ncjay
Junior Member
Posts: 71
|
Post by ncjay on Jun 19, 2021 14:48:41 GMT -8
John, thanks so much for your analysis and proposed solution to this situation. I'm still wrestling with the problem (using the existing hardware I have) and here's what I've found. a) if I use USB to charge the batteries to the point where the green LED comes on, I get about four days' worth of use out of the sensor before the battery is depleted again. I know the AQI uses a significant amount of power, but even placed in the sunniest part of my yard, the solar panel cannot keep the battery charged. There's a small bump on the battery voltage graph when the panel is in full sun in the morning - we're talking .05V or so - but the voltage always winds up the day below what it was in the morning. (See attached screencap) I guess this could be due to a less-than-robust battery, or something in charging circuit or the solar panel itself. I have a second battery (currently allocated to the Lightning sensor, but temporarily not in use) that I'm charging up now and will swap in as a test. b) I've ordered what I need to perform the low power modification described in the AQ assembly and test manual. I'm hoping that the combination of this mod and possibly a swapped-out battery may give me operation for longer than four days. It's certainly sunny enough here in NC to keep solar devices charged under normal circumstances (for example, the WeatherRack2 is mounted in the same location and has no trouble). If this is a workable temporary solution, I'll be happy to wait for the eventual Sun Air 3 board to handle those just-in-case situations. c) I had mentioned early on that I couldn't move the AQ sensor out to that position with the WeatherRack2 because the signal from the transmitter didn't make it into the house using the WeatherRack2 kit stock antenna. I had a spare RTL-SDR stick, and a spare port on the active splitter fed by a discone antenna up about 25 feet and close to the instruments' location. This combination provides a robust signal, so I should be able to move the sensor(s) where I need to. More excitement awaits - I'll follow up with my results from (b) above. Thanks ncjay
|
|
|
Post by SDL on Jun 21, 2021 10:21:42 GMT -8
ncjay,
You are rocking. What size and brand battery are you using? A link would be good.
The low power modification will absolutely work. We are using that in our test unit and it's been running for months.
BP
|
|
ncjay
Junior Member
Posts: 71
|
Post by ncjay on Jun 21, 2021 13:01:16 GMT -8
It's exactly the 6600 mAh one recommended in the manuals, from Adafruit.
I found it interesting that the battery in the Lightning Sensor took two full days to charge via USB, while the one I have in the AQI Sensor only took a day. (It's the same battery in each unit.) Does this indicate anything of interest or concern?
I still have the Lightning Sensor on the bench here while awaiting the new Thunderboard. And the AQI sensor is there too, awaiting its low power mod. I'll keep you posted.
Thanks ncjay
|
|
|
Post by SDL on Jun 24, 2021 14:44:17 GMT -8
Show me the current chart. I'm thinking that the solar panel is working, but I need to see the current on the AQ board.
Not sure about the lightning sensor. Same exact battery? You indicate it is. Can you swap the batteries and repeat the experiment?
BP
|
|
ncjay
Junior Member
Posts: 71
|
Post by ncjay on Jun 28, 2021 15:53:11 GMT -8
No need to swap the batteries because - I performed the low power mod. I did it five days ago, and there's been mixed clouds and sunshine since then - and my battery voltage is now a nice horizontal line. Screencap is attached. I consider this fixed, and thanks so much for your help. Maybe the low power mod should be the default configuration for new units? Thanks ncjay
|
|
|
Post by SDL on Jun 30, 2021 14:48:59 GMT -8
ncjay,
The reason it is not the default configuration is that it requires soldering. And that is a big step for some people. But, after you learn how, it is hard to see how you got a long without it.
BP
|
|
|
Post by Rick@oti on Sept 4, 2021 18:07:58 GMT -8
What's the latest on the SunAirPlus3 board? Thanks.
|
|
|
Post by SDL on Sept 5, 2021 11:32:21 GMT -8
Being manufactured now. Fingers crossed.
BP
|
|
ncjay
Junior Member
Posts: 71
|
Post by ncjay on Sept 13, 2021 10:10:02 GMT -8
We had a run of four days starting September 7 with days 1, 3, and 4 with overcast skies. The battery ran out of power just after sunset at 19:47 on September 9, and the sensor didn't come alive again until 10:57 on September 10. Uptime since then has been as follows: September 10 on: 10:57 off: 18:36September 11 on: 12:02 off: 14:52September 12 on: 12:34 off: 14:50September 13 on: 12:35The sensor is in the sunniest part of the yard (along with other solar-powered instruments), with direct sun from around 09:30 until sunset.
