Temperature Spikes and Spike Removal

[Super-T]

I've also noticed that small spikes can appear when the wind speed increases above a stiff breeze. But then again it could be the wind direction is more varied at higher wind speeds. Hard to tie down as it never does it while watching. Perhaps the AtoD thresholds between measurements is too close with overlap occasionally.

[serowe]

Connectors in ANY data cables will cause probelems at some stage. For every connector you use on a data line you increase the liklihood of either losing data, corrupting data or just completely severing the connection.

Plus if you ARE using Cat 5, if you make it yourself, make sure you follow the standards for Cat 5 cabling - don;t just wire them straight through (pin 1 to 1, 2 to 2....8 to 8). If you need to know the correct terminations do a search for Cat 5 and realise, then, that there are two basic standards as well Also, if making them yourself, make sure you have the correct crimping tools and, unless you *do* have enough experience, don;t try putting what you would call wall plug (female) RJ45 connectors anywhere along the line - they are designed for static, totally non moveable, permanent fixing and require the proper tools for terminating the cable.

Also - when you are running ANY data cable, keep them as far away as possible from both AC and DC power lines and also speaker cabling for a HiFi system (it is essentally AC albeit low voltage). This can be difficult to do if you run them through the roof of a house because you will have power and lighting circuits up there and all of these can cause interference.

DC doesn't spound like it will do much, but if it turns on and off it can generate enough of a spike to inject garbage into a data line.

Final point - keep your data lines as short as physically possible without putting srain on the cable(s) and don't, whatever you do, be tempted to use Cat 5 cabling outdoors without complete protection for it along it's entire cable run - it is not designed for outdoor use and it WILL get damaged very quickly (even if you can't physically see external damage to the sheath.

[serowe]

I've also noticed that small spikes can appear when the wind speed increases above a stiff breeze. But then again it could be the wind direction is more varied at higher wind speeds. Hard to tie down as it never does it while watching. Perhaps the AtoD thresholds between measurements is too close with overlap occasionally.

Wind direction would tend to be more stable at higher speeds, not more variable..

[Super-T]

Only if it's not being diverted by local structures.

[serowe]

But it still won't vary consistently - if you are getting swirls then, obviously, the wind vane is in the wrong position to start with - but that hasn't been mentioned here so has to be presumed to not be the cause here.

[yv1hx]

Connectors in ANY data cables will cause probelems at some stage. For every connector you use on a data line you increase the liklihood of either losing data, corrupting data or just completely severing the connection.

Plus if you ARE using Cat 5, if you make it yourself, make sure you follow the standards for Cat 5 cabling - don;t just wire them straight through (pin 1 to 1, 2 to 2....8 to 8). If you need to know the correct terminations do a search for Cat 5 and realise, then, that there are two basic standards as well Also, if making them yourself, make sure you have the correct crimping tools and, unless you *do* have enough experience, don;t try putting what you would call wall plug (female) RJ45 connectors anywhere along the line - they are designed for static, totally non moveable, permanent fixing and require the proper tools for terminating the cable.

Also - when you are running ANY data cable, keep them as far away as possible from both AC and DC power lines and also speaker cabling for a HiFi system (it is essentally AC albeit low voltage). This can be difficult to do if you run them through the roof of a house because you will have power and lighting circuits up there and all of these can cause interference.

DC doesn't spound like it will do much, but if it turns on and off it can generate enough of a spike to inject garbage into a data line.

Final point - keep your data lines as short as physically possible without putting srain on the cable(s) and don't, whatever you do, be tempted to use Cat 5 cabling outdoors without complete protection for it along it's entire cable run - it is not designed for outdoor use and it WILL get damaged very quickly (even if you can't physically see external damage to the sheath.

I would aggregate connecting the cable shield to ground whenever it's possible to do this.

[serowe]

I would aggregate connecting the cable shield to ground whenever it's possible to do this.

Except Cat-5 doesn't have a shield

[Stephenwx82]

no but its a twisted pair which does the job nicely....

not a single spike in coming up on 2 weeks of running with this setup the way I did it....

[Skyline]

To respond to the points made by Serowe above...

The Cat5e shielded cable I have used has no joins from the circuit board to the transmitter. I also used a high quality crimping tool to make sure the plug was well terminated. The cable is currently on the roof as a temporary measure but I cant get it any further away from any cables which may be hidden in the walls of the house. The point you make about using the right terminations is interesting. I have used two of the twisted pairs, one pair (I assume) for the direction sensor and one pair for the wind sensor. The other two pairs are not used. Whether this is the correct or best way to use the Cat5 in this particular case it is beyond my electronics knowledge to say, but I'm happy to try a different configuration if anyone knows the answer?

Last night was almost completely calm, negative temperature spikes of around 4 degrees C....

