New Preinfusion for La Marzocco Linea Mini R

Need help with equipment usage or want to share your latest discovery?
User avatar
Jake_G
Team HB
Posts: 4319
Joined: 6 years ago

#1: Post by Jake_G »

The new Linea Mini R officially released on Valentine's Day touts "An all-new two valve pre-infusion system offers soft saturation and a self-cleaning flow restrictor." Here's a general discussion on the new machine:
La Marzocco Linea Mini R

So, what's new about preinfusion on the Linea Mini R, and why are two valves needed?

Let's start with how the OG Mini (and the GS3, and the Micra) handle reservoir-fed preinfusion, or pre-brewing. The process is broken down into 3 steps, all adjustable via the app (or the button panel on the GS3):
  1. Pre-Brewing On: The pump starts and the 3 way valve is energized, filling the headspace above the puck. This phase lasts the duration of "Pre-Brewing On" as set in the app/panel.
  2. Pre-Brewing Off: The pump continues to run and the 3-way valve is shut off, venting the puck to the drain box. The puck stays vented for the duration of the "Pre-Brewing Off" timer.
  3. The pump continues to continue running and the 3-way valve is again energized, reapplying pressure to the puck until the shot is complete.
The primary problem here is that phase 2 above is inverted. What ought to have been done all along is to turn the pump off and leave the 3-way valve open, allowing the pump pressure to dissipate through the puck. So why did LM put a second solenoid in there?

Well, let's take a look at the old Integrated Brewing Group (IBG) and understand what the 3-way valve is doing when it switches states.


So, here we have a view of the back of the IBG with the Gicleur (item 20) threaded into fitting 11 that feeds the 3-way valve. The top of the IBG is through fitting 9 by way of the cross fitting. The cross fitting receives cool water from the pump and hot water from the HX in the steam boiler. It also feeds the brew pressure gauge. Just to review:
  • Water from pump feeds the HX, the boiler fill valve and the cross fitting.
  • The cross fitting feeds the top of the IGB
  • The top of the IGB feeds the 3-way valve by way of the gicleur.
The 3-way valve then switches the group at the bottom of the IGB between being vented to the drain box (solenoid "off" or "closed") and being connected to the top of the IBG (solenoid "on" or "energized" or "open").

As I mentioned above, inverting phase 2 of the pre-brew process would make the reservoir-fed pre-brewing process 100% better than it is now. So why did LM feel the need to put a second valve in here? My hypothesis is that is has to do with the HX in the steam boiler, and an unnatural concern that LM has with allowing the HX to boil dry. If you open the 3-way valve without turning the pump on, the water inside the HX within the steam boiler will begin to boil if the pressure in the HX drops below the pressure in the steam boiler. This is because the steam boiler is, by definition held at the boiling temperature that corresponds to its steam pressure. So, if you vent the pressure in the HX to the group, the HX water will begin to boil and push water into the IBG.

So what?

Well, that's kind of my stance. Steam forces water from the HX into the top of the IBG, and that pushes water through the solenoid and into the group. No big deal. There is a few hundred ml of water between the HX and the IBG, so it really isn't an issue. As soon as the pump turns on and pressurized the HX, all the steam condenses back into water, so you're not running into any issues of getting air in the system.

But, I think that LM - in an abundance of misplaced caution - decided that this is unacceptable. So they over-engineered the IBG to account for, and prevent this situation from ever occurring. I applaud them.

Here is the new IBG:


It is fed by the HX in largely the same manner as before, but now, we have a close-couple solenoid block that has not one, but two solenoids. Why's that?

Well, the IBG is still largely the same. The top is filled with water from the cross fitting, but now, rather than feeding the 3-way valve by way of a gicleur threaded into a fitting, a 2-way solenoid (open or closed) is threaded into the top of the solenoid block (item 7 on the diagram above). When this valve is energized, water can flow from the top of the IBG to the 3-way valve below (item 6). The 3-way valve now seats against the gicleur (item 8), which is why the gicleur is now "self-cleaning". The 3-way valve works the same as before. But now we have a two-way "block valve" which can completely isolate the group from the top of the IGB, and the HX that is feeding it. With this, the new Preinfusion settings on the reservoir look something like this:
  1. Preinfusion On: The pump starts running, the 3-way valve is energized, and the block valve opens. This continues for the duration of the "Preinfusion On" timer.
  2. Preinfusion Off: The pump may or may not continue running (doesn't really matter), the 3-way valve stays energized, and the block valve closes. This continues for the duration of the "Preinfusion Off" timer.
  3. The pump continues running (or turns back on, knowing LM, they probable keep it running), the 3-way valve continues to continue being energized and block valve reopens, applying pump pressure to the puck for the remainder of the shot.
Why do I applaud them for this overengineering? Well, for one, I'm an engineer. So... there's that. But far more importantly, look at what this allows you if you are plumbed...
  1. Preinfusion On: The 3-way valve is energized, and the block valve opens, allowing line pressure to saturate the puck. This continues for the duration of the "Preinfusion On" timer.
  2. Preinfusion Off: The 3-way valve stays energized, and the block valve closes, allowing a bloom phase. This continues for the duration of the "Preinfusion Off" timer.
  3. The pump turns on, the 3-way valve continues to continue being energized and block valve reopens, applying pump pressure to the puck for the remainder of the shot.
I believe this is what LM calls:
"Hybrid mode for plumbed machines
Flexible brewing with traditional and new pre-infusion phases."