So I guess the sensor has defeated its second battery. I'm going to take it down after it goes offline again this afternoon and give it a charge via USB, but really, this is unsustainable. Fingers crossed, as you have recommended, for the SunAirPlus3 board. ncjay wake forest nc
|
|
|
Post by SDL on Sept 13, 2021 14:44:21 GMT -8
SunAIrPlus3 is being manufactured as we speak! I can send you a prototype, which would do a better job and would be a good test. I've got a couple of boards coming in later this week. BTW, how hot is it getting? I had the Palm Springs beta unit kill a battery (it was AfterShock, but same thing).
I'm struggling to understand what is going on here. If you turn the switch on and off on the Sensor, does it start working again?
If you charge the battery (and it looks like it is already at 4+ Volts) does it start working again?
On your Mini Pro Plus board, is the watchdog timer switch turned to on?
BP
|
|
ncjay
Junior Member
Posts: 71
|
Post by ncjay on Sept 13, 2021 16:04:44 GMT -8
Well, I just brought the unit inside to charge it via USB. The first thing I noticed was that the blue LED on the Mini Pro Plus board was lit, and that the red and green LEDs at the top right of the SunAirPlus2 board were both lit. And they still are, at this point, before unscrewing the board to plug in the USB. I don't remember this to be the case in previous iterations.
Doing a power cycle on the Sensor as you suggest appears to have no effect. The sensor is not getting unusually warm - at least, not Palm Springs-warm; the highs have been in the upper 80s here the last week.
However! The watchdog timer switch has been turned to OFF, i.e. away from the "JP9 wDog Out" legend. Is this going to be the solution to the problem? How does the watchdog timer work? At any rate, I've turned it to ON now, and I'll be charging the unit up overnight. I'll report back once it's charged up. (That was successful with the previous battery, but only in the short term. This will be the first time the new battery has been (re-)charged.)
I appreciate the offer of a SunAirPlus3 prototype, and would be happy to help test it. At this point I'm just doing a lot of testing anyway....
EDIT: This took no time at all to charge - around an hour or so for the LED to go white. When I gave the battery its first charge (right out of the Adafruit shipping box) it took much longer. I guess I'll just turn it back on in the morning, put it back outside, and hope for the best.
EDIT #2: The sensor didn't come online at all today. I opened it up and the red and green LEDs are lit, as is the blue LED on the Mini Pro Plus board. At this point all I can do it let it stay out there, I guess, and hope the eventual SunAirPlus3 board (or perhaps a different/additional suggestion) can help get me going again.
thanks ncjay
|
|
|
Post by SDL on Sept 15, 2021 11:59:49 GMT -8
Ncjay,
Here are some answers:
1) I really would like to know the voltage on your battery when your sensor dies. Do you have a voltmeter? You can measure from the pin next to the battery to the ground pins in the middle of the bottom part of the board.
2) The watchdog timer may very well have something to do with the non-recovery when the battery charges up. The software periodically "Pats" the Watchdog so it doesn't time out (about 60 seconds). If it does time out, then the assumption is your program has gone bye-bye (which happens to computers that don't get good power!!) and resets the processor.
3) I should have these prototypes back in a couple of days. PM me your shipping address.
4) Check to see if you have dislodged the 433MHz transmitter from your board.
BP
|
|
ncjay
Junior Member
Posts: 71
|
Post by ncjay on Sept 18, 2021 12:10:59 GMT -8
Thanks, BP. Here are my results.
I've had no readings since September 13. I brought the sensor inside, opened it up, and noted the red and green LEDs on the SunAirPlus2 board were both lit as normal. The blue LED on the Mini Pro Plus board was also lit.
I corrected the watchdog timer switch setting, and reseated the 433 MHz transmitter board (it seemed fine as it was). I gave it a power cycle, and ... no readings. I did notice an occasional series of four or five flashes from the yellow LED on the Mini Pro Plus board - that LED is located just below the 433 MHz transmitter.
I charged up the battery (again, for good measure). It only took an hour or so. Power cycled and still no readings.
I took a voltage reading as you described, and measured 4.15 volts. This corresponds with the last reported voltage on the AQI graph page from Sept 13.
I've PM'ed you my shipping address for the prototype SunAirPlus3 board. Thanks again for your consideration.
So at this point I'm at a loss. Is there anything else I can check?
thanks ncjay
|
|
|
Post by SDL on Sept 20, 2021 8:19:10 GMT -8
sounds like the solar system is working correctly. Even if the AQI is not detected, you should be sending out packets. Can you recycle power again, but run the testWirelessSensors.py file and see if you are getting data?
BP
|
|
ncjay
Junior Member
Posts: 71
|
Post by ncjay on Sept 21, 2021 12:50:58 GMT -8
No readings from the AQI sensor from testWirelessSensors.py after running for over an hour. I'm only seeing the WeatherRack 2 and the F016TH sensors.
I'm out of ideas. How about you?
thanks ncjay
|
|