Lack of temperature stability on a Bricoletta - will PID help?

[dawgcatching]

My machine is a Fiorenzatio Bricoletta. I was able to borrow a Fluke multimeter/thermometer recently, and noticed that, while my temperature flush was nice and consistent (a 2-second flush past boiling, 20 second rebound time resulted in a 201 to 201.5 brewing temp every time), I had problems with temperature stability during the course of the shot. After 17 seconds, my temperature inside the puck started to drop off pretty rapidly, decreasing about 1 degree F every 1.5 seconds. I wasn't aware of this previously, and was wondering why the last 5-8 seconds of my shots went blond very quickly, not to mention the bit of sourness that seems to be on the top of a ristretto, which seems to be more evident in the first sip rather than throughout the whole cup.

Would a PID solve this problem? Or, am I looking at something else to rectify my lack of temperature stability? It may just be too small of a machine to be temperature stable though a 25-second shot.

Has anyone upgraded their Fiorenzato Bricoletta with a PID unit, or a similiar size prosumer/small commercial machine? Was it easy to install yourself (or on a similiar HX machine)? Are there any vendors that perform this service?

[terryz]

The Bric is an interesting machine. I like it over all, as the build quality is great. The HX used is a tad difficult to tune however. I have been playing on and off with Chris Tacy's Bric for over a year now, and I can tell you this much....there is no easy way to tune this beast.

I have come up with some suitable resolves however and I hope to publish my findings in the next few weeks. My goal was to tune the HX, so that you could simply start using the machine without a large flush. So far we are at about a .5 oz. flush and stable through a 3 oz pull. The final work will be completed this week.

PID is a possibility, but this tune up of the HX is most likely the answer.

[2xlp]

the blonding shouldn't have anything to do with temp

the sourness on the first sip could be from beans that are too fresh giving a green-tasting crema

how old is the machine? do you have hard water? is it possible that scale buildup is giving you bad heat-transfer in the hx tube ?

[edwa]

I haven't gotten the time to upload the video yet but I had wanted to start a thread similar to yours. I have the E61 adapter installed and have been using it for almost 2 weeks. I taped the same temperature drop while pulling a shot that had 2 shots pulled immediately before it.

I had thought I might have incorrectly installed the adapter or that the Fuji digital controller/reader was reacting too slowly.

I will look forward to terryz's future posting.

[terryz]

the blonding shouldn't have anything to do with temp

the sourness on the first sip could be from beans that are too fresh giving a green-tasting crema

how old is the machine? do you have hard water? is it possible that scale buildup is giving you bad heat-transfer in the hx tube ?

Actually the blonding has everything to do with temp in this instance. The rapid decline of temp on the Bric is interesting and while the machine under tight preparation will make nice espresso it does seem to have a drop during the extraction that is not beneficial to straight espresso.

One thing you could do is increase the overall pressure / temp and do like mike does and flush 4 oz or so.

[dawgcatching]

I was getting the blonding with Hairbender (the normal blend that we use, as my wife works at a Stumptown-affiliated coffeehouse). The cooling toward the end isn't terrible: the shots are typically on par with what I get when I go to the cafe in Portland, but something is wrong in those last few seconds, as confirmed by the Fluke readout. With the Dolce Vivace blend, I was only able to get 90% of the results that I had at the cafe, probably in that it is a very temperature-sensitive blend (it still tasted great, though).

Terry, regarding the tuning of the HX you are speaking of: how is the HX adjustable? Can one control the rate of heat transfer by adjusting the water circulation rate of warmer water? What is the proceedure for this? Is it a hardware upgrade that you are working on for commercial sale?

Also, the flush that you mentioned that Mike was doing: how does that proceedure work? If I flush 4oz before pulling a shot and not letting it rebound, the machine, according to the Fluke, will already be about to dive below the optimum brew temp. Is he doing a longer flush 20-25 seconds before inserting the PF, which in effect allows a greater volume of water to come up to brew temperature? I was flushing to below boiling (at our elevation, about 205.5F) plus an additional 2 seconds, wait 20 seconds, then pull the shot. I tried flushing boiling water plus 4oz 25 seconds before the shot, and the results (according to the Fluke) were basically identical (I haven't tested it by pulling shots yet, though).

