Hi
I have not had a chance to fire up the air compressor yet...
when ever i find time, the baby is asleep.
and now we are away for the weekend....so next week.

data history is the one major thing I need to figure out too
I can get the data, but I have not had time to decode it....
Brian

Small discovery or clarification.

Earlier I argued against the fact that temperatures in general were stored with 4 BCD digit. I thought that only one decimal was used E.g. 21.3 deg C.
I based that on the fact that the data show everywhere is with one decimal and I never saw anything else than a 0 as what would be the 2nd decimal. I see now that I was wrong. When you run the station displaying Celcius the 2nd decimal is always 0.
But when you change to Fahrenheit the 2nd decimal is actually used. The unit still stores the value in Celcius though!!. So for accuracy we should make our different softwares read both decimals. This seems to be valid for all things measured in degrees C incl dewpoint and windchill.

Kenneth

[ Diese Nachricht wurde geändert von: Lavr am 2003-04-24 12:46 ]

Hi all,
Hi Kenneth,
Could you tell me more about register 0x54C,
I can't figure how to use delay in this register.
Wind dir or speed in wired mode transmit data
very often, every 2 sec during 2 sec, or it can be
during 8 sec with sometimes pause of 8 sec.
I think it is usefull to test these data when reading memory (saved data) but iam not sure.
So what is your way of seeing that ?
thanks,

_________________
Claude Ocquidant

The address 54C seems to be used for two things.

The station uses the memory areas as a timer for the collection of data from the outdoor sensors.
It seems to calculate the new start value based on the value in address 54D.

The countdown happens in addresses 550-54F.

As I see it the PC program reads the value regularly and from the value read it can then calculate the time till next data collection point. As I see it all it uses it for is to show the "HF reception" icon read.
I have not found much connection with the station ignoring data communication. That happens also in the period between. So unless you make a program with a GUI I see no use of it.

Kenneth

I am working on decoding the history data. This is one where 3-4 brains could be better than one because I have not cracked it yet.
I have not given up either.

Here is what I know. I have 13 data sets where I both have the binary data from memory and human data from the PC program.

