Don't Ask...

It would appear that the Velleman VM110 isn't all it's cracked up to be. The calculated voltages are somewhat adrift from the measured voltages, which means that the temperatures recorded by my datalogger for the thermal system are a little adrift. Not the end of the world as I have the new DACIO card to go into the system shortly, but all the values recorded so far are just a little on the high side. That means on good days, the system hasn't been as spectacular as I originally thought, and on bad days, the output has been even lower.

With an error of about 0.2v at worst, this corresponds to an error of almost 19 Celcius on the logger. 10mV/K from the sensor, or 10mV per degree...divide that into 0.2v and there is the magnitude of the error. Not only, but also, the errors are worse at higher temperatures...nonlinearity.com...grr.

I did struggle with the DACIO today. The unit has a PP3 battery snap as it's designed to run from a 9v battery. I 'mocked' up a battery using a spare snap and an old power supply I found sculling about, and although it worked seemingly ok, the analogue measurements were somewhat erratic to say the least. Turns out the power supply was actually a battery charger for a model car, and the smoothing and regulation electronics are all in the car. The PSU is simply a transformer/rectifier and was producing almost a volt of ripple on the DACIO chip. In turn, this was putting a small AC component on the Vref of the ADC's, so the values were reflecting this. A smoothing capacitor sorted the problem out nicely. As the final unit will be running from a steady regulated DC supply originating from the solar batteries, this won't be an issue in the real world.

The moral of the story is, don't expect accurate measurements when your measuring tools aren't up to snuff!

So DACIO tested, I'm happy with my coding, the temperatures should be far more accurate and all I need to do now is construct a decent prescaler to offset and gain the sensors. Lucky the pumping system works on pure differentials and not actual temperatures so the system itself hasn't been compromised by dodgy ADC's...just my measurements. Ho hum.

Oh, and I didn't get the job I interviewed for either. It's been one of those days...