Sandaysoft

Hi Stephen,

Can you keep us informed how you go?

You will know when / if it fails. The console will show -- where the numbers should be and Cumulus will flat line at the last good value.

Meanwhile I have bitten the bullet and I am playing around with the solar / lux module. I did some measurements and noticed that the supplly rail dips from battery volts by 200 mv when transmissions occur. There is a 100 uF elect cap across the supply in the solar module. I have replaced this with a 1000 uf cap and now the dips only about 20 mV. Will leave it running as a test to see how it goes.

Phil

_________________
http://www.auzhq.net/weather - using a VP2+

Hi Phil,

I sure will keep you informed

in return please let me know how replacing the cap goes...

I thought as much I've owned stations for years and the universal signal for no sensor signal is the dreaded - or -- hehe

Power cycle reset gave me 0's and I shone my LED maglite onto the sensor to check and it showed readings

Data registered into Cumulus (brilliant program, I swear by it alongside many other stormchaser friends of mine.)

I owned a 1081 previously but desired to upgrade to the 3081.

Hopefully I get no further issues *touches wood*

station had signal all night long, solar sensor ran for 2 hours this morning and has promptly crashed, how disappointing!

I'll power cycle it once more and if that doesn't resolve it, contact fine offset...

note: this station is not even 1 day out of the box yet!

power cycling is doing nothing, no signal....

majorly disappointed!

opened up the casing, no tin on the oscillator...

think its soldering time!

tin applied to the oscillator but I also went through the circuitry with a fine toothed comb to make sure that nothing was shorting, the power supply input/output wires from the solar panel may have been touching at the point they contact the pcb, I had to bend them outwards as it looked to me there was a short being caused there, this would definitely make sense as to why some people are reporting non charge from the panel.

An observation to make, I applied the tin last and as soon as I did the little LED on the solar panel was flashing at it's proper interval under the tungsten lamp I was using, if thats not an indication for bringing it back to life I don't know what is!

Anyways I shall keep everyone posted on if this fix worked, the oscillator was sitting over the earth rail but I doubt the contact was doing much, I gently moved it off it and applied the solder very very minutely.

Hi Stephen - exactly the same failure mode I am experiencing.
Reset and it works all night and then about 2 hrs after the sun wakes up - bam!

Point to note - I have learnt the correct way to get readings. A simple reset of the solar unit (which also resets the temp humidity unit) is not enough. As soon as you reset you need to bolt over to your console and press the down arrow for 5 secs+, you should see a receive symbol constantly on at the top RH corner of the display. The console is now searching for the transmitter.

The reason for this is because its now obvious to me that the console only enters receive mode for about 1 second when it "expects" to hear from the transmitter (which is once at T zero + 48 secs and once at T zero + 60 secs and multiples of those there after). The Temp / Hum / Wind / Rain is transmitted every 48 secs, and the solar / lux every 60 secs. When you reset the outdoor stuff it is going to be very unlikely the new transmission times will match the old ones although the transmission interval remains the same. Hence the need for a resync. All in aid to save console battery I am sure.

It is shame this is how it works because a simple work around for our little issue would be to make a timer and reset the solar unit periodically, but because of the need for a resync this makes it all too hard.

On the mod, it ran all night in my test room, and was still running even though the one outside had failed 2 hrs earlier. I am not too excited about this (but have some hope) as I am worried its the action of the increased light level by the sun that is the issue.

Non the less I decided to improve things a little and put a bigger electrolytic cap (and a brand new one too) of 1500 uF (6.3v) and my test now shows negligible volt drop during transmission (< 1mv). The cap is now properly mounted using hot glue and connected using two short insulated wires to the board. I have also put a 0.1 uF ceramic in parallel as I noticed some high frequency noise with my oscilloscope. The noise is roughly halved. I wanted to make sure the noise was not part of the problem.

If I get this thing running for more than 48 hrs I will post pictures of what I have done. Basically the 1500 uF, 6.3V cap ith 0.1 uF cap soldered on the end is fitted on the same side but where some batteries are designed to go (but not used in our case). This allows the board to fit back in with the bigger cap with no issues.

Lets see how it goes - 1 hr with fluctuating light levels and all is still good. Going by history if I reset in the morning it will fail mid afternoon - so fingers crossed its still working tonight.

Phil

_________________
http://www.auzhq.net/weather - using a VP2+

Hi Phil,
maybe installing some (rechargeable) batteries may help?

_________________
Conder, Canberra Weather

BTW I agree Cumulus is fantastic and as soon as my Paypal Acct recovers from Christmas spend I will be donating to Steve. He deserves it.

Also I was in Jaycar wand was shocked to see how mucy the 1081 is there (or the equivilant - it looked like the 3081 but no solar / lux circutry) - $199. I got two 3081 units for $228 from eBay and to be fair all the functions of the 1081 work fine (except for possible wind dir sensor improvements).

Phil

_________________
http://www.auzhq.net/weather - using a VP2+

Yeah thought about that (actually thinking about a super cap), but there is no voltage present on any terminal where the batteries would go

_________________
http://www.auzhq.net/weather - using a VP2+

Possibly a Zener diode or something hasn't been populated?

_________________
Conder, Canberra Weather

Jaycar charges way too much for the 1081, its not a bad station at all but being tech savvy like all other fine offset users having another parameter to trace is just like putting a free liquor sign in front of an alcoholic haha.

1hr running so far and she's still ticking over

I'm not surprised you've had to change the cap, but with negligible voltage drop it should enhance the reliability well theoretically....

I didn't realise you needed to push a console button, thank you for that tip! I've just been ripping out the batteries alone.

Arghhhhhh!

Just checked and it has failed again.

I am getting more and more convinced that there is an issue with the increased light levels perhaps causing an analogue input into the solar microprocessor to go out of range thus making it hang or maybe that the firmware in the processor is not handling some situation that occurs.

I am heading out to puts a dark rag over the sensor and see if that affects things.

BTW - I noticed that the web links to the Fine Offset site are not working - has anyone had any luck getting onto their site lately? Does anyone have a contact (the OP or one close to it referred to one)? I don't want to give up if I can help it.

Phil

_________________
http://www.auzhq.net/weather - using a VP2+

Its possible - there's a lot of un-populated circuitry on the board. It looks like there's room for a display, one or two more ICs and some other connectors of soughts.

Phil

_________________
http://www.auzhq.net/weather - using a VP2+

that seriously sounds like a processor issue, mine is working in all levels of light at this point in time, its partly cloudy so its alternating between cloud and blue sky and yet its still ticking over.