First the human data





Event P      Ti    Hi   To     Ho  DP      WC    Wind    Rain


#    [hPa]   [°C]  [%]  [°C]   [%] [°C]    [°C]  [m/s]   [mm]


01   1017,4  27,0  22   10,5   45  -0,9    10,5  0,4 ESE 145,0


02   1017,4  26,8  22   10,5   45  -0,9    10,5  0,0 NE  145,0


03   1017,5  26,7  36   10,5   45  -0,9    10,5  0,0 NE  145,0


04   1017,5  26,7  36   10,5   45  -0,9    10,5  0,0 NE  145,0


05   1017,5  27,2  36   10,5   45  -0,9    10,5  0,0 NE  145,0


06   1017,5  27,1  36   10,5   45  -0,9    10,5  0,0 NE  145,0


07   1017,4  27,1  50   10,3   46  -0,8     8,9  2,1 SE  145,0


08   1017,5  27,1  43   10,3   46  -0,8    10,3  0,0 SSE 145,0


09   1017,5  27,0  41   10,3   46  -0,8    10,3  0,0 SSE 145,0


10   1017,4  27,0  35   10,2   45  -1,1    10,2  0,0 SE  148,1


11   1017,4  27,1  33   10,2   45  -1,1    10,2  0,0 ESE 148,1


12   1017,4  27,1  31   10,2   45  -1,1    10,2  0,5 SSW 148,1


13   1017,4  27,1  29   10,1   45  -1,2    10,1  1,1 ENE 148,1





Then the binary data





18  17-15  14-12  11-10 9-8  7-6  5-4  3-2  1-0


5   00 4   0 FD   45    35   C0   E6   30   42


2   00 0   0 FD   45    35   C0   E6   30   40


2   00 0   0 FD   45    57   EE   F6   30   3F


2   00 0   0 FD   45    57   EE   F6   30   3F


2   00 0   0 FD   45    57   EE   F6   30   44


2   00 0   0 FD   45    57   EE   F6   30   43


6   01 5   0 FD   46    7A   1C   E6   28   73


7   00 0   0 FD   46    69   05   F6   28   73


7   00 0   0 FD   46    64   23   F6   28   72


6   00 0   1 03   45    55   7D   E6   24   8A


5   00 0   1 03   45    50   9B   E6   24   8B


9   00 5   1 03   45    4B   B9   E6   24   8B


3   00 B   1 03   45    46   D7   E6   20   A3





Or in the other direction





  Event 1   2   3   4   5   6   7   8   9  10  11  12  13


Nipples


1-0    42  40  3F  3F  44  43  73  73  72  8A  8B  8B  A3


3-2    30  30  30  30  30  30  28  28  28  24  24  24  20


5-4    E6  E6  F6  F6  F6  F6  E6  F6  F6  E6  E6  E6  E6


7-6    C0  C0  EE  EE  EE  EE  1C  05  23  7D  9B  B9  D7


9-8    35  35  57  57  57  57  7A  69  64  55  50  4B  46


11-10  45  45  45  45  45  45  46  46  46  45  45  45  45


13-12  FD  FD  FD  FD  FD  FD  FD  FD  FD  03  03  03  03


15-14  40  00  00  00  00  00  50  00  00  01  01  51  B1


17-16  00  00  00  00  00  00  01  00  00  00  00  00  00


   18   5   2   2   2   2   2   6   7   7   6   5   9   3





Here is what I have guessed so far.





Nipple 18 is wind direction. It fit 100%





17-15 is the wind speed in binary. It is m/s * 10





14-12 fits rain. It is in the same steps as the sensor counts.


  Ie. Rain total binary in steps of 0.518 mm





11-10 fits Outdoor humidity in BCD. But it seems a waste of bits.





9-8.  There is a fixed factor of 2.424 with Indoor humidity. Strange





7-6  No correlation found so far. Must be combined with other bits.





5-4  The nipple 5 changes one up and down in sync with pressure.





3-2. Follows steps of Outdoor temperature but makes no sense value-wise.


     It jumps 4 steps binary per 0.1 degree step in temperature.





1-0. Some trend with Indoor temp but then jumps at data point 7. Some bits


     must belong to other field.

I will try and get some more data picked up but give it a thought. Maybe you guys see a pattern that I don't see.

My guess is that Dewpoint and Windchill are calculated by the PC program based on the other data. Pressure seems to be very effectively packed. I could use help with that.

Kenneth


[ Diese Nachricht wurde geändert von: Lavr am 2003-04-26 18:15 ]

Hi,
Binary Data:
Right,
I also found TOut needs adjustment
TOut seems for me to be 5 low nibble and 6 high nibble.
a high nibble for Tin could be in 6
Pressure seems in 0 1 2 3 and need adjustement with another high nibble. I draw graphs and show same variations in 0 1 2 3 and particuliarly in 1 and 2 than actual pressure data.

What about nibble 19 ?
Seems a special one but exists ?
I found nothing by myself for this nibble.



_________________
Claude Ocquidant

[ Diese Nachricht wurde geändert von: anonymous am 2003-04-26 18:09 ]

Note that I count the nipples from 0 so there are 19 in total and the last was the wind direction.

I am not sure I understood what you said
QUOTE
TOut seems for me to be 5 low nibble and 6 high nibble.
a high nibble for Tin could be in 6
UNQUOTE

Can you explain that in a different way? Are you counting bytes instead of nipples?

Kenneth

Hi Kenneth,

Zitat:

Can you explain that in a different way? Are you counting bytes instead of nipples? Kenneth

Yes I read 10 Bytes that is 20 nibbles
We could read 13 bytes.
Here are 10 BYTES

Code:

AB CD EF GH IJ KL MN OP QR ST 01 23 45 67 89 11 11 11 11 11 xx xx xx xx xx 01 23 45 67 89

Hum,
Forget all that i made a mistake
Sorry
Claude

[ Diese Nachricht wurde geändert von: anonymous am 2003-04-26 23:17 ]

One more nut cracked.
I have decoded the indoor and outdoor temperature of the history memory area.

The temperature are in nipples 4,3,2,1,0.
Nipple 0 being the first nipple, 1 the next etc. Nipple 4 is the most significant 4 bits.