[DaveC]

My machine is a Fiorenzatio Bricoletta. I was able to borrow a Fluke multimeter/thermometer recently, and noticed that, while my temperature flush was nice and consistent (a 2-second flush past boiling, 20 second rebound time resulted in a 201 to 201.5 brewing temp every time), I had problems with temperature stability during the course of the shot. After 17 seconds, my temperature inside the puck started to drop off pretty rapidly, decreasing about 1 degree F every 1.5 seconds.

Would a PID solve this problem? Or, am I looking at something else to rectify my lack of temperature stability? It may just be too small of a machine to be temperature stable though a 25-second shot.

A PID would not solve the problem.

I am not intimately familiar with the Bric, but I believe it has Horozontal boiler with HX unit going thru it at an angle vertically. If so you might want to try decreasing the water level a little in the machines boiler, this will expose more of the HX to the steam area and give better heating of the input water. Ensure you don't lower it too much and expose the heating element.

Lowering the boiler water level can be achieved by adjusting the autofil probes depth within the boiler (from outside, takes a few seconds to do once the case is off). Before doing take some measurements of existing probe depth and estimated depth of heating element, so you know where the water level is in the boiler and how much shallower you can make it before heating element is exposed.

The above should still allow you decent brew temps after flushing, but reduce the falloff in brew temp during the shot (it will additionally make the group run slightly hotter). Steaming power will be reduced slightly, but it shouldn't be a problem.

[edwa]

I have to ask the obvious. Would the temp stability increase if insulation were added around the boiler?

Here, I hope are 2 videos showing my temp. decreases during a pull. I'm using 6 day old West Coast Roasting Toro DeCaf. My heat cycle is from 0.91 to 1.1. These shots were made with the temperature read-out in mind so forgive me for the early blonding you will see.

Assuming these went through google's verification you should see some clips.The first clip is of a shot pulled from a long idle of more than an hour. This is without any pre-flush. Basically, I walked up, ground my beans letting the PF stay warm in the group as long as possible. Built my PF, then flush and go.





The second clip is of the following shot after rinse, wiggle, grind and build.




dawgcatching, how do these temperature decreases compare to yours?

Ed

P.S. On a humorous note, you can see a reflection of the TV news that's on a big screen 38 ft away in the chrome of the drip tray.

[jesawdy]

Edwa-

I wonder if you had stopped your cooling flush closer to the 12-13 second mark on the second video, if you would've been a lot closer to the final temp of the first video. In video 2, you start out cool, the temp comes up, overshoots the 210F (at which you stopped the water dance on video 1) to ~212F very briefly and then drops pretty quickly even beyond 210 by the time you stopped the water, and then keeps dropping in the shot. I'd try one where you come up to the 212, lock and load, and see what happens.

Now my HX machine is very different from yours BUT, if I had just pulled a shot, no idle time, I would've stopped my flush no later than the 14 second mark in video 2.

I gather from listening and counting, that you are doing about a ten count after the end of the dance in each video?

[HB]

Thanks for posting the videos Edwa. Although I've not evaluated the Bricoletta, I would caution against drawing conclusions based on the readings from Eric's TC adapter. As noted in Monitoring Brew Temperature - E61 & Silvia, its readings and those of the thermofilter aren't well correlated until the latter half:

The prior chart only showed thermofilter readings. Below is an example of the two location readings shown together. To demonstrate the attenuating effect of the group, the adapter reading was taken from the bottom of the well (nearer the HX exit), not turn "D" as Eric's instructions suggest.


Temperature within the valve chamber exit (blue) versus thermofilter (red)

You can see the difference in the lower reading in the video below:


The long flush are because I had recently done performance tests and the boiler setting was at the highest recommended setting (actually it's still at the same setting because it's so darn easy to nail the brew temperature, I haven't bothered lowering it to shorten the flush). But if I did a flush and go as rapidly as shown in your video at its normal setting, the brew temperature would have plummeted in the second half just as yours did.

Without seeing thermofilter readings (or over the lip TC readings), I'm not certain what to make of your findings since I don't know how well the adapter's readings correlate to actual brew temperature for your machine. For example, the placement of the probe does matter. I intentionally positioned it deeper in the well than Eric recommended because I wanted more volatile readings. I found through usage that I could more accurately target brew temperatures with the deeper probe. This trick may not work with your PID readout because its refresh rate is probably slower.

[edwa]

Hi Jeff,

Prior to installing the adapter my procedure was to flush 7 oz.s from a long idle, build my PF and then flush 3 oz and pull. The next shot only got a 3 oz flush. This came from some feedback I had gotten here and are actually based on a countdown from end of dance.

