tdtool status / temp.hum sensor values?

Moderator: Telldus

Sprocki62
Posts: 7
Joined: Thu Jan 18, 2018 11:08 am

tdtool status / temp.hum sensor values?

Post by Sprocki62 » Thu Jan 18, 2018 11:08 am

Hi folks,
Happy New Year to all of you!

I'm using a Tellstick DUO on a Linux system.
Maybe I'm too stupid to understand the last two lines of this output:

jm@ubuntu10:/var/log$ tdtool -l
Number of devices: 9
1 Kleiner Flur OFF
2 Wohnzimmer Schrank OFF
3 Wohnzimmer Couch1 OFF
4 Wohnzimmer Couch2 OFF
5 Wohnzimmer Couch3 OFF
6 Wohnzimmer Fenster OFF
7 Balkon OFF
8 Buero OFF
9 Esszimmer Fenster OFF


SENSORS:

PROTOCOL MODEL ID TEMP HUMIDITY LAST UPDATED
fineoffset temperaturehumidity 255 -20.4° 255% 2013-01-02 14:17:08

jm@ubuntu10:/var/log$


Does that mean that the sensor with the "fineoffset" protocol is not supported?
The values are always the same although the time under "LAST UPDATED" is frequently updated when the DUO is blinking in blue...

Any idea?
Thanks in advance for your help
Jonas

jkp
Posts: 42
Joined: Thu Jan 18, 2018 11:08 am

Re: tdtool status / temp.hum sensor values?

Post by jkp » Thu Jan 18, 2018 11:08 am

Sprocki62 wrote:Hi folks,
Happy New Year to all of you!

I'm using a Tellstick DUO on a Linux system.
Maybe I'm too stupid to understand the last two lines of this output:

...

PROTOCOL MODEL ID TEMP HUMIDITY LAST UPDATED
fineoffset temperaturehumidity 255 -20.4° 255% 2013-01-02 14:17:08

jm@ubuntu10:/var/log$[/i]

Does that mean that the sensor with the "fineoffset" protocol is not supported?
The values are always the same although the time under "LAST UPDATED" is frequently updated when the DUO is blinking in blue...

Any idea?
I have the same situation with a cheap Clas Ohlson -bought weather station. I gather it means that the protocol is not really fineoffset, but the start data is similar or the same, leading to data being misrecognized as fineoffset. Seems the data is all ones (bytes of 0xff), leading to the bogus values seen above. I've also debugged the data somewhat, and does seem like there are several bytes of 0xFF in the raw data. I'll probably try to debug this some more, don't know if I can make progress. Would help if there'd be some kind of raw mode - that would require changing the firmware. I have compiled custom firmware but still somewhat uncertain of the bricking risk.

jkp
Posts: 42
Joined: Thu Jan 18, 2018 11:08 am

Re: tdtool status / temp.hum sensor values?

Post by jkp » Thu Jan 18, 2018 11:08 am

Hi,

An update on this - I've got now a flashing environment which is easier to use, and a better idea that there shouldn't be much bricking risk. Also, I think I'm making some progress - wrote some code to show bit sequence lengths, and looks some signal patterns stand out, which could be the real payload data.

The weather station I have is product number 36-5136, http://www.clasohlson.com/fi/V%C3%A4der ... r365136000 - CORRECTION: no it's not, it's got phase of the moon whic the 36-5136 doesn't have, and it's probably 36-4440 http://www.clasohlson.com/uk/Weather-St ... r364440000 - reads "Nexus" behind the device. Extra sensors are 36-4441, the cheapest temp & hydro or even just temp sensors I've seen, so this looks good value for money.
Last edited by jkp on Fri Jan 04, 2013 10:57 pm, edited 2 times in total.

Sprocki62
Posts: 7
Joined: Thu Jan 18, 2018 11:08 am

Re: tdtool status / temp.hum sensor values?

Post by Sprocki62 » Thu Jan 18, 2018 11:08 am

Hi jkp,
sounds good... :-) Please let me know about your progress.. I'm also uncertain updating the DUO with a new self compiled firmware.

Thanks + best regards
Jonas

jkp
Posts: 42
Joined: Thu Jan 18, 2018 11:08 am

Re: tdtool status / temp.hum sensor values?

Post by jkp » Thu Jan 18, 2018 11:08 am

Some more progress, the humidity reading seems correct now, temperature somewhat off - when the station shows 0.3, my code shows 0.5, when the station shows 0.4, my code shows 0.7. So perhaps there's e.g. a negative value conversion fencepost error by 0.1 and a divider of 2 needing to be added.

fineoffset temperaturehumidity 99 -0.5° 86% 2013-01-04 23:46:49

What I currently have interepreted as (at least most) of the payload is 10 (ten) bytes:


data:5A 9F 1F FB EA D4 F8 FF DF 56

Starting numbering from 0 and ending in 9, last (byte 9) is humidity in hex, bytes 2-3 seem to be the temperature, don't know about the rest. (the 99 for id is made up, not from the data).

Hmm, bytes 7-8 also might be temperature (juding from the sign, when temp went negative they changed from mostly zeros to mostly ones) but differently coded.

Will need to add id recognition if available, checksum checking if available. And add code to differentiate between fineoffset & this one, and figure a reasonable name for the protocol.

jkp
Posts: 42
Joined: Thu Jan 18, 2018 11:08 am

Re: tdtool status / temp.hum sensor values?

Post by jkp » Thu Jan 18, 2018 11:08 am

The last three bytes are FF DF 56 (in hex), and last byte is 0x56. Looks like the first tree hex digits of those are the temperature quite simply coded (0xFFD = -0.3 degrees, 0x000 = 0 degrees, 0x001 = 0.1 degrees etc.) Seems to work for all measurements I've made so far, which are in the limited area 0.1 .. -0.5.

Update: Either there was a radio error or the algorithm is not that simple - got -7.0 shown as temp as the temp went a bit down after -0.6.
Update2: Appears to have been a radio error, now -0.7 is shown fine. CRC or the like would be handy to prevent errors like this. BTW the sensors show somewhere between 0.5-1 degrees lower values than other thermometers around.

So, looks like now I can read both temp & humidity.

About the device - I said it's clas ohlson's 36-5136 based on appearance, but not sure, maybe it's not - looks quite like 36-4440 http://www.clasohlson.com/uk/Weather-St ... r364440000 - reads "Nexus" behind the device.

The sensor looks like 36-4441, http://www.clasohlson.com/uk/Extra-Temp ... r364441000 - manufacturer's number is said to be IW001T, and is said to be compatible with 36-4440, 36-4742, 36-4743 and 36-5056.

Edit: Looked closer at the device I have, and it shows the phase of the moon so it isn't 36-5136 (which is branded prologue), but probably 36-4440.

jkp
Posts: 42
Joined: Thu Jan 18, 2018 11:08 am

Re: tdtool status / temp.hum sensor values?

Post by jkp » Thu Jan 18, 2018 11:08 am


jkp
Posts: 42
Joined: Thu Jan 18, 2018 11:08 am

Re: tdtool status / temp.hum sensor values?

Post by jkp » Thu Jan 18, 2018 11:08 am

A version of the patch to firmware is at http://developer.telldus.com/ticket/90#comment:3 - heavily experimental so should not be really used by end users (and requires also a modified telldus-core to actually work), only for educational purposes at the moment.

Basically what seems to be different in the protocol from fineoffset is two things

* the pulse lengths for demarkation of zeros and ones are bigger
* the decoded byte data is in somewhat of a different order (this is taken care of in telldus-core)

Then again, might be I'm misinterpreting the pulse length issue, and incorrect pulse decoding might lead to some garbage data bytes (even though at least some data bytes clearly are valid), so some more research should be done before wider use.

By the way, in addition to the temp&humidity sensor I was aiming for, I seem to be getting data from some other sensor(s?) with my code - could be my code mistakenly inteprets Oregon data, as with this code I don't see any Oregon values which I occasionally saw with previous code.

hex data:9D4F8FF6F4EA7C7FB7A7
hex data:9D4F8FF6F4EA7C7FB7A7
hex data:9B4F8FF6F4DA7C7FB7A6
hex data:5D3E3FCBD2E9F1FE5E97
hex data:974F8FF3F4BA7C7F9FA5

jkp
Posts: 42
Joined: Thu Jan 18, 2018 11:08 am

Re: tdtool status / temp.hum sensor values?

Post by jkp » Thu Jan 18, 2018 11:08 am

jkp wrote: By the way, in addition to the temp&humidity sensor I was aiming for, I seem to be getting data from some other sensor(s?) with my code - could be my code mistakenly inteprets Oregon data, as with this code I don't see any Oregon values which I occasionally saw with previous code.

hex data:9D4F8FF6F4EA7C7FB7A7
hex data:9D4F8FF6F4EA7C7FB7A7
hex data:9B4F8FF6F4DA7C7FB7A6
hex data:5D3E3FCBD2E9F1FE5E97
hex data:974F8FF3F4BA7C7F9FA5
A couple of additions to this:

* The Oregon sensor does occasionally show up also with this code, so it's not that
* The unrecognized data looks like it's the same payload data which is recognized by the sensor, but with timing (start/stop) recognized differently. This (also recognizing that the error seems systematic, multiple not recognized readings followed by mutiple correct readings) suggests to me that the start recognition for this sensor type is incorrect and should be improved on.

hex data:9D4F8FF6F4EA7C7FB7A7
hex data:3A9F1FEBE9D4F8FF5F4Eclass:sensor;protocol:fineoffset;id:99;model:temperaturehumidity;humidity:78;temp:-1.1;publishData +R


hex data:3A9F1F6BE9D4F8FB5F4EFAFAFAFAFBFAF9FBFA030F202020200F101F200Fclass:sensor;protocol:fineoffset;id:99;model:temperaturehumidity;humidity:78;temp:-7.5;publishData +R

folkeorg
Posts: 201
Joined: Thu Jan 18, 2018 11:08 am

Re: tdtool status / temp.hum sensor values?

Post by folkeorg » Thu Jan 18, 2018 11:08 am

My DUO also shows these values. I have OS WMR928N with extra THGR238N that is supposed to be supported. Now, I haven't upgraded my DUO-FW yet so I guess that is why I can't find it in the list. I have, however, another device in the list with ID:0, Temp: 0C and RH: 0%. Maybe that is some of the other sensors to the weatherstation that is showing up?

My problem is that the WMR928N keeps loosing connection to the 238-sensor. It works a few hours or days but then contact is lost and never regained again. So maybe I can read the temperature via Tellstick DUO instead.

I'll follow this thread from now on.. :-)

jkp
Posts: 42
Joined: Thu Jan 18, 2018 11:08 am

Re: tdtool status / temp.hum sensor values?

Post by jkp » Thu Jan 18, 2018 11:08 am

There's some ideas for protocol debugging (with a modified firmware) / adding new protocol support in the developer area at ticket http://developer.telldus.com/ticket/90

folkeorg
Posts: 201
Joined: Thu Jan 18, 2018 11:08 am

Re: tdtool status / temp.hum sensor values?

Post by folkeorg » Thu Jan 18, 2018 11:08 am

Now that I changed FW in DUO to ver.5 I get "real" values but still not the same as the weather station SW states.

SENSORS:

PROTOCOL MODEL ID TEMP HUMIDITY LAST UPDATED
mandolyn temperaturehumidity 11 1.5 86% 2013-01-08 08:37:30
mandolyn temperaturehumidity 1 -50.0 0% 2013-01-08 00:00:05
oregon 1A2D 71 4.1 2013-01-08 03:25:31

I have OS WMR928N w wind meter, windspeed, rain meter, outdoor temp/RH, indoor temp/RH + extra THGR238N. They show according to Ambient SW:
Out: -0,8C / 91%
In: +9,7C / 26%
Extra: +4,2C / 88% (can this be 1A2D?)

What is "mandolyn"?


How often should they update? The OS-sensors send about every 45 seconds or so but only ID 11 is regurlarly updated..

jkp
Posts: 42
Joined: Thu Jan 18, 2018 11:08 am

Re: tdtool status / temp.hum sensor values?

Post by jkp » Thu Jan 18, 2018 11:08 am

Hmm - looking at the time stamp at the Oregon reading, I wonder if that could be bad reception or a badly-functioning sensor. There's a Mandolyn sensor reading at time 8.37, but the Oregon sensor reading is more than five hours older. I gather the readings are supposed to be updated whenever the sensor sends, in other words at the same time as in the weather station probably.

Could the Mandolyn reading be from some other sensor, not yours but a neighbour's?

Mandolyn is a protocol for which support apparently was added in firmware v5. E.g. Clas Ohlson sells humidity & temp sensors sending in the Mandolyn protocol, and some generally used home inexpensive wireless thermometers seem to use this protocol. I also see an id 11 Mandolyn reading which I don't think is mine but from the neighbourhood, doesn't have humidity though.

I suspect the other (id 1) Mandolyn reading is just a corrupted reading from the "11" sensor, it's more than 8 hours old.

folkeorg
Posts: 201
Joined: Thu Jan 18, 2018 11:08 am

Re: tdtool status / temp.hum sensor values?

Post by folkeorg » Thu Jan 18, 2018 11:08 am

Well, now that you mention it, I do have an old inexpencive offline weather station from CO that I used for testing temperature in the attict. It is still there and functioning I guess. So that is probably the "mandolyn" I see. My nearest neighbour is at least 100m away and noone lives there anymore.

The 238-sensor have updated since last night but that is still not visible on the DUO. Weatherstation base unit and DUO are closely located so they should receive at the same time.

folkeorg
Posts: 201
Joined: Thu Jan 18, 2018 11:08 am

Re: tdtool status / temp.hum sensor values?

Post by folkeorg » Thu Jan 18, 2018 11:08 am

It sems that there are some more sensors entering the arena, what is "fineoffset"? Do these and the -50C-mandolyn show up due to protocol error? That the radiosignal is corrupted or misinterpreted?
And why is the no humidity from the Oregon sensor?


SENSORS:

PROTOCOL MODEL ID TEMP HUMIDITY LAST UPDATED

mandolyn temperaturehumidity 11 -12.0 73% 2013-01-13 23:24:04

mandolyn temperaturehumidity 1 -50.0 0% 2013-01-09 14:18:24

oregon 1A2D 71 3.7 2013-01-13 23:23:34

fineoffset temperaturehumidity 0 0.0 0% 2013-01-13 16:15:05

fineoffset temperaturehumidity 255 0.0 0% 2013-01-11 19:33:12

fineoffset temperaturehumidity 254 0.0 0% 2013-01-10 20:23:17

Post Reply