[Shannon Hicks]

What sensor(s) are you using with the board?  Have you reprogrammed the sketch onto the Mayfly recently, or are you using a board that was programmed last year or earlier?   Some people seems to be having issues with boards not sleeping/waking properly using the most recently SDI12 libraries, so if you’ve recently reprogrammed the board, then that might be the issue.

[Shannon Hicks]

No, LiFePO4 batteries should never be used with a Mayfly board because the onboard charging circuitry is only designed to work with 3.7v Lithium Ion packs.  There’s a summary of important battery information and options here: https://www.envirodiy.org/battery-options-mayfly/

 

[Shannon Hicks]

I haven’t seen any other templates besides the one you linked.  I never use them myself, I just lay them out in whatever pattern is the easiest, based on the sizes of the cable glands.  I prefer to stagger them as shown in this photo from the kit’s shop page, since that makes it easier to get your hand around the end part to tighten them.  I use a socket to tighten the nuts on the inside of the Pelican case because an adjustable wrench can’t usually get in there to properly grip the plastic nuts.

[Shannon Hicks]

Are you referring to an EnviroDIY Monitoring Station kit which contains a Mayfly Data Logger, or is your station based on some other type of controller?  And are talking about limiting the amount of current your monitoring kit could draw from a power source, or are you talking about limiting the output current of the kit in regards to the load the kit can supply power to?

[Shannon Hicks]

There’s nothing electrically wrong with having all 4 resistors populated on that board, but only one pair will be used for your sensor analog input, and the other pair will not be connected to anything and will therefore not be performing any function.

[Shannon Hicks]

If you only have one C-FLUOR sensor, then you only need one pair of resistors, since there’s only one analog signal from the sensor.  There ARE two ground wires, which is why I asked how many sensors you were using with this one screw terminal board.  In order to connect two Turner analog sensors to a Mayfly Logger, it’s good to use 2 separate screw terminal boards, since each sensor needs 2 ground terminals.

[Shannon Hicks]

If your station is in an open location that gets several hours of direct sunlight every day, then a 2W panel is probably sufficient if you have a cellular modem on your station.  The 2W panel is definitely sufficient if it’s just a Hyrdos21 sensor and no telemetry.  But if your station is under tree canopy or is mostly shaded, the P105 model 5W panel is larger better at generating more current with less light.  You can buy them either directly from Voltaic Systems, or from Adafruit, since they also sell the exact same 5W panel (https://www.adafruit.com/product/5367).

[Shannon Hicks]

Are you trying to measure 2 analog signals from 2 separate sensors with just one 6-pin screw terminal board?

[Shannon Hicks]

That happens occasionally when the connection between the Mayfly and the online time server gets terminated before the entire timestamp gets transmitted, resulting in an incomplete timestamp, and the Mayfly code has some smarts to check whether the timestamp it received is a valid, reasonable date.  If you ever get this error, you can just restart the logger, since (with the sketch you’re running) it will always try to contact the time server upon startup.  Once your board gets a valid timestamp, it will set the clock chip on the Mayfly and it shouldn’t forget that time unless you remove the CR1220 watch battery (or the watch battery dies after a year or two of sitting in a Mayfly with no other power source).   Once the clock on the Mayfly has been set, in the future when you restart your Mayfly, you’ll see text on the serial monitor saying what time the Mayfly thinks it is, then it’ll say it is contacting the time server, and then it should say something like “clock is already within 5 seconds of the correct time”, meaning that no adjustment to your clock was necessary.  If you aren’t getting a successful timestamp from the online server, then it probably means your cell signal is weak (or non-existent) or there’s an issue with your SIM card.  Note that when you’re first using a Hologram SIM card with a Mayfly, it can take up to 2 minutes to connect to the time server the first time, and occasionally it can require 3 or 4 reboots and 2-minute waiting periods to successfully connect when using a new card for the first time.  Once you’ve successfully made a Hologram connection, subsequent transmission will only take about 8-10 seconds.

[Shannon Hicks]

Mayfly v1.0RevA had a Max17250 boost chip that used the following resistors in order to set the output voltage at 12 volts:

R15 = 10k
R19 = 84.5k
R31 = 220k

When it was time to manufacture more boards in 2021, there was a semiconductor shortage that made it impossible to find that particular boost converter, so the design was modified to use a MCP1665 instead.  However, the first small run of Mayfly v1.1RevA boards used the resistors above from the previous circuit instead of the correct values below:

R15 = 20k
R19 = 180k
R31 = 470k

The input power selection options for the MCP1665 also didn’t behave like the Max17350 did, so before we shipped out any Mayfly boards from that run, I put a blob of silicone conformal coating on the back of the boards to cover SJ26 to make it unusable and also made a note about it on the schematic.  I also manually soldered a tiny black wire onto each board to connect the power circuit properly, as denoted by the blue line in the schematic and shown in the attached photo below.

So when it came time for the next production run, we opted to call it v1.1RevB since it would correct the problem with the incorrect resistors, add a trace to connect power circuits properly (to take the place of the black wire), and remove solder jumper SJ26 from the back of the board since it was no longer needed.  We also switched the ICSP connector from a small 2×3 0.5mm pitch header to a Tac-connect connector.

Someone asked this question about version differences a few years ago on the Technical Questions thread, and you can see my specific response here.

So in summary, the schematic currently on the website for v1.1RevA was what we designed, however the 3 resistor values placed by the manufacturer are different than specified in the documentation.  This affected only a small number of boards because we quickly made a new production run for v1.1RevB.  However not all of the documentation in various locations got updated with the latest specs.  I’ll go back and add a note to the v1.1RevA schematic to clarify which resistors were actually used on that batch, and will also update the Details and Specs page to more clearly state the specs of the current v1.1RevB board.

Attachments:

1. 

   v1p1revA_jumper-rotated.jpg
   

[Author]

Posts