor using owserver. I'm
reading this from:
/12.CC883400/TAI8570/pressure
And I get something like:
" 994.431"
when it's reading from the cache. But when I use uncached or the
server actually reads from the device I get:
"\000\000\000\000\000\000
On 3/4/07, Jan Kandziora <[EMAIL PROTECTED]> wrote:
> Cat6 may have different wire gauge. As the cable length is limited by wire
> gauge, you'll have to be more specific.
I don't have the exact same cable here with me but a similar one has
.3mm wires. What kind of range should I expect from that?
On 3/3/07, Jan Kandziora <[EMAIL PROTECTED]> wrote:
> Am Samstag, 3. März 2007 21:14 schrieb Pedro Côrte-Real:
> > Do I need to power the 18B20 with Vcc for this distance?
> >
> No. Power requirement of the DS18B20 is max. 1.5mA, which causes only a
> voltage drop in th
On 3/3/07, George M. Zouganelis <[EMAIL PROTECTED]> wrote:
> what kind of cable did you use, flat telephone, shielded or unshielded
> twisted AWG24? Have you checked for parasitic noise sources near it?
>
> once I tried a 'T' topology with unshielded AWG24s in a noise free
> environment and I got s
On 3/3/07, Pedro Côrte-Real <[EMAIL PROTECTED]> wrote:
> On 3/3/07, Paul Alfille <[EMAIL PROTECTED]> wrote:
> > Are you certain that the extension cable doesn't reverse polarity?
>
> I think not. I've tried with a cable I did myself and one I bought a
On 3/3/07, Paul Alfille <[EMAIL PROTECTED]> wrote:
> Are you certain that the extension cable doesn't reverse polarity?
I think not. I've tried with a cable I did myself and one I bought and
I've used a cable tester on them. Everything seems to check out.
Pedro.
-
I'm trying to use a 18B20 sensor connected to a USB dongle. I've put
the 18B20 sensor directly on an RJ-11 connector. I've connected the
18B20 Vcc connector to the USB Gnd. When I plug it directly to the
dongle it works fine. If I put it on a 12 meter cable it stops
working. Do I need to power the
On 2/12/07, Pedro Côrte-Real <[EMAIL PROTECTED]> wrote:
> If no one objects I'll put this in CVS tonight.
It's in CVS now. Please test it. I'm going to be out until next week
so I won't be able to r
On 2/13/07, Paul Alfille <[EMAIL PROTECTED]> wrote:
> Hmm,
>
> We're not testing the owserver, just the clients. How about write returns an
> error if you don't write the set value?
We're not testing the server but we need to test if the writes
actually do something. Returning an error should be e
On 2/13/07, Paul Alfille <[EMAIL PROTECTED]> wrote:
> Pedro, your note suggests the answer!
>
> While --fake is good for random responses, we want predictable responses to
> test against.
>
> How about --tester as a simulated adapter.
> 1. It would return known devices
> 2. It would return kno
On 2/12/07, Pedro Côrte-Real <[EMAIL PROTECTED]> wrote:
If no one objects I'll put this in CVS tonight.
I forgot to attach the code. Here it is.
Pedro.
connection.rb
Description: application/ruby
-
Using Tomc
On 2/12/07, Pedro Côrte-Real <[EMAIL PROTECTED]> wrote:
> Smaller stuff:
> - Abstract the messages into a class of their own so that instead of
> pack/unpack we can do "msg.flags = whatever". Should make the code
> prettier and probably simpler.
This is done now
On 2/9/07, Pedro Côrte-Real <[EMAIL PROTECTED]> wrote:
I wanted to use owfs in a Ruby project so I decided to port the Python
ownet client to Ruby.
Attached is a more or less straight port of the Python version of
ownet. I don't know if someone wants to check this first before I
On 2/11/07, Paul Alfille <[EMAIL PROTECTED]> wrote:
> New documentation on persistent connections:
> http://www.owfs.org/index.php?page=persistent-connections
(...)
> Yes, owserver supports persistent connections, but they must be REQUESTED
> and GRANTED. (It's only a single bit to signal this).
I wanted to use owfs in a Ruby project so I decided to port the Python
ownet client to Ruby.
While doing the port I found that Connection.read contains the following code:
while 1:
data = s.recv(24)
if len(data) is not 24:
raise exShortRead
re
On 2/10/07, Peter Kropf <[EMAIL PROTECTED]> wrote:
> Hi Pedro -
>
> I'm (slowly) in the process of implementing the changes needed for
> persistent tcp support in the ownet Python module. It'll take me a few
> more days but I'll see about incorporating your suggestions into the
> code also.
Great.
On 2/10/07, Paul Alfille <[EMAIL PROTECTED]> wrote:
> Persistent tcp connections are a new optional feature. There has been a
> recent string that explains them.
Ok, I'll look into it.
> And we'd love to have the ruby code included!
Cool. I'll clean it up and submit it.
Pedro.
On 2/9/07, Pedro Côrte-Real <[EMAIL PROTECTED]> wrote:
> I wanted to use owfs in a Ruby project so I decided to port the Python
> ownet client to Ruby.
I've just about finished this. I have a general question about ownet.
The python client creates a new TCP connection for eac
On 2/9/07, Pedro Côrte-Real <[EMAIL PROTECTED]> wrote:
> In __init__,py there's this piece of code (line 379):
>
> namelist = ','.join(list)
> #print 'Sensor.sensors namelist(%s)' % str(namelist)
> for branch_entry in namelist.split(',')
On 2/9/07, Pedro Côrte-Real <[EMAIL PROTECTED]> wrote:
> I wanted to use owfs in a Ruby project so I decided to port the Python
> ownet client to Ruby.
In __init__,py there's this piece of code (line 379):
namelist = ','.join(list)
#print 'Sensor.sensors n
81 matches
Mail list logo
