Welcome to the Cumulus Support forum.

Latest Cumulus MX release v3.0.0 build 3050 - 1 May 2019.
Legacy Cumulus 1 release v1.9.4 (build 1099) - 28 November 2014

See the Wiki for downloading either version.

Transmitter repair.

Discussion specific to Fine Offset and similar rebadged weather stations
Post Reply
andrewinpopayan
Posts: 38
Joined: Fri 19 Mar 2010 10:03 pm
Weather Station: N96GY
Operating System: Suse Linux 11.2
Location: Wyke, Bradford BD12, UK

Transmitter repair.

Post by andrewinpopayan » Tue 05 Jan 2016 3:51 pm

Hi all.

Has anyone successfully repaired a transmitter? The transmitter on mine flashes the LED every 48 seconds (or so) but doesn't put out any RF, I am watching an SDR radio tuned to 433.9 Mhz and have the bandwidth opened right up, nothing being sent out when the light flashes. I could get a spare for £15 from Maplins but as a *very* last resort. The RF stage appears to be an ST4021 capacitively coupled to the "antenna".

Any replies most welcome

Andrew

AllyCat
Posts: 1057
Joined: Sat 26 Feb 2011 1:58 pm
Weather Station: Fine Offset 1080/1 & 3080
Operating System: Windows XP SP3
Location: SE London

Re: Transmitter repair.

Post by AllyCat » Wed 06 Jan 2016 10:50 am

Hi Andrew,
andrewinpopayan wrote:I could get a spare for £15 from Maplins
Sorry, but no you almost certainly can't. :(

About 3 years ago Fine Offset changed the transmitter hardware and transmission/wireless protocol, so the transmitters now sold by Maplin (and AFAIK everybody else) is completely incompatible with your Console. It was very much a "cheapening excercise" because I believe the old stations used FSK whilst the new system is simple OOK. The only advantage of the new hardware/protocol is that it's much easier to hack or intercept. ;)

AFAIK there is no solution apart from trying to find a an old "second hand" transmitter. But that's difficult because there are no external differences between the versions. See this very recent thread (and there are others on the forum) and the sticky thread linked there.

If you're an "expert" in electronics you might be able to fix the transmitter. But I've never even succeeded in seeing any "wireless data" waveforms on a 'scope (with those early stations), because I believe it's all done via an internal (SPI) data bus.

Cheers, Alan.

andrewinpopayan
Posts: 38
Joined: Fri 19 Mar 2010 10:03 pm
Weather Station: N96GY
Operating System: Suse Linux 11.2
Location: Wyke, Bradford BD12, UK

Re: Transmitter repair.

Post by andrewinpopayan » Wed 06 Jan 2016 1:31 pm

Hmmm, thanks for the heads up on non compatability issue Alan.

To this point now, I have stripped the PCB out, given the board a flush with "Servisol" and scrub with toothbrush. I left it on a radiator overnight and it's now in a sealed jar that's full of dried out silica-gel. I figure if water has crept in anywhere, then the silica-gel should pull it out, if it does work again, I will paint the whole PCB (except sensors) with encapsulation epoxy resin.

I am going to have a snoop around the board with a multimeter and oscilloscope to check that the 4.00 and 10.00 Xtal oscillators are running and that everything is getting power. I also have a USB microscope, so should see anything obvious.

Andrew

AllyCat
Posts: 1057
Joined: Sat 26 Feb 2011 1:58 pm
Weather Station: Fine Offset 1080/1 & 3080
Operating System: Windows XP SP3
Location: SE London

Re: Transmitter repair.

Post by AllyCat » Wed 06 Jan 2016 3:02 pm

Hi,

At least you have the advantage of working with a "real" ic package. The majority of the Fine Offset boards use "chip on board" technology (under a blob of black epoxy) which are less reliable and impossible to repair.

I did open up one FO "transmitter" which has the same chip, but it also has a RTC antenna, so I think is the Clas Ohlson 868 MHz version (which stayed with FSK longer than Maplin, but is now also OOK). I remember probing around the board (and the corresponding receiver) but never saw any "millisecond" type signals consistent with the RF data.

