How to fix influence of rain on hive weight

9 posts / 0 new
Last post
wwfoste
How to fix influence of rain on hive weight

I see a fairly dramatic change in weight (2-4) lbs due to rain (see attached). I assume this would be due to an issue with mounting of the load cells and not just actual changes in the hive. Has anyone run into this and is it something that I can fix? Or is it worth the hassle and just let it correct itself after the rainfall, which it seems to have done.

Image: 
wwfoste
How to fix a scale gone crazy.

After the heavy rainfall we've had, my scale's gone crazy. Any ideas as to what would cause this and what the best approach to fixing this issue is? Is it as simple as just restarting the Pi, or will I need to take the hive off of the scale?

Image: 
Paul
Paul's picture
data shows slow death

The weight data (third column) shows a slow death over an hour and 15 minutes. This might be more consistent with water shorting a load cell than a sudden failure due to lightning. If it is water in one load cell, maybe it will dry out. But we have never had a failure like this, so if just one cell failed, I would say it is a manufacturing defect?

179 05/01 14:55 168.44 97.5 57.0 82.2 0 82.2 63.0 0.9 2.0 133.0 68.4 1014.1 0.44
180 05/01 15:00 115.87 97.7 57.0 82.4 0 82.4 63.0 6.0 10.0 172.0 68.5 1014.1 0.44
181 05/01 15:05 115.41 97.7 57.3 82.2 0 82.2 65.0 4.0 4.0 249.0 69.3 1014.1 0.44
182 05/01 15:10 114.94 97.7 57.3 81.9 0 81.9 63.0 4.9 9.0 195.0 68.0 1014.1 0.44
183 05/01 15:15 113.95 97.7 58.0 81.3 0 81.3 64.0 4.0 8.0 248.0 68.0 1014.1 0.44
184 05/01 15:20 59.71 97.9 58.0 81.5 0 81.5 64.0 4.9 10.0 179.0 68.2 1014.1 0.44
185 05/01 15:25 55.57 97.9 58.4 81.3 0 81.3 64.0 6.0 10.0 212.0 68.0 1014.1 0.44
186 05/01 15:30 46.88 97.9 58.4 81.9 0 81.9 64.0 4.0 7.0 224.0 68.5 1013.8 0.44
187 05/01 15:35 -20.89 97.9 58.9 82.2 0 82.2 64.0 2.0 7.0 183.0 68.9 1013.8 0.44
188 05/01 15:40 -42.63 97.9 59.1 82.2 0 82.2 65.0 2.0 10.0 206.0 69.3 1013.8 0.44
189 05/01 15:45 -42.55 97.7 59.1 82.8 0 82.8 64.0 0.9 6.0 250.0 69.3 1013.8 0.44
190 05/01 15:50 -47.90 97.7 59.5 82.8 0 82.8 63.0 4.9 10.0 215.0 68.9 1013.8 0.44
191 05/01 15:55 -51.56 97.9 59.5 82.4 0 82.4 63.0 2.0 9.0 252.0 68.5 1013.4 0.44
192 05/01 16:00 -49.89 97.9 59.2 82.4 0 82.4 63.0 0.9 5.0 158.0 68.5 1013.4 0.44
193 05/01 16:05 -53.84 98.1 59.2 82.2 0 82.2 65.0 2.0 6.0 217.0 69.3 1013.4 0.44
194 05/01 16:10 -35.75 98.1 59.2 82.0 0 82.0 64.0 6.0 6.0 222.0 68.7 1013.4 0.44
195 05/01 16:15 -59.30 98.1 59.2 82.0 0 82.0 61.0 4.9 11.0 184.0 67.3 1013.4 0.44
196 05/01 16:20 -464.29 98.1 59.3 82.0 0 82.0 62.0 2.9 8.0 264.0 67.6 1013.4 0.44
197 05/01 16:25 -515.06 97.9 59.5 82.0 0 82.0 64.0 0.9 6.0 209.0 68.7 1013.4 0.44
198 05/01 16:30 -727.57 97.9 59.5 82.2 0 82.2 64.0 0.9 3.0 250.0 68.9 1013.4 0.44
199 05/01 16:35 -999.99 97.9 59.4 82.0 0 82.0 62.0 2.9 6.0 237.0 67.6 1013.4 0.44

wwfoste
scale recovery

Well, just this morning it did a massive jump and recovery and is near where it started at. So I'll keep an eye on it and see what it does. I've got some scrap tin that I may use to make a rain block for the future.

Image: 
Paul
Paul's picture
Weight gain after rain

There could be two things going on:
1. Absorption of water by the wooden-ware. What is the lid made from? A metal lid will shed the rain pretty well and dry quickly. An old half rotten lid takes longer to dry and you may see the bees kick up the internal temp to dry the "roof". I have even placed shingles on the lid. If it is the wooden ware, it will dry out. If you want to eliminate that "noise", the hive needs to be in a bee house or place some kind of shelter over the bees. I haven't come up with a good design. Possibly a scaled down version of the tents used at outdoor events by vendors? One suggestion was a PVC frame with a canvas or plastic tarp top. Maybe a wooden telescoping three sided box?

2. Collection of water by the bees. Looking at the rate of weight increase (kind of slow and linear) and the rate of loss (kind of slow that looks like evaporation). It may be the bees were taking advantage of water on the landing board to bring some in. They use it to dilute honey to feed brood and maybe to stabilize the temperature (evaporative cooling) and humidity in the hive and I guess they get thirsty too.

This is the main reason we think a rain gauge at the hive is important. This was a pretty big storm system that moved through so the rain gauge at the different location probably fairly reflected the amount of rain but possibly skewed the timing. Contrast that with small local cells that pop up in the summer. In that case, the rain on the hive and the recorded amount at the different location may be totally different.

wwfoste
Weight gain after rain

I just have the standard metal telescoping top with painted wood on the side, relatively new. I'm not that worried about the noise from the rain, and I figured some of it was the bees taking it and the boxes absorbing a little bit of it. I may look into the rain gauge at some point, but I'll need to replace my USB hub, it's a Belkin but only 2 of the connections work on it. Thanks for the response on this.

Paul
Paul's picture
Scale gone crazy

Never seen anything like this before. Some theories:
1. Water got into the electronics. Problem is the Pi is still sending data and the temp/RH is still working.
2. Water some how expanded part of the frame and the expansion lifted the hive and removed weight from the scale. But your frame is metal and plastic so I don't see how that could have happened and when that happens, the weight goes lower, but not to -1000.
3. The water destroyed the load cells but yours are well protected.

Also, if you believe the rain gauge, the failure occurred after the rain.

So, I don't have a clue. Can you unplug the load cells and see what it reads with no load cells attached?
Since it went negative, and hard, I would expect to see the input voltage to the HX711 ADC near ground. Only problem is you can't really measure it since the excitation voltage is only turned on when the HX711 is sampling the cells (only a few milliseconds or less).

So, my trouble shooting advice is to 1. unplug all the load cells and see what it reads. 2. Measure the resistance of each cell with an ohm meter. The easiest way is to use a modular phone jack (maybe the square ones they mount to a baseboard when you wire a phone extension). Plug in each load cell and measure red to black and then white to black and green to black on each cell. It should be around 300 ohms. Maybe measure green to white while you are there.

One possibility - we had a few big lightning strikes. It is possible the ADC input or load cells blew. Maybe measure the resistance of the cells and if they look normal, replace the HX711? I can send you one or drive over.

One more thought. If you unplug all the load cells and it is still pegged at -1000, I'd say the HX711 input blew. If you unplug all the load cells and the HX711 output changes (probably not to zero, but to a random number (may be noisy), then maybe a cell has shorted. Then try plugging the cells back in, one by one.

wwfoste
Scale gone crazy

I'll see if I can get to this tonight. It was weird that it would do a drop, then start gaining weight back, then another big drop, then start recovering again, until it finally bottomed out.

wwfoste
Scale tracking with temperature

Ever since the scale bottomed out and came back, it's been closely paralleling the ambient temperature. I'm thinking that maybe the leveling leg is too close to the strut such that changes in temperature are causing some expansion and driving the weight change. Before I just back the leg out and just forget overload protection, does anyone else have any suggestions?

Image: 
Log in to post comments