HB wrote:...instructions provided by the manufacturer indicate D = 8.

Thanks Dan!

I got my Pro 700 at the end of December and I too noticed that running water from the group sounded a bit like flash-boiling. I don't have any precise means of verifying the temperature at the group at the moment, but I checked and all PID parameters are at their recommended values out of the box, except perhaps for the brew boiler temperature which is set at 200 instead of 199. I may lower it, and also try to find a way to measure the temperature of water running from the group, but I wonder if the sound I hear is misleading (and not at all flash boiling).

After a few months of playing with the Pro 700, I managed to confirm that with the default Fahrenheit settings, water is much too hot, as described by the OP almost a year ago.

I have part of the explanation however. Comparing the recommended settings in Fahrenheit and Celsius, everything seems to match: default for t1 is 93C, which translates to 199.4F, so either 199 or 200 fits; default for t2 is 124C, which translates to 255.2F.

However, E1 is amiss. The default is an offset of 14C, which should correspond to an offset of 25F, not 34F as recommended. The difference is a whole 5C.

Combine this with posts in which E1 should be set lower in Celsius anyway (12C, or even 10C, which would corresponds to 22 to 18F) and I now understand why I was struggling with over-extraction.

Furthermore... is parameter d really units-dependent? Looks like d=4 is preferred in F despite the official recommended value of d=8.

Neutro,

I am curious did you ever decide on a specific setting?

Yeah, I should perhaps have updated this thread --- discussion on temperature with the Pro 700 spans numerous threads now.

In my case, which must be stressed is different from others, the default Fahrenheit settings result in flash boil at the group so it's clear the temperature is too hot. With the default Celsius settings (E1 = 14C, T1 = 93C), flash boil disappears but there's quite a lot of vapor coming from the group head. I'm not sure if it's customary or not; but after all 93C is not that far from boiling point.

In one thread I posted measurements with a thermal infrared camera. While high-tech it's not necessarily precise because one has to make assumptions (e.g. regarding the emissivity of water coming out of the group). But I found out E1 should be at 12C with this method, although this was probably a lower bound.

Another user (spearfish25 I think?) suggested another method to adjust E1 which is brilliant. You begin by setting E1 to 0, and thus having T1 reflect the brew boiler temperature. You start with T1 a bit under the boiling point (e.g. 95C) and then let water through the group, before inching the temperature up (96C) and testing the waters again. At the first hint of flash boil within the group, you know that water is 100C at the group. At this point, T1 - 100C is the value you should use as E1.

The only flaw in this method is that the hotter T1 is, the more pronounced the temperature drop in the group should be, so it's not entirely linear. Since we want colder temperatures at the group than 100C (say, 93C), the temperature difference found with this method should be a higher bound. In my case, I found... E1 = 14C, which is the default Celsius setting.

It's also the setting I use for now, and it works well. It must be said that while I sure got over-extraction when I was using the default Fahrenheit settings, I now believe that most of my sink shots were due to a mix of not grinding fine enough and not preparing the basket correctly. That being said, if I found 12C with the thermal IR method, and 14C is an upper bound, perhaps I should use E1 = 13C as a compromise! In any case, E1=14C and T1=93C works well for me now.

The discrepancy between the recommended E1 in Celsius and Fahrenheit hasn't been resolved as far as I know. User erics measured the temperature at the group with a Scace and found the Fahrenheit settings to be sound. So it looks like it's somehow machine-dependent. Profitec did not reply to my inquiries on this topic. As for the P, I, D settings, I use the recommended ones. Most methods for adjusting the PID settings requires access to the output of the PID (i.e. the actual command sent to the process). I think the settings are okay even if they produce a small overshoot (1-2 C). The PID should be adjustable so as to not produce an overshoot, and still get the target temperatures as fast as with the overshoot. The temperature stabilizes fast enough for my needs anyway. I have no idea if the temperature is normalized internally or if the units affect the PID settings that should be used.

neutro wrote:Another user (spearfish25 I think?) suggested another method to adjust E1 which is brilliant. You begin by setting E1 to 0, and thus having T1 reflect the brew boiler temperature. You start with T1 a bit under the boiling point (e.g. 95C) and then let water through the group, before inching the temperature up (96C) and testing the waters again. At the first hint of flash boil within the group, you know that water is 100C at the group. At this point, T1 - 100C is the value you should use as E1.

The only flaw in this method is that the hotter T1 is, the more pronounced the temperature drop in the group should be, so it's not entirely linear. Since we want colder temperatures at the group than 100C (say, 93C), the temperature difference found with this method should be a higher bound. In my case, I found... E1 = 14C, which is the default Celsius setting.

You are correct, sir. That was me. I think my method also gets impacted by the flow rate through the group. If I set my PID to 201 and run water through the group without the portafilter in place, I get a tiny hint of flash boiling. It's just a slight rumble, gurgle as the water comes out. But without the portafiler and coffee in place, water is traveling at higher flow from the boiler and out the group. I suspect it's losing less heat during that journey than if it's under pressure and having a slower progression through the group. On the other hand, my offset value could be just plain wrong.

Neutro,

Thanks for the detailed response.

John

Sorry to bring up an old thread but did the HB community decide in any official PID settings for the Pro 700?

Is the 34 offset agreed upon? Also 199 and 255 I in F?

Is programming in Celsius the way to go?

I don't think there is really a consensus. Some people, including @erics, measured the output of the machine with a Scace thermofilter and vouch for the default Fahrenheit settings (255F boiler, 34F offset). The fact is, the offset is mathematically proven to be different in Celsius and Fahrenheit. Recently (can't find the thread right now, will update comment if I do), I've read a user getting an email from Profitec saying that the settings in Celsius (93C boiler, 14C offset, which translates to 25F) are correct.

In the end though, temperature is easy to modify, and now that I have played with my Pro 700 for more than a year, I see that grind settings and basket preps decidedly have much more impact on taste than temperature. I for one currently use the default Celsius offset (14C) but boiler set to 95C, so a tad hotter than default, and I get good results. I will continue looking for improvements but going to default Fahrenheit settings definitely produces flash boil at the group for me, so I'm pretty sure it's too hot.

So to make a long story short, no consensus yet, not sure if there will be, but still can do great shots with the machine!

A quick update.

I'm not sure how to say this, but I think I changed sides on the issue.

After a 3-week trip to Australia, I became more accustomed to espresso I guess, and returning home to my newly repaired machine, I felt the taste was off. I decided to re-test using an E1 of 19C / 34F.

I have no idea if the water hits the coffee at the right temperature or not -- and I do hear flash-boil when I let the water run from the group -- but it seems to me that the coffee just tastes better. This is with now one year experience working with a bottomless portafilter and VST basket, so my basket technique has probably improved a bit, which might explain my previous uncertain assessment.

I think the lesson here is just try both settings and chose the better tasting one; or buy a Scace if one wants to be 100% sure.