Humidty reading very low - wires swapped?

[mattsimis]

While I appreciate this isnt a support forum for Fine Offset, I just thought Id see if anyone can save me dismantling my 8meter weather poll again!

I took it down to wire up a DC Adapter instead of the batteries in the transmitter. That was all fine, stronger signal and no more 5mth lifespan on batteries. However I was reviewing the data in Cumulus and noticed that when its dry out Humidity is reading "--" to 10% and in the rain its reading 50-80%. Previous figures would have been near 90%.

I was wondering if I accidentally swapped the RJ45 connectors on the back of the transmitter and it could be reading perhaps the Rain sensor for Humidity and vice versa?

[Gina]

I was wondering if I accidentally swapped the RJ45 connectors on the back of the transmitter and it could be reading perhaps the Rain sensor for Humidity and vice versa?

No. The connectors are for the wind sensors and rain gauge - the humidity is part of the transmitter/temperature unit.

[mattsimis]

Ah right thanks. In that case I gather I can force a flat +20% on the readings in Cumulus to get it closer to reality...?

[EvilV]

Ah right thanks. In that case I gather I can force a flat +20% on the readings in Cumulus to get it closer to reality...?

I think I'd want to check other local readings before doing that. As the season goes on, I wouldn't be the least surprised if humidity changed. I'm no meteorologist, but it is my impression that my humidity figures are far different now in early May to what they were in February when I first got the machine. Then it was 95% day after day. Now much lower, in the 70% range. I expect there is a function related to temperature influencing it. I'm not sure where you are, but my figures relate to the north of England about nine miles from the sea.

Check other local stations first, I say, before introducing random correction figures. By doing so, you will in any case be able to actually correct the numbers so they relate to reality rather than guesswork. Also, if you do have certainty that the unit is misreading humidity, check that any ventilators that allow air to come into the unit to be samples are clear. There might be some very easy solution to the problem if there is one.

[mattsimis]

True, I should check other home users stats. Im in Ireland, near the midlands. I looked at Air Ports and stuff (http://www.met.ie/latest/reports.asp) and their Humidity figures were 60-70% when mine was 17%.

EDIT: I think my unit is broken (or not liking the 4.5v its getting vs the 3.2v the batteries gave out), the humidity figures went upto 90% today but then smoothly dropped to 15% again. That doesnt sound right!

[Gina]

I think my unit is broken (or not liking the 4.5v its getting vs the 3.2v the batteries gave out), the humidity figures went upto 90% today but then smoothly dropped to 15% again. That doesnt sound right!

Why are you running it off 4.5v? It might be OK to up the voltage a bit but I think 50% increase is pushing it. They're designed to run on 3v. If 4.5v is what your PSU puts out, I'd recommend putting a couple of silicon diodes in series to drop the voltage to just over 3v.

[mattsimis]

I think my unit is broken (or not liking the 4.5v its getting vs the 3.2v the batteries gave out), the humidity figures went upto 90% today but then smoothly dropped to 15% again. That doesnt sound right!

Why are you running it off 4.5v?

Cos I was just using what was onhand...
Ill go sort out something better now.

[Gina]

Cos I was just using what was onhand...
Ill go sort out something better now.

Well, I hope you haven't blown it up!

[EvilV]

I think if your humidity is so much different to other reliable local stations you certainly have got an issue. Gina is right for sure. It's quite a leap to jump from 3v to 4.5, and also the typical wall wart things are not that tightly regulated, so it might actually be higher in such a low current device as the FO. Without a load, those things can float higher than their stated voltage, even though when you put a demand on them they come near enough to their rated voltage.

I'd go straight back to a pair of new batteries and then see what happens. Hopefully it is ok.

[beteljuice]

It is not impossible that your unregulated psu is over-running the board / components so much that local heat is been generated - enough to 'dry out' the air in the sensor unit.

Try an old mobile 3v charger if you have one.

[Super-T]

If it's an unregulated 4.5 volt supply with hardly any load on it, it could well go to 6 volts. Have you measured the output or just read the label?

[mattsimis]

If it's an unregulated 4.5 volt supply with hardly any load on it, it could well go to 6 volts. Have you measured the output or just read the label?

Measured the output with no load at all.


I put in a regulated 1.2amp, Voltage selectable unit last night (and set to 3v).

Humidity reading did seem to jump, to the 70-80% range from under 50%. I thought that might be a little low still, so I put a 10% calibration on it in Cumulus, though looking at the trend graph Im now hitting 100% throughout the night, so maybe thats not needed.

So yeah, it seems the overvoltage was causing the erronous readings.

[Gina]

I put in a regulated 1.2amp, Voltage selectable unit last night (and set to 3v).

Humidity reading did seem to jump, to the 70-80% range from under 50%. I thought that might be a little low still, so I put a 10% calibration on it in Cumulus, though looking at the trend graph Im now hitting 100% throughout the night, so maybe thats not needed.

So yeah, it seems the overvoltage was causing the erronous readings.

Q.E.D. Good, luckily it seems you didn't do any permanent damage

[mattsimis]

I put in a regulated 1.2amp, Voltage selectable unit last night (and set to 3v).

Humidity reading did seem to jump, to the 70-80% range from under 50%. I thought that might be a little low still, so I put a 10% calibration on it in Cumulus, though looking at the trend graph Im now hitting 100% throughout the night, so maybe thats not needed.

So yeah, it seems the overvoltage was causing the erronous readings.

Q.E.D. Good, luckily it seems you didn't do any permanent damage

Indeed and Im a lot happier not having to change the batteries continiously too!

[beteljuice]

It can take HOURS for the RH sensor to settle after a sizeable 'jump' - they are quite slow in their (final) response.

Leave any calibration alone for a day or two