I wonder if it could be possible for the firmware in the hub could correct the wind direction reading if the Tempest isn’t oriented 0° N? I figured it could do it by using the location information that’s provided by the user. Thoughts?
How is it going to know it needs correcting?
It won’t “know” but it will spit out the data as if the Tempest is oriented correctly (even though it might not be).
In the “Advanced Settings” for the “Device” is a widget that will allow you
to “adjust” the offset angle, if that helps.
I do know about that option, but you will have to know exactly how off your readings are in order for that to work as intended. This proposal will automatically do any adjustments for you.
I think this would only be possible if the Tempest had an on-board compass/magnetometer. Otherwise I simply can’t see a way for the Tempest to know how it is oriented.
Again, how is it going to know it needs to correct the data?
Don’t GPS receivers include a compass? (I am naive about such things, though).
no, gps doesn’t have a compass, but often it can determine orientation when you are moving in a straight line.
Anyway the unit doesn’t have gps, so it couldn’t use it anyway.
The only way to autocorrect the orientation is if it consistently, over a very long time, is giving an offset compared to the expected wind direction. If it does it might correct for that (currently it doesn’t). As weatherflow apparently love those algorithms they might implement it, but I rather see them work on other stuff. Most people can easily determine true north and orient the unit accordingly. Do you really care if it would be off be a few degrees?
I sure agree about the “few degrees”, given the variability of surface winds.
I’m curious, though…when I installed my Tempest, the “station info” showed it was at an elevation of 19.7 feet AGL, which is pretty much correct (chimney mount)…I figured it must have a GPS in there in order to figure that out! Also, when I look at the tempestwx map, it shows the exact location of the Tempest on my house!
I guess I haven’t thought about it long enough @GaryFunk. @peter would be correct. And @sunny, I’m one of those weirdos that wants everything to be exact. Lol. I just figured it could filter the calculation through just knowing the location of the unit. But you’re right, it would also have to know its current orientation.
the default high during installation is set to exactly 6 meter (which might or might not translate to 19.7 feet). At some point you gave the station that location.
Thank you all for correcting my “mental model”! Obviously, I’m a “new kid on the block” – having come from a couple of decades in the 1-wire arena, since replacement parts are now hard to come by, I’ve been waiting for something like the Tempest to replace all that. I appreciate your insights…and patience! And, yes, while I was on the roof putting up the Tempest, I probably did use my phone to say “here it is”…but that was a couple of months ago…
This is a feature request where we are supposed to clarify the feature without too much discussion. But the request is unclear.
It is not clear in the current instructions that during setup it seems to take ground level at your phones location during setup for your hub altitude above sea level and it sets 6m (19.685’) extra to your Tempest height above ground. You can change both values in settings.
The user can already provide the ‘orientation’ information to adjust the wind direction offset in settings. See @tomw_mail comment above.
Question to @jgentry Justin: Are you satisfied if we close this request?
I do know about that option that you’re referring too and it’s great if you know how many degrees you’re off by. However, my station’s location can throw those compasses off and I’ll have to find a landmark that represents 0° N to fine tune my Tempest orientation. I figured if WF could conjure up an algorithm that could automatically correct orientation issues, it would be awesome. Currently my Tempest is oriented fairly well but I’m not sure it’s 0° N. I’ll find out when I get this compass in the mail https://www.brunton.com/collections/outdoor-compasses/products/truarc-3?variant=907594737 and go back to the station’s site to fine tune it. I’m a accuracy freak so I want the setup to be as perfect as possible. Lol. Plus, this station’s location is perfect to record data during hurricane landfalls.
For the ‘accuracy freaks’… if Weatherflow installed something to measure magnetic North it would be influenced by local variations which change with time and nearby magnets and devices that generate magnetic fields. Meaning a compass would only be approximate.
The difference in the angle between magnetic and true north can be as much as 180 degrees opposite in some places.
However the solution without obvious land marks available might be a type of sun dial if for your location you know when the sun is at true south or north.
This web site explains options:
If the bird spikes were installed precisely you could use the shadow.
More likely because it is on top of the pole you need a guide lower on your pole. What I do is attach my alignment bracket onto my pole where I can site along it standing behind the pole. Then I lower my pole and attach my Tempest in alignment with the bracket. I mark extra lines with a pen onto the Tempest to help.
From your location you might be able to sight a pole or something along Dogwood street as true North.
Thanks! You’re definitely right about the compass. Here is a picture of the current setup. The black circle is where the station is located. The pin mark is a close approximation of where the arrow of the Tempest is currently pointed.
Stand behind the pole and clamp something that you can sight along towards the North aiming at where Dogwood street bends or to something you can match on maps as true North. I screwed an aluminium cross that I can bend to match N/S/E/W. Then adjust the Tempest to match the alignment on the pole.
Thanks! I’ll be tinkering with it when I go down there again.
@dsj gave some tips also. Thinking about putting a fence (maybe a couple) posts up where that curve is and line up the unit . From what I can tell from the map, my unit is already pointing at the curve. So If the curve is true N, then my unit shouldn’t be off by much if any.