After I installed the adapter the first discovery was how hot these shots were. More than ever before I feel like I am truly surfing a temperature wave as its never the same. Including back to back shots, though I haven't gone more than 4 in a row. I have to be careful not to end the flush too soon otherwise the temperature never drops far enough. If I pull above 210 the temp doesn't drop far enough and I get a hot shot. With this in mind I have been watching the falling readout trying to gauge when to lock and pull rather than count from the end of the dance.

Perhaps this is the wrong approach?

[edwa]

Hi Dan your post came up after I finished writing my last response to Jeff.

First, any opinions on whether insulation would help?

Interesting to note that your temp readout stabilizes near the end, something I've yet to achieve.

[HB]

I wouldn't expect insulation to make any significant difference.

Although we speak as if all HX machines are operated similarly, it simply isn't true. For example, the Quickmill Andreja Premium / Vetrano perform optimally with a longer rebound than La Valentina, which prefers a modest 10-25 second rebound depending on the pressurestat setting. The Expobar Lever and Elektra A3 are on the opposite end of the spectrum, where the flush and go (no dawdling!) is the only choice at all but the lowest pressurestat setting. By all reports, the Bricoletta performs best with a short rebound similar to La Valentina.

I haven't disassembled these HXs to measure, but if I were to guess, the key difference between these various E61 incarnations with regards to flush regime and temperature profile would be the HX length / immersion, HX volume, and most importantly the injector length. This controls the mixing that occurs in all HXs.

Without knowing anything about your machine and only having seen the video, I would try a slightly longer flush and slightly longer rebound (somewhere between 25 and 30 seconds). That usually shores up droopy temperature profiles for semi-commercial HXs - up to point. Some manufacturers intentionally design in a drop and there's not much you can do about it, the Cimbali Junior being the prime example.

[luca]

Although we speak as if all HX machines are operated similarly, it simply isn't true.

Thankyou!

I haven't disassembled these HXs to measure, but if I were to guess, the key difference between these various E61 incarnations with regards to flush regime and temperature profile would be the HX length / immersion, HX volume, and most importantly the injector length. This controls the mixing that occurs in all HXs.

Also, I'd think that thermosyphon restrictors play a big part. I gather that some US expobars have a completely unrestricted flow, so you have to flush heaps, then extract straight away. If you check out gimme coffee's blog, they mentioned a tricked-out mirage with different restrictors in each head so that one runs cooler than the other.

Cheers,

Luca

[jesawdy]

Prior to installing the adapter my procedure was to flush 7 oz.s from a long idle, build my PF and then flush 3 oz and pull. The next shot only got a 3 oz flush. This came from some feedback I had gotten here and are actually based on a countdown from end of dance.

After I installed the adapter the first discovery was how hot these shots were. More than ever before I feel like I am truly surfing a temperature wave as its never the same. Including back to back shots, though I haven't gone more than 4 in a row. I have to be careful not to end the flush too soon otherwise the temperature never drops far enough. If I pull above 210 the temp doesn't drop far enough and I get a hot shot. With this in mind I have been watching the falling readout trying to gauge when to lock and pull rather than count from the end of the dance.

Perhaps this is the wrong approach?

edwa-

I just wanted to add that my HX temp management at the moment is by the "seat of my pants", no thermocouple adaptor, no temp meter, no thermofilters. Just sight, sound and taste.

When you use your old technique for your second shot after no idle (and ignore the temp readout) do you really notice a difference that you can identify or taste?

[edwa]

Jeff,

In all honesty, no, the taste doesn't suffer. An additional variant has been the lack of my usual blend. The roaster wasn't happy with his supply of beans from Brasil so he gave me some El Salvador Montelon to try. I'll have to wait until next week. However, I did have some left over Kid O during the transition and was happy with that. A friend who's opinion I trust tried some and was quite pleased also.

I've been told that the Fuji controller has a fast response but I wonder if anyone know's how much difference there is between it and the Fluke or even Omega?

Ed

[jesawdy]

I've been told that the Fuji controller has a fast response but I wonder if anyone know's how much difference there is between it and the Fluke or even Omega?

I'm glad you asked (and I hope Dan knows on the Fluke), because I've looked before and can't find the sample specs on the Fluke.

I have an Extech Digital Thermometer (not a datalogger), and its specs say 2.5 samples per second. The Omega HH506RA is also 2.5 times per second. I'm pretty certain that it cannot log at that interval though, more like one logged sample per second.

