Single boiler machine, right?

Yep, single boiler -- no steam.

Coffee tastes OK... I needed more time with it.
I say that because shortly after getting it working (about 18 months ago), I gave it to my dad, who lives 900 miles away. So I don't see it often.

However, I recently was decommisioned since the OPV valve I made is getting a little flakey. I keep urging him to ship it to me, but in his defense, it does weigh in the ball park of 50 lbs. He doesn't need to be messing with shipping that.

So there it sits, waiting for the next time one of us drives from FL to NC, or vice versa... I'll have it back by Dec (planned drive) at the latest. Then it will be out with the vibe and I'll put the procon I have laying around to use. Externally.

Also plan on changing the element. I originally used a water heater element and despite the coating which I was careful to preserve, there are traces of rust. May also throw in a PID I have sitting around. The control system worked well, but the sensor was in a well, and had to sense thru a thin wall of stainless--not ideal. If I go PID, I can use a TC that's in direct contact.

After all that, the plan is to submit it for review by the local HBs (this includes the thee HB himself!).


ciao

lino

lino wrote:Coffee tastes OK... I needed more time with it.

Certainly, the idea of creating an LM-like saturated group with an E-61 is terrific (assuming you want a flat temp profile, of course!) An E-61 has the big advantage of not making the water exit the group on its way to the portafilter.

lino wrote:Also plan on changing the element. I originally used a water heater element and despite the coating which I was careful to preserve, there are traces of rust. May also throw in a PID I have sitting around. The control system worked well, but the sensor was in a well, and had to sense thru a thin wall of stainless--not ideal. If I go PID, I can use a TC that's in direct contact.

If you're concerned about responsiveness, I would recommend a thermocouple in a 1/16" stainless sheath, fitted into the boiler with a compression fitting. That's plenty responsive enough for a boiler, without getting into the hassles of direct contact. FWIW.

Hey Andy,
I should clarify. That TC arrangement is what I meant. I have no intention of trying to get a bare TC bead in there. I'm nuts and like doing things the hard way, but I'm not that crazy. I was [sloppily] using the colloquial definition of a TC...

I'd plan on doing just as you said, as I also did on the La Peppina (in two spots) which I've posted around here somewhere... Here:



ciao

lino

Even more cutaway pictures (click to see the original 1.4M image). Thanks Barry!


Portafilter-side facing up, diffusion block in place


Portafilter-side facing up, diffusion block in place, angled above


Grouphead "gooseneck" shown attached to boiler

Hi everyone -

Throughout this thread I've seen several references to modified LM groups, LM hybrid group, and the LM GB5 group. I'd be super excited if anyone could compare all these in a post! I got a chance to see the hybrid internals at ESI quite some time ago, and thought that boiler/group design (pre-heat, solenoid shuffle) was making its way into the GB5.

malachi wrote:I worked on a machine for a while that had the Blackwell/Schomer banjo tube mods done and actually found it to be far more stable. Of course, you want to talk about service PITA?!?

I've heard conflicting results about lengthening the banjo tube. I'd love to learn more about it, or have the opportunity to work on a machine with this mod (anyone in/near Seattle using it??)

I'm about to do a banjo lengthening or "Pratt" solenoid implant into the grouphead (my personal vote for all time best mod). I don't suppose anyone has a picture of a head with the long tube? I understand it's roughly 21" long. The stock banjo tube is slightly less than 1/4" diameter and I suppose a welded compression fit to a true 1/4" tube would work, but I'm looking for better ways. Has anyone used a thinner diameter tube to get more coils and therefore length? Too thin would affect pressure but may actually enhance preinfusion. Thoughts would be appreciated. Especially if tips can be given on the how-tos of affixing the solenoid inside the body in lieu of a lengthened tube. thanks.

On a side note, to do variable length preinfusion I found an old time-delay relay that ranges 0.1 to 10 seconds. I can have the brew switch activate the solenoid and time-delay relay, and have mains pressure preinfuse while the time-delay relay activates the pump a few seconds later. A dial on the relay varies the time delay.