The whole value of the 5 nipples (20 bits) are used in the calculation for BOTH temperatures.

Tindoor = (value % 1000)/10 - 30 [C]
Toutdoor = (value - (value % 1000))/10000 - 30 [C]
Where % is the modulus operator.

I made some programming examples. In both the variable "value" is the interger value of nipples 4-0.

In Perl you can make this little program as example.

#!/usr/bin/perl





$value = 0x62873;


$ti = ($value%1000)/10-30;


$to = (($value-($value%1000))/10000)-30;


print "Ti = $ti - To = $ton";

Or in C we can take advantage of the interger division only leave the interger part of when dividing two integers.

int main(void)


{


    int value;


    float to, ti;





    value = 0x62873;


    ti = (value%1000)/10.0-30.0;


    to = (value/1000)/10.0-30.0;


    printf("Ti = %f - To = %fn",ti,to);


    return(0);


}

It was a hard nut to crack but after having made many samples writing directly to memory and reading the resulting temperatures, and after I dreamt about it at night and put it all up in an Excel sheet it became clear to me how the two were combined.

So now I am attacking pressure. We will get there. I noticed that Rain is relative to something not part of the 19 nipples of data.

Kenneth


[ Diese Nachricht wurde geändert von: Lavr am 2003-04-27 20:21 ]

Here is next step in the decoding of the history data.

Nipples 9-5 (next 20 bits) are Air Pressure and Indoor Humidity.

The formulars are similar to the ones used for the two temperatures.

Pressure= 1002.2 + (value % 10000)/10
If pressure is greater than or equal to 1502.2 then you subtract 1000.

Indoor temperature =(value-(value % 10000))/10000

Bed time. I continue tomorrow with next parameters.

Rain is a bit strange. Hints are welcome.

Kenneth

[ Diese Nachricht wurde geändert von: Lavr am 2003-04-28 00:18 ]

Next chapter of the history data.
Nipples 11-10 is Outdoor Humidity in plain human readable BCD (Binary coded decimal, one nipple per digit). Ie. 0x62 = 62%.

Kenneth

NOTE NOTE !!!. CORRECTED THE NIPPLE NUMBERS.

[ Diese Nachricht wurde geändert von: Lavr am 2003-04-29 15:07 ]

Keep up the good work Kenneth!

I have been busy adding support for the ELV WS2500PC data logger in weather display

I have it working now, just polishing off

Brian

Rain history decoded.

I got the rain figured out. Or at least enough to understand how to use the data.

Rain is stored in history record nipples 14-13-12.
The value is binary and steps 0.518 mm/step.
The absolute value does not seem to be related to anything else than an internal "household" value inside the station.

Every period the current 12-bit rain count value is stored as history data (Let us call this value for a specific history record 'RAINCOUNTn'). The PC program uses it this way.

FIRST time the PC program is started without a history data file but connected to the station the program fetch the current rain value from addresses 4D7-4D2 (BCD coded in 0.01 mm. We can call this RAINref.
At the same time it saves the rain count from the last history record (the one the pointer points to at the moment). We can call this RAINCOUNTref The PC program always just saves ONE record the first time and for rain this becomes a reference record. From then on each new history record is fetched by the PC program and the total rain is then calculated as using the RAINcount

RAINtotal = RAINref + (RAINCOUNTn - RAINCOUNTref)*0.518
Result is in mm rain.

Kenneth

Last two history items

Wind speed is nipples 17-16-15.
It is simple.
Windspeed = value in binary / 10 [m/s]

And Wind direction is also simple.
Nipple 18.

Direction = value * 22.5 degrees. North is 0 and degrees are clockwise on the circle.

Kenneth

[ Diese Nachricht wurde geändert von: Lavr am 2003-04-29 12:26 ]

Hi Kenneth

How are you getting the data history?
I am sending:
$82
$B2
174 (decimal...this increases 4 next time)
154 (decimal...this increases 12 next time)
$EA


and then I get a reesult, but the temperature in the result using your formula does not seem to work (but yet it does on some other data I captured using the heavy weather sofware),so I think I am doing something wrong!
Thanks
Brian