I have a Eurotherm PID that I have used as a temp display and they do not publish a sample rate that I can find. This Fuji PXR spec sheet seems to indicate an input sample rate of 0.5 sec. Does the display update that quickly?, I don't know.

[dawgcatching]

I have to ask the obvious. Would the temp stability increase if insulation were added around the boiler?

Here, I hope are 2 videos showing my temp. decreases during a pull. I'm using 6 day old West Coast Roasting Toro DeCaf. My heat cycle is from 0.91 to 1.1. These shots were made with the temperature read-out in mind so forgive me for the early blonding you will see.

Assuming these went through google's verification you should see some clips.The first clip is of a shot pulled from a long idle of more than an hour. This is without any pre-flush. Basically, I walked up, ground my beans letting the PF stay warm in the group as long as possible. Built my PF, then flush and go.






The second clip is of the following shot after rinse, wiggle, grind and build.




dawgcatching, how do these temperature decreases compare to yours?

Ed

P.S. On a humorous note, you can see a reflection of the TV news that's on a big screen 38 ft away in the chrome of the drip tray.

My machine temperature profile is different than yours, according to what the Fluke said. I was getting an almost linear temperature readout (201 to 202, depending on my flush) for 15 seconds, then a steep decline (between 15-25 seconds, I am seeing a 3 to 3.5 degree F drop). You are getting loss of stability from the end of pre-infusion throughout the whole shot, but slowly. Then again, I am measuring the temperature in the puck, whereas you are measureing the brewing water input temperature (or is that a PID readout)? The puck will have a higher heat capacity and hold temperature constant, even if the brewing input temperature is slightly decreasing over time. Our readouts aren't likely showing us the same data. Has anyone worked out a heat exchange analysis for an HX grouphead, portafilter filled with coffee, being heated by water?

[erics]

I've been told that the Fuji controller has a fast response but I wonder if anyone know's how much difference there is between it and the Fluke or even Omega?

The read-rate or display update for the various meters is:

Fluke 51,52,53,54 Series II is 4/second
Omega HH506RA and Extech 421509 is 2.5/second
Fuji PXR3 is 2/ second
Typical 1/32 DIN Chinese PID is 4/second

All of these are certainly fast enough for what we are trying to observe and/or record. The source for the above is published information with the exception of Fluke - that was taken from notes I had made a while back as a result of conversation with their engineering gurus.

On another note, the Bricoletta comes in a few different models and Ed has the one with the rotary pump. Therefore, (and I'm just using this as an example) a 15 second flush for Ed is a SUBSTANTIALLY greater water volume than a 15 second flush for machines equipped with a vibration pump. So, this nice machine may require a little more time in the lab than normal.

Skol,

Eric S.

[beta14ok]

So, this nice machine may require a little more time in the lab than normal.

My Isomac Mondiale is prolly the most like the Bricoletta of any machine out there. (ie E61, rotary, large boiler, HX, plumb-in)

I think that there are some "trends" with HX machines, but as "dawgcatching" demonstrates, there are some not too subtle differences.

My surf profiles are VERY similar. Everything happens fast on Mondiale (including between shot recovery and the need to re-cool the head). Sorry, I don't have video. The temperature hump during the head cool-down is very familiar to me. I usually "pull-out" of the cool-down surf at around 208-205°F. once I lock-in the pf and start the shot, I do see an initial drop of a couple of degrees, but then I see a modest recovery. I just pulled a shot which stabilized at ~201°F for the majority of the 25 sec pull.....tasted just ok

Does this happen every time? ....no
Do I think that temp stability on these HX machines is controllable or easily definable? ....no
Do I have any clue what temp the puck is actually seeing? ....no
Has the temp probe helped my reproducibility? .....no
Have my efforts improved over the basic techniques covered in Dan's "How I Stopped Worrying" article?? ...no
Do I need more "time in the lab" with lb after lb of espresso? ....you betcha
Do I wish that I was nailing a "god-shot" everytime? .....yes
Will my next machine be an HX machine? ....prolly not
Is that gonna happen soon? ....definitely not!
Do I make an enjoyable after supper 6 oz latte? ...yep
Am I content "to stop worrying"? ....well.........mostly yes, and I'm not selling Mondiale
Am I looking forward to Terry's proposal on how to tune the Bric's HX? ....yes!