Keith

Hi Keith, sorry for the slow response to this and your email!

I looked at 2 ways to do this last year. The advantage I have is that I have dozens of old groups heads that I can cut up with worrying too much about the consequences.

There are basically 2 solenoid bodies out there for espresso machines, the square head which has the 2 o-rings and mates against a flat block



The other is the type that Marzocco use for the group and what is usually used for 2 way operation like autofill, hot water taps etc...



The first method I tried was with the square head. I blocked off the regular solenoid valve inlet and outlet pipes on both sides of the group.



Then the brass group was modified on the right hand side to provide the space for the solenoid valve. The rough finish on the brass was filed, sanded and then polished to create a smooth surface so that the valve head could fit true and the o-rings could seal. That was the hard part because the group is a really awkward shape to work with. Now, the solenoid requires an inlet from the boiler and an outlet that will go onto the coffee via the banjo tube. I then drilled 2 holes into the side of the group that would line up exactly with the ports on the solenoid.

I could have just let the inlet hole be, but I decided to draw the water from the boiler (like on the old Marzocco manual group machines), so a 10 cm copper pipe was made so that one end protruded down into the boiler and the other went up to the inlet hole of the valve. It was then brazed into place and then I dressed up the brazing with some drill bits so the surface would seal with an o-ring. The other end of the pipe (into the boiler) was fitted with a 0.7m gicleur.

The following pic shows the basic principle using the old manual group. There is a tube that runs down into the boiler where the water is drawn from.



The banjo tube was then cut so that it was placed into the other hole, brazed, dressed and finished. The square head of the valve was then held onto the side of the group by 4 socket head screws (drill 4 holes and tap). The drilling of the holes was hard because the brass isn't all that thick and it was easy to drill completely through.

Amazingly the valve required only minor tweaking to get it to seal correctly and was tested using cold water for a couple of days before I heated it up.


The other way of doing it involved cutting into the group and brazing in the rectangular head. Sadly I can't for the life of me find the pics I had of this. I remember it was a lot more difficult than the 1st method and it was a struggle to get the valve head in and have enough clearance for the steel stem and the coil to fit properly.

I still have the first version, the second version is somewhere but these are just groups and had been taken off the machines. Maybe I'll put them back soon to do some testing with the Scace device. It was a long time ago and my lasting impression I had of both groups was that it messed with the infusion times which is why I added a pre-infusion switch to it.

Paul

Thanks Paul.

I ended up taking the "easy" way out and lengthening the banjo tube. That was NOT easy! I finally got it after scrapping a few copper lengths and cutting the banjo tube to within an inch of its life, and am just about finished implementing a plan I've had for a while which is to provide constant circulation to the grouphead via a circulation pump or impeller, and to preheat the inlet boiler water to try and get a very stable and predictable temperature. I used a potable hot water Taco circulator pump and tapped holes in the group caps of each of the two groups on one of the two boilers (4 head machine). So the pump sucks water out of one head just above the banjo bolt and returns it into the other head. I preheated the boiler inlet water by simply disconnecting head number 3 outlet from its flowmeter and connecting it to the inlet of the first boiler. So the second boiler provides hot water for the first boiler inlet. And finally, the time-delay relay as mentioned above for preinfusion. (And PID controller and thermocouple array as standard procedure of course.) So far I've pressure tested at 150F and the circulator works fine. I have yet to do a full blown test. So little time...

Keith

Well done Keith. The banjo tube mod should be simple but I imagine squeezing the serpentine path in there isn't easy.

Maybe all mods are now redundant because there is a new LM group out there. There's already the awesome hybrid group from Bill at LMI with the valve at the back of the group, but now it looks like they are producing another version of this with the solenoid valve and flowmeter embedded up near the group cap. I saw it a few years back and I think they waited a few years to get it ready before rolling it out. Definitely some (if not all) the USBC machines used this as Gimme in Brooklyn has one with this group config. So on both these groups the only time the water really leaves the saturated group is when it finally hits the coffee.

Paul