Looking for a test of actual E61 dual boiler brew temperatures - Page 2

Need help with equipment usage or want to share your latest discovery?
mx125 (original poster)

#11: Post by mx125 (original poster) »

Jeff wrote:Perhaps the most reasonable option is to accept a one or two degree Celsius shot-to-shot variation as "close enough". Even though I could control my E61 HX to within a 1°C range and my current machine similarly, I don't find changes much less than 2°C to make an obvious change in the cup.

There's a balance between cost and performance. E61 castings and parts are dirt cheap compared to engineering and fabricating a new group. There's no need for tight tolerances as there would be with a close-coupled group that needs to mate to the boiler at 9-bar working pressure. A thermosiphon sort of works with the relatively low temperature differential (maybe 5-10°C for a dual boiler, compared to around 30°C below 123°C for a 1.2 bar steam boiler HX), well enough that I don't know any manufacturer that has a circulation pump.

My guess is that the technology will come from two fronts. One, like the BDB, how to mass-produce a temperature-stable espresso machine at a premium, but not outrageous price for the high-end kitchen appliance crowd. The other, I'm guessing, will be energy (cost of operations) savings in the commercial market. Smaller boilers, better-insulated groups, flash heaters, and integrated energy management seem to be a way to get some significant savings for the commercial operator. The VA Eagle One series is, I think, just the first of these. My crystal ball isn't clear enough to know when that will make a big impact on the hobbyist-consumer market. There's still the attraction of decades of posts on the Silvia and E61s coupled with the "that doesn't look like an espresso machine" factor that is a big driver of, I'm guessing, 90-95% the machines out there above the basic, "15 bars" buzz-box level.
Also helpful and somewhat comforting.
Particularly interesting to see that thermosyphon should still function on db. I've seen some dismiss value which I didn't quite understand. There is still a variance which drives flow, albeit lower delta from hx boiler..... which should separately be a shortcoming of hx.

And your view of development path is logical and likely. Fair to say the saturated (or otherwise ) premium tech is either a long way off at our accepted price point or more likely in the very long to never realm. At 3000 cdn for hx and 4000 cdn for db .....7500 cdn must be a very small home buyer universe.

DaveC

#12: Post by DaveC »

mx125 wrote:The second link in particular tells a story that makes sense.....
The heat spilling from e61 means a big delta vs boiler. (I guess the thermosyphon is not very effective?) and hence running shots simply raises the temp in succession.
The thermoregulation of the E61 depends on a thermosyphon, for that to work, there has to be a temperature delta between the group and the boiler...the larger that delta, the faster it goes. The group looses around 85 watts continually at brew temperature, so the thermosyphon has to provide that energy and flow fast enough to do so. For this to happen, a certain temperature difference must exist. This difference is larger or smaller, depending on then efficiency of the thermosyphon, pipe size, flow resistance etc.. For any machine, a balance is struck between brew temp and boiler temp...which leads to a particular offset for that machines design..

When you pull a shot, the thermosyphon no longer runs and the shot water is cooled by the group resulting in the desired brew temp..If the offset is 12C, then a small amount of energy is added to the group based on the temperature differential. Once the shot has finished, then the thermosyphon resumes...if the group ends up very slightly hotter than it was before, the thermopsyphon simply runs slower until the previous steady state was achieved. usually by the time you are ready for your next shot, the system will be stable again.

The big problem comes when manufacturers set up their machines and do temperature testing (not all do). These tests are often a series done in a lab, don't reflect the way someone would actually use a machine and tend not to perform well in the home. When I come up with PID settings for manufacturers, it typically takes me 5-7 days of testing. It's based around how machines are actually used e.g. sometimes hours between shots....
★ Helpful

mx125 (original poster)

#13: Post by mx125 (original poster) »

Thanks.

So in your experience, even though the convection of an e61 DB is lower, it still does produce a consistent temp management with repeatable recovery and shot to shot profile? I see your point on the differential mapping.. However, once complete, your tests support the prior embedded link data with a ~1c shot to shot temp profile ?

Separate topic but I don't see any of the inexpensive "saturated groups" with group head simply attached to a small boiler in any way comparable to a true integrated / Saturated commercial or LM design. That's a whole different thread which has been covered but a "Silvia" as example is a disaster of instability in a long list of design elements. No relationship between cheap "saturated" and commercial saturated/integrated.

User avatar
cafeIKE
Supporter ❤

#14: Post by cafeIKE »

from a long while back on the original Vibiemme DB, after cold water injection mod which was implemented in later versions:



Not all e61 DB are cut from the same cloth, so the graph is only representative of one machine

Most users will struggle to make shots every two minutes or less, so this is worst case. Profile [droop/rise] can change slightly with scale accumulation.

