[HighdB]

Whilst I have just joined the Speeduino sites and am in the process of setting up my own to run in a couple of cars I have been using Arduino boards for several years at work for Data logging at remote solar sites.
Logging Time, light level, temperature, rain, battery voltages and various pin logic levels to a SD card. They are a self powered unit with a sample rate between 1 and 10minutes depending on what we were looking to capture with a battery life in excess of 3 years.

Over those years some interesting problems were found, our general setup was a Uno or a Teensy with a shield for an SD Card/RTC and another shield as a breakout for Analog and Digital inputs we had an interest in recording. These were for installation on site and left often for 3-12 months to log data.

Initially we found a high failure rate of the units, either locking up, powering off or recording random data that simply could not exist. After a lot of testing we found the culprit to be the connectors between the Arduino and the shields being noisy or intermittent, when these were replaced with a soldered pin between the board and shields the problems all stopped. Whilst the time between failure was high, often months apart it did raise concern over the use of that style of plug for reliable operation. Often the problem could also be reproduced with light finger pressure on the connectors.

I only raise this purely to see if anyone else had seen random problems with no apparent reason and whether anyone else has thought on this.

[PSIG]

Interesting project you had. I have not seen any issues on Speeduino related to the board-to-shield connections. In-fact, of the millions of generic Arduino and similar board projects on the web using similar connections, I have not perceived repetitive complaints or common issues with them.

That said, there are two known issues with most dev boards and micro-controllers in-general, which are "first-read" error and floating inputs. First-read error is the input circuit reading before the input stabilizes, and is why you can find multiple calls to read for the same value in Speeduino code in order to obtain accurate results, even with much shorter time between reads as your logging project. Floating inputs (not attached to a circuit or pulled even slightly high or low) give no signal to read, leaving it to "read" any stray signal like an antenna, such as electrical charges on your finger tip. Again, these are not Speeduino issues, but common and widely-known micro-controller traits that Speeduino has taken into account.

It is fairly common to hear concern that a bare Mega board loaded with Speeduino code reports some very odd results in TunerStudio and is due to the floating inputs described. Once attached to a Speeduino board and sensors the odd signals cease. As Speeduino is open in both software and hardware, you are free to modify or improve the code, or solder the pins or different connections if it has value or just feels right.

David