Tempest Charging Questions

Please do open a support ticket
Let them double check with more in depth data than we can get. Maybe something is wrong and panels are not giving enough and when it becomes critical like now, it shows when in mid summer, there is way enough.
Mention this post in your ticket. This way they see already what was tried and seen. Might help to get a quicker reply.

i did create a tcket several hours before I started discussions here. I figure it could take several days but after that i plan on shutting it down if it continues fail to charge. I have a raspberry pi capturing udp data and no sense in wasting power. I came here just to see if i was overlooking something. I WAS… I thought it was the solar module(s) not Battery. I never looked at the batts until you woke me up.

Thank you, Eric and Sunny, I appreciate you and all others who have helped. I have gotten my money worth regardless.

Bill W

At least wait until WF support can take a look. Remember, the hub takes very little power to operate.

I’d like to keep it running but I am not going to invest any further. Like said earlier, I have gotten my moneys worth already. Learning new communication options was valuable. WS and UDP were new protocols on the local level.
I will not switch it off until its dead in the water with no hope. I simply can’t move it myself even if that would help.


Booster pack would be fantastic. Living in Fairbanks Alaska I don’t see the sun for a few weeks during winter when it fails to clear the hills. And it is dark most of the time, too! Can Tempest be brought inside and recharged with a light bulb?


gizmoev did it and it works though not easy these days as we all use leds and that won’t work whatever power you set.

that probably isn’t true, as it depends on the wavelength of the led. Use one that matches the peak wavelength of the solar panels and you should be fine.

That assumes you know what wavelength the panels use and that you can find an LED with that output. It is likely much cheaper to just get a 250W heat lamp like I did as those are often available, at least in the US, for bathrooms and livestock heating.

but most solar panels use a very broad absorption spectrum, so it doesn’t matter too much. You should be able to charge using leds.
I have a spare (broken) tempest, I’ll test with some leds. But not now as my station is in transition due to new hub arriving, and I don’t want to confuse the current problems by adding another tempest now)

It matters in the fact that LEDs are not broad spectrum output devices so there is very little energy to capture. Incandescent, on the other hand, is a broad spectrum output, though not usually in the UV end of things. Feel free to experiment. I’d be interested in the differences if you have both light sources as options.

Hi, I have a very similar problem. Living north of Seattle is certainly gray and dark this time of year. My Tempest has a lot of trouble maintaining charge. I tried a Halogen work light (300W) at a distance of 2 Ft. The affect was negligible. After doing this for several days I swapped out the Halogen for a RGB 40W LED flood light set to color of White at a distance of 6 inches. In this LED there is no “white” source other than turning on the Red, Green and Blue components simultaneously. I will try to post the graph as well. On the graph, the slope for charging really took of (if you can say 72 hours of continuous light exposure is taking off LOL. It did work but boy, did it take a while. The charge lasted a few weeks and now I am back to where I started. I really wish Weatherflow would get that booster on the market. I really need it.


I made a corner reflector out of a couple of sheets of Styrofoam to concentrate the light on all cells better. It didn’t take but a day to charge back up to probably half charge. This was with a field test unit running full throttle without any power saving modes setup yet.

1 Like

so, that settles that. It does charge just fine with LEDs (imagine if you would use 250W instead of 40W :-). LEDs are so much more efficient. :smiley: :smiley:

But thanks for the graph. I would like to get more graphs like this from people where the unit doesn’t charge for a long time (or better the numbers). From that I would like to calculate what the base power usage is, with everything working, except for the wind sensor, and calculate how much a single wind measurement takes. Using that I would like to calculate how long the unit takes from 2.65V to 2.355V depending on the number of wind samples taken. My guess (just guessing here as I don’t have sufficient data yet) is that it could take 8 weeks to do so if only sampled once a minute And perhaps 5 weeks if sampled 4 times a minute.


You dont need to be concerned. I charged mine with a RGB 50W light. Set color to white. (R G and B full open) It charged in two days to 2.78. When you stop charging it drops down to about 2.56 in no time. After that it’s going down slowly. The upper voltage window does not have much energy. 2.56 is great.

1 Like

it’s part of the characteristics of this type of battery. Look at the curve in https://www.researchgate.net/figure/The-OCV-SOC-characteristic-of-the-LTO-based-battery-measured-at-different-temperatures_fig4_324117029

Mine came with no charge and took about 2 full days to even come to life around Dec 26th. The power had stayed pretty snug to 2.4, but slowly started creeping up, and just today it briefly shot up to what it looks to be 2.8 and change. It was sunny but not out of ordinary. I was wondering if there was a firmware update or something.

The battery voltage won’t change due to a firmware update. 2.8V is the max charge voltage. There really isn’t much energy between 2.8V and 2.6V so don’t be alarmed if the voltage drops rather quickly initially.

If you put a link to your station in your profile we could look to see what firmware you have. Right now the latest is v160 for the hub and v134 for the Tempest.

1 Like

I just rigged 2 x 15W LED Floodlights on a homemade 3D printed setup to charge my Tempest. It was down at 2.37v and updating every minute. After 16 Hours it was back up to 2.8v under the artificial

Here is a graph of the charging sequence, every unit on the x axis is 30 minutes.

I also pointed the lights vertically down on the top of the Tempest at the Light Sensor and the light equates to about 800W/sqm.


Great idea. Congratulations!

Great setup! Looks like you are planning on using this often. I’m hoping more for a permanent solution from weatherflow.
For my little experiment ( Tempest charging with artificial light ) I had them in the same orientation as yours, but just put them on the floor together with my tempest.
Later I realized that my lights were labeled as 15W but had both rgb leds and warm white leds. Coming from china that probably means that the combined output is 15W, but I couldn’t turn both sets on at the same time. It took 2 days to charge the tempest.