Sunday, June 2, 2013

Monitoring a Londinium

The Londinium blog recently featured a box full of Amprobe TMD-56's ready to start a series of measurements on this machine that has been a bit of a rave recently. The guys at HG-One like it a lot, for instance.

I have done extensive measurements on the Londinium myself for a while and it has taught me a lot. In the beginning there were one or two fellow users who had less desirable temperature values and I learned how to 'trip' the machine with some unusual procedures into behaving oddly and I also got quite a routine in recovering. It was like teaching oneself the recovery out of a stall in a glider airplane. Back in the years when I trained in a glider, the risk was greater than just having a sink shot from the espresso machine so then it was good I had a qualified instructor in the back ;-)

With the Londinium, the Artisan software did assist me greatly though, in acquiring the ability to see the approach of the stall that I had been invoking.

In the process, I've been connecting probes to different parts of the group. At first very low in the front and on the side,  later further upwards on the front, a little sideways and one cm from the neck of the group where the hot water comes in. The two probes were useful to note the difference of temperatures between the two: if the difference between two points was small and both had a similar rate of climb, the group was functioning properly and any air I had managed to get into the sleeve had been successfully flushed out (short 175ml flushes, 2 minutes apart).

First I had the laptop running Artisan in the kitchen next to the machine, but since I have more work to do and I can't do it all standing up in the kitchen, I bought a 5 metres USB extention cable that promises to boost the signal so no data is lost.

Getting to know the Artisan software took me a day of struggling with details, at first quite puzzled by the user interface but then, once I 'got it' I began to love it a lot. I believe I have only learned twenty percent of what Artisan can do for me and that's a lot already.  Installing the drivers for the proper communication with the Amprobe was a bit of a hassle but following the instructions carefully, step by step, was successful after I had first failed when I did it by guesswork.

Setting the Amprobe to communicate over "Channel 0" was a little trick not mentioned in the Amprobe manual but luckily explained on the website and in a posting I found on a forum: "To SET CH/ID to 00,00, by pressing "T1-T2" key and power key for more than 6 seconds with the meter powered down. The meter will set channel and ID to 00,00 status. The second display will show 00, which means that the channel and ID has been set to 00." It looks scary at first but just following the step by step instructions saved the day for me.

One of the graphs I've been saving of the group nicely warming up. A wobble in the line around 70 degrees Celsius would for instance signal an inadequate heat transfer.

I tried to see how the temperature in the group may be affected by the portafilter being locked in or left off the machine. This doesn't tell you much about the effect on coffee extraction of course. If you leave the PF off, the group gets 1.5 ºC warmer, which could compensate for the moment you lock in a cooled off PF. If you leave the PF on, the group doesn't need to deal with a cool PF just before you pull a shot.

Three espresso shots within 45 minutes and the group idling nicely. Note that this does not tell you anything of the temperature inside the puck during extraction. More about that aspect below...

A shot during a session when I did 13 fast back-to-back pulls. Note the yellow probe cables leading to the group.

Peter van der Weerd of is making me a new basket with a temperature probe but in the meantime I improvised one that could give a few first indicative readings.

Probe cable leaving the PF, fixed in place so moving the PF around will not make the stiff wire change position in the puck. If the puck gets disrupted, so do the readings.

Pulling a shot.

The new basket will have a probe like this:

This specific T-type probe responded too slowly and the people at TCDirect are now making one with the thermocouple connected to the metal sleeve so it will be more sensitive to the specifics of the temperature swings a coffee puck endures during a typical espresso extraction.

As it is, I get readings like this. The spot where I have connected a thermocouple to the group is idling around 80ºC and there's a dip when the PF is taken out as cool air from the kitchen floats by, then at the pull the group warms up and the puck probe is briefly at 94ºC during pre-infusion, then extracting at 93ºC, the group and PF cooling down while I enjoy my espresso.

When I decided to lower the PressoStat from 1.4 to 1.2 bar, I used Artisan again to evaluate the results.

And right after I installed the NRV upgrade I did another measurement, noting that now the group was even more willing to remain at a perfect level for excellent espresso.

Plus a very recent one, probe on back of group: