Photos of the insides of Fine Offset sensors.

[kayedee]

Does anybody know what the small shiny metal tubular component alongside the DCF77 timer receiver is, looks like a capacitor to me, but it has no identification markings at all on it.
I was getting very low humidity readings and also lost communication with the DCF, so took the transmitter apart and found one leg of the aforementioned broken off.
Hoping someone out there will know.

[AllyCat]

Hi,

I'm not sure which photo you're referring to, but if it's a tube less than 2mm in diameter, with no polarity markings and both wires coming out of the same end, then it's almost certainly a 32.768kHz "Watch" crystal. Accurate timing between the transmissions (every 48 seconds) is essential for good operation of the system.

They're "always" 32.768 kHz and normally cost just a few pennies. But if you can't find an economical supplier, they can often be salvaged from a cheap battery-operated clock. This thread discusses some issues with such a cyrstal in the WH3081, but sadly the pictures now appear to be missing.

Cheers, Alan.

[kayedee]

Alan,
I apologise for belated reply, somehow I did not get notified of new comment on the topic, (will check my profile).
Yes that is the one, thank you for knowing, so as it is a crystal to maintain TX timing at 48 secs, is not the reason I am no longer receiving the DCF77 signal then, strange why I have no trouble receiving the 868Mhz signals.
Any ideas?

[voxit]

It must be the crystal for the DCF77 receiver,

http://www.conrad.com/ce/en/product/168432/Quartz-for-general-applications-TC-26-Frequency-775-kHz-Type-TC-26-x-H-31-mm-x-83-mm

[AllyCat]

Hi,

Yes voxit seems to be correct (and welcome to the forum).

The Transmitters do seem to use a 32.768 kHz "watch" crystal for timing, but it's "hidden" on the far side of the RJ11 connectors on the transmitter PCB(s) I've examined (sometimes also two empty holes with a symbol can be seen on the PCB).

So yes, the one by the DCF receiver is presumably a 77.5 kHz crystal (or as required for the country of use), see for example this pdf.

Cheers, Alan.

[kayedee]

Thank you both for your replies, I will order a 77.5 Khz crystal and a replacement Humirel HS 1101 sometime and solder them in.

[Philip]

Kayedee

77.5 Khz crystal? is that for the German version of the receiver?

Philip

[kayedee]

Philip,
No it would apply to all European sold devices that have the DCF77 radio clock receiver.

[dio45rus]

Hi all!
My WH-2081 lose outdoor hum and temperature values, but wind speed, direction and raindrops is still alive. Any suggestions whats wrong with transmitter?

[AllyCat]

Hi,

As far as I know, on "later" models both the temperature and humidity are measured by the "Humidity" sensor module (not by a thermistor as shown in those early photos). I don't believe anybody has found a source of separate (new) replacement modules, but somebody did report success in swapping the modules between their transmitter and Console (so then the external but not internal temp/humidy was working).

Replacement transmitters are available, but it can be difficult to identify one with the correct frequency/protocol for a particular Console. However, one might be "cannibalised" to repair the original transmitter.

Cheers, Alan.

[dio45rus]

Hi,

As far as I know, on "later" models both the temperature and humidity are measured by the "Humidity" sensor module (not by a thermistor as shown in those early photos). I don't believe anybody has found a source of separate (new) replacement modules, but somebody did report success in swapping the modules between their transmitter and Console (so then the external but not internal temp/humidy was working).

Thanks, Alan! This "Humidity" sensor module seems like simply DHT-11 sensor, available on most radiojunks.


One more plus to this idea that data, as I clarified, is recieved but is wrong (-3276oC and 255% humidity)

Replacement transmitters are available, but it can be difficult to identify one with the correct frequency/protocol for a particular Console. However, one might be "cannibalised" to repair the original transmitter.

Cheers, Alan.

Yes, its available on AliExpress, for example. But buyng is not "gentlemens way". More intresting repair this item.

[steve]

One more plus to this idea that data, as I clarified, is recieved but is wrong (-3276oC and 255% humidity)

Those are just the "no data" values- all FF. It's odd that you're seeing those values, as Cumulus would normally look for those and ignore them, and the console would normally display "dashes".

