Cumulus Support

Yesterday I noticed flat lined graphs and had a look at my WH-1081 Console to find that the display was showing random segments and no comms with the computer.
This morning I took the console apart, cleaned the board and removed a couple of solder splashes but could see nothing obviously wrong.
Re-assembled and when booting the console, most of the segments turned on as they should but then the display returned to nonsense with only a few recognised numbers, and no reponse to touch at all. The processor is not playing nice!
To use colloquial slang....I think it's sucked a Kumara.
Anyone have any experience as to what could be wrong?
Console is about 2.5 years old so maybe this is the life of these Fine Offset Stations.

Sounds like it might be that strip of rubbery stuff that connects the cct board to the display. There's trouble if you don't get it back in exactly the right place. Watch out for dirt or other contamination too. And make sure you get it the right way round. I had problems with that when I took my first one apart.

Another possibility is bad solder joints - another thing I had trouble with. Quality control? What quality control?!!!

The rubber strip stayed in place when I dismantled and the segments do seem to do the startup test but yes...tricky to align and you need to get enough downward pressure on the strip.
Couldn't see any cold joints but I might just put on the magnifying googles and go over it with a hot iron...nothing to lose if it's not going anyway. I'm pretty sure the processor is running but get no comms to Cumulus either.
It was a bit strange because it just turned up it's toes and stopped....unusual in this sort of equipment.
These stations seem to be a lot cheaper in the UK and maybe I should get the latest Solar one and get my cousin to freight it to NZ.

Problem solved. Another case of bad solder joint somewhere. I went over the whole board and reheated the joints, plugged it back in and it started up OK. Now I just have to change the settings to what they were before.
I hate soldering SMD's even with high magnification

[edit]
Spoke too soon.
After a short time I get -159C indoor and Rel Pressure of 412.5hpa. Figures are ok after initial boot. Needs more investigation before hooking up to Cumulus again. Maybe I will resolder the internal sensors also....didn't do them before.
I also have the screen blanks every 48 secs and then comes back....something to do with the transmission I think.

Super-T wrote:Spoke too soon.
After a short time I get -159C indoor and Rel Pressure of 412.5hpa. Figures are ok after initial boot. Needs more investigation before hooking up to Cumulus again. Maybe I will resolder the internal sensors also....didn't do them before.
I also have the screen blanks every 48 secs and then comes back....something to do with the transmission I think.

Is it a coincedence that the unit updates every 48 seconds as well. Could be something loose (maybe).

Seems really odd.

I've had the readings from the outside sensors vanish on 48 seconds and then reappear - i think due to the thickness of my walls and the transmitter range probably being very close to the maximum, but the whole screen to blank doesn't make much sense

Leaves me wondering whether on transmisison receipt, the processor is drawing too much power causing the LCD to fade

Daft question and I'm sure you have - worth trying fresh batteries in the receiver/display console?

When connected to the computer, the USB supplies power to the console - the batteries are disconnected by a relay. You can hear it click when you plug in or unplug the USB.

I have no explanation for your problem, I'm afraid, apart from a dry joint or bad connection.

try replacing the 3 elco's.
most common problem of using low quality elco's to produce as cheap as possible.

All fixed and running for now.
Bad joint somewhere. I went over the sensors with a soldering iron and they came back up.
Have had some of "Gina's Problems" with negative and/or positive spikes on the external temperature.
Didn't have to change the elco's but that would have been next in line. I keep a good supply of capacitors for replacing in equipment around 5 or six years old. Starting to get lots of LCD monitors showing up with caps problems. Still the odd motherboard and power supply and even DVD/Video players. At one stage I had bad caps in the Dishwasher Power supply. Bad caps cause so many problems.
Must be my time for electrical problems. Yesterday the washing machine died. Took it apart and there was an "exploded" transistor. Had to get the whole replacement unit as I had no way of knowing the faulty device ID. About to plug it all back in and cross fingers I don't get another spike

After my problems with the Fine Offset Console I lost about 36 hours of data and used DaveNZ's data as he is close by. Thanks Dave!

I then had to do a few corrections but one thing I did notice is that if I deleted the today.ini then Cumulus seemed to rebuild it. Then it occurred to me that perhaps an enhancement later where if the other ini files in \data are deleted then Cumulus would go to a routine of rebuilding them from the monthly files. I understand this could take a while but it could save heaps of time struggling through data.....let the computer do the work. Perhaps there could be choices as to what date to rebuild from?

Just a thought.

Super-T wrote:Then it occurred to me that perhaps an enhancement later where if the other ini files in \data are deleted then Cumulus would go to a routine of rebuilding them from the monthly files. I understand this could take a while but it could save heaps of time struggling through data.....let the computer do the work.

As it happens, this same thing was going through my mind this morning. The current situation is that the month.ini and year.ini files are created from the highs and lows in dayfile.txt if they don't exist. But as you point out, there's no option to build them from the monthly files as well/instead, and often dayfile.txt doesn't actually contain a particular high and low as at any time in the past (or even now) some items may not be in there (e.g. dew point). And there's no option at all to rebuild the alltime.ini file.

I was contemplating adding some more highs and lows, and then started thinking about all the posts that would appear pointing out the obvious, that I have already pointed out - that they don't apply retrospectively. So then I got to thinking about a facility to trawl all of the data files. In my head, this was an extension of the alltime record editor (with new similar editors for this month and this year highs and lows) where the editor would find the corresponding item from dayfile.txt and the monthly logs, and offer the option to replace the existing record with one of the values found from the files. It would need to offer this on a 'per item' basis, because most of the time the existing entry would be correct and the values found from the log files may well not be exactly the same (because of the 'snapshot' effect), so a wholesale replacement of the existing entries would sometimes lose data.

That was when I realised how much work was involved in doing it properly and decided against adding any new highs and lows just now

That's what I find when developing software - you think of things (or others suggest them) faster than you can implement them. The ToDo list just grows and grows!

Lol....good answer Steve.
Perhaps you could whip something up over Xmas while the rest of us are enjoying ourselves
Merry Xmas to you and yours.