I'd love a BttW saturated group, but the hideous factor I'd have to endure from the missus keeps the e61 in the kitchen instead of the shop 8)

PeetsFan

#15: Post by PeetsFan »

mx125 wrote:The second link in particular tells a story that makes sense.....
The heat spilling from e61 means a big delta vs boiler. (I guess the thermosyphon is not very effective?) and hence running shots simply raises the temp in succession.

That makes sense and unfortunate. The question though ( not to create new topic ) is what are the alternatives. I had a La Spaziale. Different boiler hybrid and I understand not significantly more stable.

The LMLM is awesome but a LOT of money. A different class. I don't love my e61 reality but don't see much I could have done in prosumer space. No amount of BDB performance stats or value comparison will get me onboard.
Bezzera has MN and DE models, and the DE models have PID-controlled, electrically heated group heads. So that's an alternative.

boren

#16: Post by boren »

mx125 wrote:I don't love my e61 reality but don't see much I could have done in prosumer space. No amount of BDB performance stats or value comparison will get me onboard.
I'm in a similar situation. I want to move away from E61 due to its poor temperature stability (at least on my machine, see the first link in Ryan's comment). I don't like the style of the BDB, but it seems great otherwise so I'm still considering it. The other machine I'm considering is the Lelit Elizabeth. It provides excellent temperature stability, thanks to boiler sitting directly on the group. Based on this test with a back-to-back shots and a Scace portafilter, temperature stability more than satisfactory:

PID target: 95c

Shot #1
@10s 94c
@30s 94.7c

Shot #2 (6 min later)
@10s 93c
@30s 95c

Shot #3 (2 min later)
@10s 93.9
@30s 95.8

Shot #4 (2 min later)
@10s 94.3
@30s 95.6

Shot #5 (2 min later)
@10s 94.3
@30s 95.7

Shot #6 (2 min later)
@10s 94.3
@30s 95.5

Shot #7 (2 min later)
@10s 94.8
@30s 95.9

This is much better performance than what I'm getting with my single-boiler E61 machine, PID and all. If other E61 machines, and specifically the Bianca V3 could come close, I would definitely prefer them over the Elizabeth, but as you suggested in the OP, finding test results is not easy.

User avatar
HB
Admin

#17: Post by HB »

boren wrote:Based on this test with a back-to-back shots and a Scace portafilter, temperature stability more than satisfactory...
Keep in mind that the Scace thermofilter itself may flatten the profile given the heat transfer from the "pseudo puck". It's an old plot from 2005, but demonstrates the point:


See Measuring brew temperature for related discussion

So the range deltas of [0.7, 2.0, 1.9, 1.3, 1.7, 1.2, 1.1] above are likely larger in actual use. Whether that impacts taste is another matter (I am skeptical that it does).

A couple other points worth mentioning:

When many posters write "stability", they mean a flat temperature profile (= small delta between start and end temperature), not repeatability (= same temperature at predetermined point). I personally am more interested in repeatability than a specific brew temperature profile shape. The Scace thermofilter also has a slightly faster flow rate than most real pours. It's a great calibration tool, but wasn't designed to precisely replicate an actual coffee puck (see WBC Procedure for Measurement of Brewing Water Temperature for a lot more detail on the design and limitations). If you want to measure brew temperature in use, a simple over the lip thermocouple will do.
Dan Kehn

boren

#18: Post by boren »

HB wrote:Keep in mind that the Scace thermofilter itself may flatten the profile given the heat transfer from the "pseudo puck".
I would assume that even with these methodology limitations the results would still be comparable if the same method is used to test two different machines. Practically speaking, I'd like to know if the Bianca would provide results that are more similar to my E61-SBDU machine (where one shot may average temp be 93c and another, two minutes later, be 89c), or if it would be closer to the (excellent) results I quoted in the previous comment for the Elizabeth.

I also care more about temperature repeatability, as long as stability is reasonable. If it changes from 92c@10" to 94c@30" (or the other way around) it's an acceptable 2 degrees range. If it changes 4 degrees in that time period it's definitely more than I want or expect from a machine that costs this much.
If you want to measure brew temperature in use, a simple over the lip thermocouple will do.
That's how I measured the performance of my machine, but it's not a great solution. The thermocouple prevents a proper seal so there's coffee dripping from the side of the portafilter during extraction. It also probably makes the grouphead gasket not last as long as it would otherwise. On the other hand, it beats buying a $600 gadget that I'll only use a few times.