[killah78]

Hey Alan,

hope this topic is not too old. I am interested in buying such a WH3080 and to decode the signal via Raspberry.
Found your post that your WH3080 sends via OOK.
Now I found out that there was a model change in 2014(?). I guess like with the WH1080 there is a change from OOK to FSK (or other way round).
My question is, do you still own this weather sensor?
If yes, could you check the battery-case if you will find a stick with a "pass"-code? As far as I know the current model is pass18.

That would be very helpful finding the right version.

Thanks a lot.
killah78

Hi,

Welcome to the forum.

Yes, the WH3080 (or at least my version) transmits three different messages (IIRC of 8, 10 and 11 bytes) of a similar format but with different addresses. The normal message (external temp, humidity, rain and wind) and the RCC message are the same as the documented (OOK) messages for the 1080 (the RCC message replaces the "normal" message for a few minutes after the hour).

The Solar message is transmitted once each minute, generated within the "Pod" but passed to the transmitter via the 6-core cable (so it's possible to monitor/test it without a radio receiver). It contains the Lux in three bytes (a pure binary 24-bit number) in tenths of a Lux and the UV in pure binary in one byte (nybble), but I'm not sure of the exact contents of data bytes 2 and 3. The following is part of the header code from my (unfinished) PICaxe decoder:

; Fine Offset (Solar Data) message BYTE offsets (within receive buffer)
; Examples= FF 75 B0 55 00 97 8E 0E *CRC*OK*
; =FF 75 B0 55 00 8F BE 92 *CRC*OK*
symbol FOrunio = 0 ; Fine Offset Runin byte = FF
symbol FOsaddo = 1 ; Solar Pod address word
symbol FOuvo = 3 ; UV data nybble ?
symbol FOluxHo = 4 ; Lux High byte (full scale = 4,000,000?)
symbol FOluxMo = 5 ; Lux Middle byte
symbol FOluxLo = 6 ; Lux Low byte, Unit = 0.1 Lux (binary)
symbol FOcksumo= 7 ; CRC checksum (CRC-8 shifting left)

Cheers, Alan.

[AllyCat]

Hi,

Welcome to the forum. Yes, my WH3080 transmitter does indeed have a "pass 18" sticker, but I'm not sure that's very informative because AFAIK there have not been any issues or changes to the transmitter. The Solar Data (WH308x) stations have always used the OOK format and later the basic WH108x stations were changed from the earlier, more expensive, FSK to cheaper OOK. It appears the 433 MHz versions may have changed first (~2012), then the 868 MHz, and somebody in USA recently still received a 915 MHz FSK (but maybe it was "old stock").

The "issues" (which may not have been fully fixed even now) were "lockups" of the Solar Pod (microcontroller), and USB lockups of the Console USB PC interface (not fixed and apparently now ported over to the WH108x stations). Certainly the Solar Pods have been "improved" and I believe FO changed the model number of later stations to WH3083, but I've never had one to test.

Cheers, Alan.

PS: Looking at my data format extract above, I think only 6 bits of the High Lux byte are actually used (i.e. 22 bits in total), giving a maximum value of ~ 400,000 Lux.

[killah78]

Hi Alan,
thanks for your reply. I'm still trying to even receive a signal from a WH3080 station. Unfortunately it is not my own, so it's complicated to search through the logs and filter out wrong receipts. But when this will run, I plan to buy an own one.

Perhaps you can give me some hints.
1. As you say you have the pass18-sticker: Does your station send via 433Mhz? Do know if there are WH3080 stations sending via 868Mhz? In several blogs I can read 433 while in others they mention to receive via 868Mhz. Was there a change in the frequency? Or are there differences in different brands?

2. When I demodulate the signal, is this "compatible" with the 1080? At the moment I am looking for these pulselengths:
1 = 400 - 621 us
Pause = 805 - 1150 us
0 = 1495 - 1840 us
Can you confirm these lenght?


Thanks for your anwer.
Regards,
killah78