SJM wrote:You can build a thermofilter like this...

For those who are interested, Susan elaborated on this point in Thermofilter Build.

I find it much more useful to use something that can log and graph the entire temperature profile during brew. One such device can be found by searching ebay for "USB thermocouple".

I noticed one from overseas in a post you made at coffeegeek, how is the accuracy on that one?

I also found this one (http://sine.ni.com/nips/cds/view/p/lang/en/nid/208177)

It works fine, although I haven't compared accuracy to anything that I'd trust (ie, NIST traceable). I usually use a P1000 RTD read by a microcontroller, but this isn't suitable for most people.

So I made the device from the Gaggia group. For my machine, I needed a much larger hole in the basket to simulate flow but other than that it works well. I am worried about needing to calibrate my group measurement device due to using a cheap dual input ebay thermometers and a K-Type Thermocouple as my temperatures seem a little cool.

Does the thermometer matter significantly in this regard? The ebay one claims to have good accuracy but I'm always suspect of such products and was considering an Amprobe TMD-56 (I would love a Fluke but not enough used to justify the price). I have been trying to determine the best thermocouple but there seems to be inconsistent info on their accuracy. It seems T-Type would be more accurate at this level but by a centigrade or so (1c vs 2.2c accuracy)?

I have also been following the DB flush thread but felt this post would be better served here.