Datasheets can lie!

So, when working on the permanent version of my weather station setup, I have installed at Atmega 1284p as the MCU, and have a DHT22 and WIZ820io ethernet module installed so far…

So, after getting a web server running on the board, which was reporting the temperature and humdity, it was time to power it standalone for the night… easy… just plug in the 5v power supply… or so I thought…

Plug in the power supply… and nothing happens. The WIZ820io lights just flashed for a moment and nothing else happened.

Plugged in the FTDI programming cable again and plugged in back into the computer… and it worked again! Strange. So I plugged in a external USB power supply, and powered it via the FTDI cable… and it also worked. So I left it overnight, and it was still happily running the next morning.

From the WIZ820io datasheet

Upon checking the datasheet… it reports that the WIZ820io module is 5v tolerant (with an absolute maximum of 5.5v). No problem… the power supply is outputting 5.3v, so is still within spec… or is it? So, I put it on a adjustable power supply, and run it at 4.8v… runs without any issues (and so it should… FTDI cable on the computer was outputting 4.7v) Anyway, ramped up to 4.9v… still ok… 5.0v… and the lights on the WIZ820io go out…

So it would appear that the 5.5v (max) tolerant WIZ820io didn’t like getting any logic signals above 5.0v! Since I didn’t want to use a fancy power supply… I just added a 1N4004 diode in line, and the resulting 0.7v voltage drop fixed it up perfectly. So the 5.3v power supply is now running my raspberry Pi and weather station as intended now!

Work in progress weather station / datalogger


Posted on 19th August, 2013, in Uncategorized. Bookmark the permalink. Comments Off on Datasheets can lie!.

Comments are closed.

%d bloggers like this: