[James P]

See attached, a picture is worth a 1000 words. This one has the 4400mAh battery. Installed the new micro sd card yesterday.

Attachments:

1. 

   Screenshot-little-spokane-with-new-sd-card-2-23-24.jpg
   

[James P]

Turns out, it was the micro sd cards. We had two units with Sandisk Ultra 32gb cards, and two with different cards. One was a Sandisk “industrial” 8gb card, I forget the other one but it was an older 4gb card. The Sandisk 32gb cards use far more power — there is some info online but I don’t fully understand it. What I know is the Sandisk “industrial” 8gb card uses far less power. All three of our sites that are operating (the fourth got chewed on by a beaver and we just got the sensor cable repaired) are working great (aside from some sensor issues).

This station does have a bigger solar panel and we installed a 6600 mAh battery. There is no voltage drop during non-solar periods: https://monitormywatershed.org/sites/SpokaneR-SpokaneValley/

The stations with smaller solar panels (all provide at least 500mA) and 4400mAh batteries exhibit the same behavior between solar charges, though with slight voltage drops during non-solar periods…but the battery fully recharges every day.

We had a great test week about a month ago where it was either dense fog or snowing, solar was minimal with very little solar charging, and I estimate the batteries probably would have lasted at least several weeks under those conditions.

Here’s the micro sd card:
https://www.amazon.com/dp/B07BZ5SY18

From my research there is a 32gb Kootion card might be good too but we haven’t tested it yet. I think I calculated that a 4gb card could store 2000 years of our data, so 4gb (tough to find these days) or 8gb is plenty.

[James P]

I have a multimeter that can measure amperage. I’m not sure how to get it inline with the circuit though without putting together a couple wires with the proper jst connectors…I’m not an expert in this and don’t want to short the board.
Any suggestions on where on the board or how to test?
Thanks!
James

[James P]

Thanks for all your help. I believe that the bigger solar panel with >1A peak current, a 1A charging current to the board (jump SJ15), and a freshly charged battery should remedy the winter issues.

Regarding consumption, I took one unit home, mostly charged a LiPo 3300mAh battery, here is what I found (no external sensor connected, transmitted via cellular every 15 min, 70F ambient temp, in my office with no solar panel connected):
Starting voltage 3.851
12 hours later = 3.760 (.091 volts difference)
+12 hrs = 3.669 (.091 volts difference)
+ 12 hrs = 3.593 (.076 volts difference)
I ran it another couple hours and it seems the voltage drop slows a bit over time like above between hrs 24 and 36.
According to Shannon Hicks, “Our Mayfly v1.1 boards with EnviroDIY sim7080 LTE board and a CTD sensor will see a decrease in the battery voltage overnight of 0.045v during a 12 hour period with absolutely no sunlight, and transmitting the data to the portal every 5 minutes.”
So our board is using 2x the power even though it’s only transmitting every 15 minutes and had no external sensor connected. Not sure if a battery that’s lost some of its capacity would cause a faster voltage drop?

James

[James P]

I’ve had a 5v 2.0A usb charger plugged in for 3.5 hrs. The yellow charging light is still on. I think it should take ~2.5 hours to fully charge a 4400mAh battery with a 2.0A charger, so it should be done.

Does someone know if the board is getting any feedback from the battery on whether it’s accepting a charge or not? If there is not feedback from the battery, then I think the yellow light is a ‘power is now available to charge the battery’ light.

What’s interesting is that our first solar panel provided 180mA peak current…I think that means an uncharged 4400mAh battery would need over 24 hours of peak sun to fully charge. Given we’re very lucky to get 3 hours of usable charging light in a day this time of year, it makes sense that our batteries are not keeping up. The bigger solar panel we bought has 1,670mA. So it will come down to consumption during the approximately 22hours a day that has no sun (and typically cloud-obscured sun). And maybe a larger capacity battery to get us through the lean weeks.

[James P]

Thanks! So the yellow light was on because solar was charging, that’s good news. I assume that light is powered by the sun and not the battery.