The duration of the "old" transmissons is only around 50 ms (much longer with the later protocol) and I think the LED has an independent control (i.e. it's not just across the supply/enable pin of the TX). I'll be interested to hear if you manage to get it working again. But working with the later models is much easier, with the protocol documented by for example Kevin Sangeelee (who has posted on this forum) on his website here.

Cheers, Alan.

andrewinpopayan
Posts: 38
Joined: Fri 19 Mar 2010 10:03 pm
Weather Station: N96GY
Operating System: Suse Linux 11.2
Location: Wyke, Bradford BD12, UK

Re: Transmitter repair.

Post by andrewinpopayan » Wed 06 Jan 2016 3:12 pm

Image

Hopefully a picture of the board is here.

AllyCat
Posts: 1057
Joined: Sat 26 Feb 2011 1:58 pm
Weather Station: Fine Offset 1080/1 & 3080
Operating System: Windows XP SP3
Location: SE London

Re: Transmitter repair.

Post by AllyCat » Wed 06 Jan 2016 3:31 pm

Hi,

Yes mine looks very similar (but with an antenna "rod" for the RCC). I've only just noticed the "433." on the PCB near the antenna, mine says "868" although the printed antenna is identical!

I hadn't previously noticed that Kevin's web page has a link to the Raspberry Pi forum, where they have apparently hacked an "old" Clas Ohlson WH1080, which confirms the FSK and SPI bus. The thread dies about a year ago, probably when FO also changed the CO protocol.

Cheers, Alan.

andrewinpopayan
Posts: 38
Joined: Fri 19 Mar 2010 10:03 pm
Weather Station: N96GY
Operating System: Suse Linux 11.2
Location: Wyke, Bradford BD12, UK

Re: Transmitter repair.

Post by andrewinpopayan » Wed 06 Jan 2016 4:04 pm

Hi Alan.

So yours is up on 868MHz, which nowadays I find odd because OFCOM seemed to have grabbed every frequency up from 500MHz for yet MORE mobile fone space. There it is on my board 433.92, I am spoilt for receivers from 100MHz upwards to 2GHz-ish.

The board is back in the desiccating jar, I have also pulled a vacuum in there, as well as stood it on a radiator, so I can be pretty sure it isn't water causing the problems. :ugeek:

Andrew

andrewinpopayan
Posts: 38
Joined: Fri 19 Mar 2010 10:03 pm
Weather Station: N96GY
Operating System: Suse Linux 11.2
Location: Wyke, Bradford BD12, UK

Re: Transmitter repair.

Post by andrewinpopayan » Wed 06 Jan 2016 6:25 pm

Hi Alan.

The Si4021 is not a friendly little chip to investigate, however looking at the "standard" usage application notes, I can see a little bunch of SMD inductors around the RF O/P stage, candidates for the multimeter, especially L3 which supplies power to the >finals< :geek:

I have 3.12 V Vdd, the 10.0MHz Xtal is too high an impedance to measure with my frequency counter.
Image

Image

leccyrob
Posts: 2
Joined: Wed 12 Jun 2019 10:58 am
Weather Station: WS1080
Operating System: Ubuntu 14.04

Re: Transmitter repair.

Post by leccyrob » Wed 12 Jun 2019 11:12 am

Hi Guys,

I'm probably a couple years too late to post on this but came across your posts while investigating a very similar fault......
I've got a bit of experience at soldering SMD's (Surface Mounted Devices) so I had a look into my WS1080 which started failing to transmit intermittently and more and more frequently not transmitting.
Moving the transmitter within a few feet of the receiver didn't help the issue so I opened it up and found a SI4021 transmitter IC with it's 10Mhz crystal.
The PCB appeared to have a lot of gunge/hot melt glue type compound all over it so I decided to remove the Si4021 (with a hot air desolder tool) and the crystal too, cleaned up the PCB and resoldered the components.....
This appeared to do the trick and all worked well for about 3 weeks. Then it started to not transmit more and more often once more......
I've now ordered a new 10Mhz crystal and a Si4021. Will try just changing the crystal first and then the Si4021 transmitter IC if it still fails.
Will post my results back here in about a month with (hopefully good) news.