Let's discuss this. Has anyone verified the operation? Am I wrong?

Cheers!

- Jake
LMWDP #704
★ Helpful

PaulTheRoaster
Posts: 261
Joined: 18 years ago

#2: Post by PaulTheRoaster »

There's also that the steam boiler might be off or it might be refilled (but it seems like the Micra at least does not refill during brewing).

Maybe?:
You have a really long blooming time. You steam a quart of milk (that's why you got a Mini), dropping the steam boiler temperature enough that the HX starts pulling brewing water into the brew boiler.

(As you say, for plumbed machines, you do really need two valves for a bloom phase.)

sboplataz
Supporter ♡
Posts: 61
Joined: 10 months ago

#3: Post by sboplataz »

Thanks for explaining some of the inner workings; I found this fascinating!

hao
Supporter
Posts: 17
Joined: 3 years ago

#4: Post by hao »

Thanks for the detailed description, I'm also fascinated by stuff like this! I did have a couple questions - can you elaborate why the gicleur positioning makes it "self-cleaning"? and related to this, where actually does the gicleur sit? the three options to me seem to be a) between brew boiler and 2-way solenoid; b) between the two solenoids; and c) between the 3-way solenoid and the brew head (lower portion of IBG). I can imagine "self-cleaning" here to mean that water can flow both ways through the gicleur, which would make the most sense with option c.

also, having the new solenoid valve be 2-way makes complete sense to me, but comparing the diagram to the old linea mini parts diagram, it looks like both valves are 3-way, which doesn't make sense to me, so perhaps it's just the new diagram that's incorrect right now?

thanks!

User avatar
Jake_G (original poster)
Team HB
Posts: 4319
Joined: 6 years ago

#5: Post by Jake_G (original poster) »

PaulTheRoaster wrote:There's also that the steam boiler might be off or it might be refilled (but it seems like the Micra at least does not refill during brewing).

Maybe?:
You have a really long blooming time. You steam a quart of milk (that's why you got a Mini), dropping the steam boiler temperature enough that the HX starts pulling brewing water into the brew boiler.

(As you say, for plumbed machines, you do really need two valves for a bloom phase.)
The HX is hydraulically independent from the steam boiler, so there is no way the HX would "pull" under any circumstances. Thermal expansion is of course a real thing, but you're talking about a couple grams of movement between room temp and boiler temp for the volume of water that is in the HX. So if you take an OG Mini and build a few bar of pressure in the group and then turn the pump off, the pressure will decay as the water saturates the puck and you'll have residual pressure in the group equal to whatever the steam boiler pressure is until the pump kicks back on. If you steam a bunch of milk during this phase, the heating element might kick on, but LM historically limits this phase to no more than 10 pr 15s, so I don't see room for too muck dynamic variation within the bloom phase. And not that you asked, but there is a check valve after the pump. So, there is no way the pressure will just bleed back to the reservoir.
sboplataz wrote:Thanks for explaining some of the inner workings; I found this fascinating!
You're most welcome!
hao wrote:Thanks for the detailed description, I'm also fascinated by stuff like this! I did have a couple questions - can you elaborate why the gicleur positioning makes it "self-cleaning"? and related to this, where actually does the gicleur sit? the three options to me seem to be a) between brew boiler and 2-way solenoid; b) between the two solenoids; and c) between the 3-way solenoid and the brew head (lower portion of IBG). I can imagine "self-cleaning" here to mean that water can flow both ways through the gicleur, which would make the most sense with option c.

also, having the new solenoid valve be 2-way makes complete sense to me, but comparing the diagram to the old linea mini parts diagram, it looks like both valves are 3-way, which doesn't make sense to me, so perhaps it's just the new diagram that's incorrect right now?

thanks!
The gicleur is self-cleaning because the plunger in the solenoid valve seats against it. As such, any potential buildup is disturbed every time the valve cycles.

The gicleur sits between the 2-way valve and the 3-way valve. This way, dirty dregs after the shot never back flow through it. While it seems a good idea at first blush to have 2-way flow through the gicleur to clean it, the dregs from the group are orders of magnitude more dirty than the brew water coming from the top of the IBG.