Our next step will be to replace the battery if the full overnight charge doesn’t work. I’ll set those two DIP switches to off after I try to fully charge the battery.

 

[James P]

Yesterday I attached a powerbank to the usb-c port for a little over two hours. Two yellow lights were on, one near the usb-c port and one lower down. I also installed a 10w solar panel. When I detached the powerbank, the yellow light near the usb-c port turned off but the other one stayed on. The station held voltage for two hours after that, then dropped to 3.5v volts.

Does anyone know what that other solid yellow light is for? Would like it to not be on and drawing power, unless it’s the “solar charging happening right now”” light.

I am thinking about pulling the unit, bringing it home and charging it with a phone charger (5v 2A) overnight. Maybe my powerbank didn’t quite get the battery over the hump. It’s older and probably not a ‘fast’ charger.

Attachments:

1. 

   2yellowlights.jpg
   

[James P]

I don’t believe any lights are on during the 15 minutes between data transfers. In the summer and fall lights would start blinking every 15 minutes, then stop for 15 minutes, etc. I’m not sure if there was something happening during the 15 minute break. Once I get the battery charged up with a powerbank i’ll check the dip switches.

Crazy, today at 1:30pm the battery voltage read 4.958. 15 min before that nothing, after that nothing. It’s been cloudy and snowing all day.

I don’t understand how the battery voltage could go from low to max to low again in 30 minutes….could something be going on that would drain the battery in less than 15 minutes? The last time i was out there i opened the otterbox and it looked like every light was randomly blinking, nonstop. But after I reconnected the solar panel that didn’t happen.

 

[James P]

Thanks for the data, Our site has much more exposure to sun than in a redwood forest, and  is doing something similar:
https://monitormywatershed.org/sites/SpokaneR-SpokaneValley/

Note the times on the website are accurate but when you look at the data the times are not accurate — sun isn’t up at 5am. I wonder why the clock reset..
The last transmit via cell was 2:45 pm on 11/27. Yesterday Nov 28 was a full sun day all day…so it seems that that solar panel doesn’t run anything, all it does is charge the battery, and the battery never got charged enough to transmit via cellular.

We have version 1.1 boards.
The cell module says SimCOM – SIM7080G, P/N S2-108HB-Z308J
Our only sensor is a Hydros 21 CTD.

It sounds like the bigger solar panel might do the trick. I also saw in another discussion going to a 6600mAH battery. So this panel https://voltaicsystems.com/10-watt-panel-etfe/ with this battery https://www.amazon.com/dp/B0137IPVY6. And get the battery charged at home via usb before bringing it out into the field.

Though for the good of the order I wouldn’t mind also trying 4 AA 1.5v energizer lithiums into the usb port and see how long they last. Something like this https://www.amazon.com/dp/B08P1L6JQT with a usb adapter https://www.adafruit.com/product/4536.

[James P]

Hi thanks for responding. I’m sure our system is undersized for our location. Today was a pure sunny day, last week we installed an extension pole and cable so the solar is up plenty high to get direct sun. But today we got no readings sent via cellular.

Our battery is a LiPO 4400mAH 3.7V.

The solar panel is a 1W 6V
Open Circuit Voltage: 7.7V
Peak Voltage: 6.5V
Peak Current: 180mA
Peak Power: 1.2W

Trout Unlimited’s guide says that any reading above 4.0V is a fully charged battery. But this winter we’ve seen the reading above 4 and an hour later no more data. So it doesn’t seem like the battery is truly fully charged because it would last more than an hour.

If anyone has a recommendation on the max solar panel and battery I’d appreciate it. The other option which I think would work is to get a pack that holds 6 or 8 AA energizer lithiums and use that, but we would need to program the mayfly to not draw any juice until reading the meter or transmitting data. And we could maybe set it to send every hour instead of every 15 minutes. I’ve had game cameras sit out all winter and take hundreds of pictures and videos without draining those energizer lithium batteries. I don’t know what a game camera draws while not taking a picture, I know the clock stays accurate.

Thank you!
James

[Author]

Posts