Different sensor reading on Nano vs. Uno?

I've just moved an almost-complete project from my Uno to a Nano (clone), which I intend to install and dedicate permanently to what I've built.

There are two temperature sensors in the project which connect to A2 and A3. Now I've moved from to the Nano, my readings appear about 50% higher. What could be the reason for this?

Thanks in advance.

They are both running 5V, 16 MHz, and using the same Aref level? (Internal Vcc, internal 1.1V, or external level)

I haven't specified the Aref level in my code, and there was nothing connected to the Aref pin on the Uno. Do the two boards default to different values?

No, I think both will use the internal 5V reference as the default. The Nano clone is 5V, 16MHz also?

I have tried setting the reference to 'DEFAULT' in my code, and 'EXTERNAL' with the 5v pin connected to the REF pin. The reading is the same. Disconnecting 5v and REF causes the reading to shoot up, confirming it is being used as reference.

Any suggestions gratefully received!

CrossRoads: The Nano clone is 5V, 16MHz also?

Yes, it is 5v 16MHz. Here's the one I bought:

http://www.ebay.co.uk/itm/111210189157

Sorry - another piece of possibly important information...

If I measure the voltage on my Uno it's at 4.94v; on the Nano it's at 4.62v. In percentage terms, that's almost exactly the same difference between the temperature readings on the two boards.

Could this be the cause?

Possibly the regulator on the Nano is not as well calibrated as on the Uno.

I presume that you're getting this with the power coming from the USB, what does the 5V pin actually read if you use the power jack to supply 9V from a battery snap or DC transformer?

This is what the AREF pin is for, to feed a known good voltage supply, from an external regulator for example.

spynappels: I presume that you're getting this with the power coming from the USB, what does the 5V pin actually read if you use the power jack to supply 9V from a battery snap or DC transformer?

...I get 4.94v when it's powered from a 9v battery.

Mystery solved, perhaps? I'll rewire up my project and see if that fixes the readings. Thanks!

Can you limit the readings you are taking to be between 0 and 3.3V? Then you can use 3.3V into Aref with the External setting selected.

JohnLincoln: Was the 50% mentioned in your original post accurate, or just a rough guess?

That was a rough guess. It's clearly hotter in here than I thought...

Have rewired to the Nano and powered from 9v battery. That seems to have resolved my issue!

Thanks, folks - very much appreciated. I think I'll have to lay my hands on a 9v power supply and stop relying on USB power if I need such accuracy.

If you really need accuracy than a real Voltage Reference is needed, such as:

http://www.digikey.com/product-detail/en/REF194ESZ/REF194ESZ-ND/1007337