Skyline

[serowe]

Personally I wouldn;t be using the same piece of cable for more than one sensor. So one distinct cable for temperature sensor, anbother for anemometor, another for rain gauge etc. This is how I have mine set up and the only spikes I get come from the temp sensor itself for a defined period when sunlight strikes the walls of the building it is on. It did have a well known issue with 'open circuit' readings but shielding the sensor stopped this as well )many Wm-918 users have had the same issue so more a design flaw than a 'problem')

One thing you don't mention - how far away are they? One thing I learnt the hard way with Davis instruments - go over 15 metres and you start to lose signal. Remember these units are not designed or meant to be large distances away - thus they have an extremely low transmitting 'power' and even with good quality cable and very good connections, they will still have an issue with lng distances. Personally I would keep them no more than 10 metres away from the main console (you can then - depending on the type of connection to the computer - make this distance as long as possible to help move everything away from you).

Remember - if you are using software (whether it be VWS, WD, Cumulus or whatever) you don;t really need to see the actual station console - mine is situated under my house (because that's where my servers and UPS/power supplies are all located) and the onlt time I see it is if I have to physically reboot a server or need a bottle of wine from the cellar that is also under there

Stephen

[Skyline]

Thanks for that. The only reason I moved the whole shaboodle in the first place was because of the high temperature spikes you get in the sun even with the screen fitted. As for distances its 6m from the wind/rain sensors to the transmitter and about 15m direct line of sight through the window from the transmitter to the console. I'm not sure if you can easily put in separate cables for the anemometer and the wind direction. The WH1081 has them wired in series i.e. the anemometer is wired into the direction sensor and then one cable runs back to the transmitter. I dont suppose thats a problem because I think they dont share circuitry, just the RJ11 plug socket. However, the transmitter also only has one RJ11 socket so you would have to connect them back up again to plug them in.

I think I am going to try to move the whole thing closer to the console anyway, that way I can use the original 2m cable (which ironically never gave spikes even when sited 50m from the console) and still put the transmitter in the shade.

Skyline

[Skyline]

Just thought I would provide a final update on this in case anyone was interested in the outcome.

Have now moved whole station so it is on the roof just above the room where the base station sits so just 2 or 3m away through the wall. This allowed me to shorten the new CAT5E cable on the wind section to just 2m or less and still put the temperature transmitter in the shade. Alas, this has not solved my problem and am still getting those irritating spikes which destroy your temperature max and mins. The only thing I can say for certain is that it is the anemometer that causes the problem. The wind direction sensor can run all day without causing a spike, but as soon as the anemometer is connected the spikes appear.

So I am resigned on this one for the time being like everyone else with this type of station (does anyone with a WH1081 actually NOT get this problem at all?). I'm just going to extend my logging interval out to 20 or 30 minutes as this will reduce the frequency of the spikes and run cumulus during the daytime when the p.c. is in use so the spike removal function operates.

Long term - get a different station!

[Tau Bootis]

Hi

I have never had a problem with temperature spikes until my TV aerial mast (& wind sensors) snapped in the gales a couple of weeks ago. As soon as I re-connected it the interference started; disconnect the anemometer and the severity halved; disconnect both sensors and it was stable. My logging interval is 5 mins by the way.

In my case the spikes are not big, only between 0.1 & 0.3 C but they are constant rather than intermittent, at first the spiking was quite bad, then I cleaned the RJ11 plug and socket in the wind direction sensor, as it was covered with an inappropriate grease which I think was part of the problem; originally the grease was applied after the plug was inserted but after I had unplugged it, it covered the contacts.

At one point the temperature sensor flat-lined for several hours but fixed its self before I got chance to look at it, I have looked inside the sender unit and visually it looks clean.
At the moment the spiking/interference is 0.1C so still some work to do.

I think the RJ11 plug on the anemometer is a bit iffy and I may replace it as my next move, another thing that could have changed is the way the way the bog standard phone cable that I am using has been routed down the aerial mast, at the moment it is bunched together with 3 TV aerial cables! They were there before but not so tightly bunched, not sure whether they can interfere or not, I guess I could unclip it as a quick test.

To sum up, it must be something at the wind sensor end as it was perfectly fine before the aerial broke and I had it all to bits. I also had to repair the anemometer not too long ago and didn’t have any problems then.

If I find the cause I will post an update.

[Skyline]

I would be chuffed to bits with a variation of 0.1 to 0.3C with a 5 minute logging time. That sounds really good.....

[MickinMoulden]

In my case the spikes are not big, only between 0.1 & 0.3 C but they are constant rather than intermittent

Ummmm...How can 0.1C be a "temp spike" when the minimum it can change IS 0.1C? That is not a spike. Also, TV arials cables can hold high voltage, they're not supposed too, but with some of the connections I've dealt with (probably the units were not grounded sufficiently), I've had shocks through them. 0.3C is still pretty good. I've get normal temp changes of .3 to 1.0C in a minute, however I do get "spikes" of 0.9 and higher, so it's a balancing act. I have my spike removal at 1.1C. It started when I moved the outdoor unit further away and also had extended the wind/rain gauges. I feel I may fix it by moving the indoor unit closer by extending the USB cable. I will test this when I'm bothered