Hi... I'd like to get a rough idea of what's going on temperature-wise with my machine. I've read about Eric's thermocouple mod but I'd like to start by not making any changes to the machine (partly because I'm not handy, partly because of its newness and its being under warranty).

Can someone give me a procedure and a parts list for making a PF with thermocouple, including recommendations for suppliers of the various elements?

Thanks in advance for any advice!

A bottomless portafilter gives you a number of options. The first is a Schomer basket: embed a thermocouple in a regular basket (drill a hole, put in the TC, and seal with epoxy glue), fill it with ground coffee and pull shots as usual, reading the temperature. The second is to get a Scace thermofilter -- a basket with a flow restrictor and a TC so no coffee is needed to get the flow to espresso making rates. The final option is to make a DIY restrictor basket by putting TC in a basket and sealing all the holes but one. This doesn't work as well, since you'll get 62.5 mL (the amount of a double including puck absorption) in about 15 seconds instead of 30.

Jim, can you recommend a particular thermocouple, and also temperature readout device?

A T-type thermocouple will be the best bet, its temp range is right where you want it for 'spro, and depending on what one you get, they have a very fast readout. The one which comes with the scace is a .25sec readout, IIRC.

A fluke is the best thermometer to use, but any therm will work as long as it is compatible with the TC you choose. Remember, you'll get what you pay for here. Buying a cheap TC and Therm will not give you the accurate readings during the shot you're looking for.

My personal opinion is that for home espresso, the whole T-type, Fluke thing is so far over the top that even binoculars won't see it, you'll need the Hubble.

The Scace thermofilter is designed for qualifying machines for Barista competition. This requires high accuracy and repeatability to be fair and meaningful when comparing machines to others measured at other places by other meters. At home on the other hand, if your readings are slightly biased, the bias will be repeated on all your measures, so as far as understanding the effect of changing temperatures, it will make no difference

You can get K type thermocouple digital multimeters, with the TC for about $15. You can get 2 and 4 channel temperature loggers, that talk to computers via USB, for around $120. These will only be plus or minus 1/2F providing you've calibrated; but they won't be Fluke or T type. Google will turn up lots of good alternatives. The cost savings are quite dramatic.

On the other hand, while your espresso will not be improved by high end measurement gear; it may give you peace of mind knowing exactly how hot your good, bad and ugly shots are.

another_jim wrote:You can get K type thermocouple digital multimeters, with the TC for about $15.

I already have a Fluke for my 'day job' and purchased the fairly inexpensive Thermocouple adaptor (80TK) so that I wouldn't have to blow a buncha cash on another meter, and got a K t-couple. It seems a bit slow, and someone gave me a T t-couple to try. Now, the 80TK is designed to work with a K-type, what would I expect to see with a T-type T-couple? A one to one off-set? A percentage of the total as an off-set? No offset?
I'm getting much faster readings from the T, but they don't agree with the K.

BTW, if you make part of your living with a meter, I cannot recommend Flukes enough. If you don't, I cannot agree with Jim more.

FYI, below is an example "Schomer-type" portafilter for measuring temperature:


From pages into Scace Thermofilter Temperature Device

The thermocouple juts out from the bottom of the basket, ending approximately at the same level as the puck's surface. It's not as repeatable across equipment as a Scale thermofilter, but I used one and over-the-lip measurements (36 gauge wire) for years with good results.

Seems like you ought to be able to rent a Scace/Fluke rig.

Psyd wrote:Now, the 80TK is designed to work with a K-type, what would I expect to see with a T-type T-couple? A one to one off-set? A percentage of the total as an off-set? No offset?

None of the above.

The error varies with the difference between the temperature at the thermometer (cold junction) and the temperature at the sensing end (hot junction). Unfortunately, the error is not quite a linear function of the difference.

Up to around 125F, the error is no more than 1 degree.

At espresso measurin' temps, though, the error is between 6 and 8 degrees. As the temperature at the sensing end gets higher, the error gets larger.

Here's a table and a graph that might help.




Jim

Back in my BS times (Before Scace), I took a double basket, drilled a small hole in the lower portion of the side-wall. Ran a run of the mill K thermocouple through it and epoxied the entire works in place. Put a piece of duct tape over the basket holes from inside the basket (where would the world be without duct tape). Then I used a sewing pin to poke a hole in through the tape in one of the basket holes. That slowed the flow rate down, still a little fast but much better than the full flow from an open double basket. It worked just fine for a while. Eventually the epoxy started to leak and had to be re-glued, then it would leak again, reglue, etc... You also have to use a bottomless portafilter. Botched together, yes, but it did work good enough to get me an idea of what was going on. I used my multimeter which had a hookup for a K thermocouple. You can also get very inexpensive meters with the thermocouple on ebay. It is no Scace, but was good enough for home use.