Razgriz_410
Posts: 3
Joined: Tue 25 Jun 2019 4:09 pm
Weather Station: I don't know yet, here to learn
Operating System: Win 8.1 64bit

Re: Transmitter repair.

Post by Razgriz_410 » Sun 30 Jun 2019 8:04 pm

Speaking of transmitters breaking, is it possible to have two redundant ones? The thing is that I'm going to buy one of these houses in Greece for the holidays, and install a weather station there; my plan is to have it online even when I'm not there, and thus, I'd like to have 2 redundant transmitters for the station, so that if one of them breaks, the other one keeps working and I don't find myself with an out-of-service transmitter and an useless station thousands of miles away.

leccyrob
Posts: 2
Joined: Wed 12 Jun 2019 10:58 am
Weather Station: WS1080
Operating System: Ubuntu 14.04

Re: Transmitter repair.

Post by leccyrob » Mon 01 Jul 2019 11:21 am

Hi Razgriz_410,
Funnily enough, I'm actually doing the opposite of what you're describing....
I've got one transmitter being picked up by 2 receivers. One receiver is on my coffee table where I can read things as normal. I then bought another receiver to work in a cupboard connected to a raspberry pi running various programs to enable me to view the weather stats in Home Assistant via the internet.
Knowing that this definitely does work, I would make an assumption that I probably could run 2 (or more) identical transmitters (that is to say the same frequency, protocols etc) to be picked up by a single transmitter. There could well be some strange issues arising from the rain gauge though which works out rain in the last 24 hours by subtracting the total recorded 24 hours ago from the total recorded rainfall to present. Once you get 2 different rain gauge totals hitting the same transmitter you may get confusing results.
Also, you may find that later models of the WS1080 contain other firmware which may (possibly???) work on a 'pairing' type system where you can only have one transmitter sending to one receiver.
Perhaps it might be worth you getting one transmitter/receiver set working on 868Mhz and another set working on 915Mhz (and maybe another pair running on 433Mhz), that way you'll have some completely separate/redundant systems working in parallel??????
Rob.

PS. I've just got those new Si4021 IC's from China- am still waiting for the 10Mhz crystal to show up......

AllyCat
Posts: 1057
Joined: Sat 26 Feb 2011 1:58 pm
Weather Station: Fine Offset 1080/1 & 3080
Operating System: Windows XP SP3
Location: SE London

Re: Transmitter repair.

Post by AllyCat » Tue 16 Jul 2019 9:38 am

Hi,

Sorry, rather late with this reply.
leccyrob wrote:
Mon 01 Jul 2019 11:21 am
I would make an assumption that I probably could run 2 (or more) identical transmitters (that is to say the same frequency, protocols etc) to be picked up by a single transmitter.
No, whilst FO indeed use only a "one way" transmission system ("send and hope"), there are two reasons why a single receiver cannot accept signals from two transmitters. Firstly, the FO protocol (the second byte, immediately after the FF run-in) uses an address (better called a signature) which is randomly set when the batteries are inserted. The Console "learns" this address when it is Powered up / Reset. Secondly, to save battery power, after learning the "time" of its partner transmitter the Console switches off its receiver for ~47 seconds and only "wakes up" when it expects the next transmission (48 seconds after the previous).

Note: A complete FO station may use three different "data packets" with different addresses (for T/H/Wind/Rain, Solar and RCC data) but these all emanate from the same external transmitter.

So the answer for Razgriz_410 is that you could use one transmitter with two receivers (which would give some redundancy). But the external sensors/transmitter are more likely to fail (because of the "weather" outside) than the Console ("USB Lockups" excluded), so a completely duplicate system would be better. Two "identical" (FO) stations could use the same transmitter frequency (independently), but a different frequency/protocol might be better, to avoid the "confusion" of the addresses changing when the transmitter batteries are replaced.

Cheers, Alan.

Post Reply