Cheap digital thermometers
 

Hi All,

Earlier this year I build an HRV but I had no good way to measure it's efficiency. To accurately measure that I need exact readings from 4 spots at the same time but I only had 2 simple thermometers.

After looking around I decided to order some cheap digital ones from a far country (dx.com). Since they were fairly cheap with free P&P I decided to order 10 so I could surely find a set of 4 that would agree with eachother.

To my pleasant surprise the package was delivered to my door in only 9 days.

After unpackaging everything and inserting the batteries (which gave me a good feel for the cheap buildquality of the items) I build a contraption to see if the 10 thermometers agree with eachother.
With some foam I put the sensors thightly stuck in a piece of aluminium rail, put it all in a nondraughty place and let it sit for an hour.

I found 2 sets of 4 that nicely agree with eachother which will let me make accurate calculations on my HRV, see attached thumbnails.

Although the 10 thermometers do agree with eachother the actual temperature measured with 2 known-good ones (also visible in the images) is a full 2 degrees centigrade lower...

"E=Fornax;29680]Although the 10 thermometers do agree with eachother the actual temperature measured with 2 known-good ones (also visible in the images) is a full 2 degrees centigrade lower...[/QUOTE]

I think your results are amazingly consistent.

If I'm not mistaken, these thermometers use 1-wire sensors which are advertised to be within 1/10th degree C, so it looks like your thermometers are pretty close to that.

You might try the same setup at different temperatures, to see how well they track. Use boiling water (100 C) and ice (0 C) as extreme points.

BTW, I have been doing some experimenting with a high quality Temp/Humidity sensor, and I am quite pleased with it's results.

In my heat pump work, I have become very disenchanted with 'digital'. I now know that it just means 'numbers, and not necessarily accuracy, unless forced to be accurate. Don't be fooled by decimals, they aren't a sign of accuracy, they are a sign of decimals.

I have tried a very large array of temperature measurement tools... cheap thrift store digital thermometers, non-contact infra-red, alcohol column, mercury column, bi-metallic, analog electronic (thermistor, lm-135, thermocouple), digital electronic (1-wire).

But when everything seems to no longer make sense, I have one lab-grade mercury column sling psychrometer that has two columns and even though it will not give me resolution to 1/10 degree, I really trust it.

I think that your selection process is really a good approach... but, be careful, you might become a scientist.

Best,

-AC

/grin

For the measurements on my HRV I am looking for resolution, plus 4 thermometers that agree with eachother, and I have 2 sets now.

I fully agree with you that 'digital' is by no means a guarantee for precision nor quality.

The 'known-good' thermometers I referred to in my OP are alcohol columns and they agree with the roomthermostat that came with my central heating furnace. These 2 were €1,95 each ($1.50) but they lack the resolution I want for my purpose.

Your suggestion to calibrate these elcheapos with melting ice is a good one that I thought about, but forgot to actually do. I opened one of the elcheapos to see if there was anything adjustable inside but there wasn't. Probably the sensor and the electronics are fine but there is no compensation for the wiring and the solderingcontacts.

There is a reason that high quality industrial PT100 sensors use a total of 4 wires to compensate in realtime for resistance issues and temperature differences, see: PT100 on wikipedia

Fornax