There is no fundamental difference between a 2-way valve and a 3-way valve, aside from the armature that the plunger and spring reside inside:
  • On a 2-way valve, the armature is closed on the coil side and the plunger blocks flow between the port that the plunger seals against and a second port that is always exposed to the spring and plunger.
  • On a 3-way valve, the armature is open on the coil side, and there is a seat inside that the plunger seals against when the coil is energized. This port goes to the drain tray and the port that the spring seals the plunger against goes to the boiler and the port that is ipen to the spring and plunger inside the armature goes to the group.
In the case of the Micra and the Mini R (and also the GS3) the seat that goes to the boiler (or the 2-way valve on the Mini R) is actually the gicleur, as mentioned above. Also, LM occasionally uses a 3-way valve with a cap on the coil side in place of a 2-way valve. I'm sure why, but they consistently do it on the GS3 tea water solenoid, whilst the steam boiler fill solenoid is a proper 2-way. Folks have theorized it has to do with better bi-directional sealing, but I'm not fully convinced there is any real benefit there.

Cheers!

- Jake
LMWDP #704

hao
Supporter
Posts: 17
Joined: 3 years ago

#6: Post by hao »

Got it, that makes a ton of sense - thanks so much for the explanation!

PaulTheRoaster
Posts: 261
Joined: 18 years ago

#7: Post by PaulTheRoaster »

Jake_G wrote:The HX is hydraulically independent from the steam boiler, so there is no way the HX would "pull" under any circumstances. Thermal expansion is of course a real thing, but you're talking about a couple grams of movement between room temp and boiler temp for the volume of water that is in the HX. So if you take an OG Mini and build a few bar of pressure in the group and then turn the pump off, the pressure will decay as the water saturates the puck and you'll have residual pressure in the group equal to whatever the steam boiler pressure is until the pump kicks back on. If you steam a bunch of milk during this phase, the heating element might kick on, but LM historically limits this phase to no more than 10 pr 15s, so I don't see room for too muck dynamic variation within the bloom phase. And not that you asked, but there is a check valve after the pump. So, there is no way the pressure will just bleed back to the reservoir.
What I mean is if HX temperature is very close to the steam boiler temp, and you put something gross in a blind portafilter, and ran a long blooming phase, and used up all the steam (let's say you've also tripped the autofill, so the steam element is off until it can refill after the shot), the HX pressure could be < brew pressure.

The check valve in the pump is enough, but the plumb-in kit comes with an extra check valve for the same reason that the check valve in my domestic hot water recirculating pump is enough, but plumbing codes require another. (So that you can point to something instead of saying "trust me there's a check valve here"). So perhaps there is some regulation that requires it to be done this way. The allowable backflow is usually zero. (Thermal expansion is not going to get water from my hot water heater into the mains, but it doesn't matter.)

There aren't any great theories on this :roll:

BaristaBob
Posts: 1873
Joined: 6 years ago

#8: Post by BaristaBob »

Jake_G wrote:The new Linea Mini R officially released on Valentine's Day touts "An all-new two valve pre-infusion system offers soft saturation and a self-cleaning flow restrictor." Here's a general discussion on the new machine...
Jake,

As one who would run this machine off the reservoir, I was told yesterday by Chris at LM Home that the double solenoids allow the prebrew function to operate "properly" with the reservoir. Though you need to use the app to set the prebrew, it does appear to provide enough pump on/pump off flexibility to fully saturate the puck without a release of pressure before applying full pressure. I'm game for this versus the normal preinfusion I currently use. I can't experiment with prebrew on my machine (BDB) because of "puck jump" if I turn my pump off and then back on (can't hold group pressure). This is also why the extraction time was extended to 90 seconds. Chris said if someone wanted to bloom the puck for an extended time (example: pump on for 2 sec., pump off for 60 sec., then full on mode), you have the room.

Does this make sense to you?
Bob "hello darkness my old friend..I've come to drink you once again"

User avatar
Jake_G (original poster)
Team HB
Posts: 4319
Joined: 6 years ago

#9: Post by Jake_G (original poster) »

BaristaBob wrote:Chris said if someone wanted to bloom the puck for an extended time (example: pump on for 2 sec., pump off for 60 sec., then full on mode), you have the room.

Does this make sense to you?
Yes it does!

My only comment is that they really didn't need to add a valve to achieve this. All they need to do turn the pump off in the "pre-brewing off" to achieve 90% of the same exact results. What they've done is above and beyond this and allows for a blooming phase when plumbed in, which prior models could not do. I'm not dogging the LMLM R, but rather imploring LM to fix the pre-brewing on existing machines.

Cheers!

- Jake
LMWDP #704

BaristaBob
Posts: 1873
Joined: 6 years ago

#10: Post by BaristaBob replying to Jake_G »

I see your point. That should be an easy fix for the machines with LoT, not so easy for older machines.

I did not ask Chris if indeed the pump actually stops in the pump OFF phase. I'll pose this question to him tomorrow, unless someone here actually knows the answer. I got too excited about the group actually holding pressure during prebrew! :roll:

But I did ask him the elephant in the room question...that being, so couldn't you do this on the Gs3 AV? His answer has me excited but with measured enthusiasm. Chris said, and I paraphrase here, "the Gs3 AV hasn't really changed since its introduction back in 2008. Development is underway for a radical upgrade for this machine." He could not tell me when the next AV would be introduced to the marketplace but it's coming.
Bob "hello darkness my old friend..I've come to drink you once again"