CRCasey wrote:If the copper tubing in the head is causing a particle problem, could some type of sacrificial material be added to the tubing to reduce the problem? I am thinking something like we see on boiler elements.

-Cecil

My working hypothesis at this point is that working on or manipulating the copper part #TL30 is responsible for the verdigris corrosion becoming a problem with the gicleur, if in fact the verdigris is a problem.

I have no basis for this conclusion, however I should have more evidence as to whether it might be true when I disassemble my machine's group head for the 2nd time later this week, when I get the new part(s) from LM. I will then carefully look at the gicleur to try to determine what is blocking its channel (metal swarf, verdigris, whatever) and will look at the copper tubing to see how firmly attached the corrosion is.

It seems logical to me that under normal circumstances, if one does not experience a gicleur clog, that the verdigris forms slowly over the surface of the copper and in effect seals itself on the copper, preventing further corrosion from occuring beyond a certain point. It is when the gicleur needs to be cleaned out, the first time, perhaps due to metal swarf or other debris left over from manufacturing, that the surface of the corroding copper is disrupted (in my case by trying to clean it off) that then opens up the possibility of reclogging the gicleur with dislodged verdigris.

I don't know if LM is going to send me just a new, modified gicleur, or the rest of the TL30 in addition. The tech who will be doing this work for me hadn't decided when I spoke with him what he was going to send to me. If my old TL30 is going back into the group head, I'm planning to soak it in CLR to try to remove all the corrosion beforehand, which will then be able to reform at a "normal" pace, hopefully avoiding further dislodgements.

We shall see.

ken

This leads to the question of why in a otherwise stainless head do you drop a pipe in there like that in the first place? Why not just put a stainless tube for TL30 in the first place.

If you noticed the outside forming the green verdigris do you see anything of the same type forming on the inside due to the water?

-Cecil

CRCasey wrote:This leads to the question of why in a otherwise stainless head do you drop a pipe in there like that in the first place? Why not just put a stainless tube for TL30 in the first place.

If you noticed the outside forming the green verdigris do you see anything of the same type forming on the inside due to the water?

-Cecil

I didn't notice it but I didn't look for it, either.

As to the question regarding the use of copper in the TL30 part, I'd have to assume that this was used because copper is easier to work with than is stainless steel. I'm no chemist, but in general I think you would want to avoid using a part made of one sort of metal, within a water containing boiler made of another.

ken

What do people think of the solution Andy linked to, which moves the gicleur to the dispersion screen? This would make the problem manageable at least.

I suspect Ken may be on to something. I remember when I was going through my repairs when I first put the group head back together I stupidly forgot that the bottom ring which holds the portafilter has to be installed with a particular orientation and I installed the ring such that the portafilter handle wouldn't face the front of the machine. I took everything apart and with the boiler cool but not drained you can see the boiler water at rest in the group head. Not exactly the crystal clear water that is going into the machine that's for sure.

Cheers- Jeremy

What about creating a "cap" of sorts out of a screen with holes .6 or smaller to slip over the gicleur?

Ken Fox wrote:I'm no chemist, but in general I think you would want to avoid using a part made of one sort of metal, within a water containing boiler made of another.

+1

kzdad wrote:What do people think of the solution Andy linked to, which moves the gicleur to the dispersion screen? This would make the problem manageable at least.

I think this is a tricky idea. Putting the gicleur between the group and the solenoid valve means that pressure relief will be flow-restricted. I think that's a bad idea that will expose the gicleur to coffee dirt, making it probably even more prone to giving trouble, and also will likely make cleaning backflushes less effective at cleaning the feed tube/3-way solenoid valve chamber.

I believe I posted previously that when I opened up my GS3's cap to remove the gicleur (and later replace with a much larger one), I found verdigris similar to Ken's:



I did my best to remove this, and have been using my GS3 since then. Today I drained and flushed my boilers, and took the opportunity to have a peek under my group cap. This is what I saw:


Looking more like it should... (flickr)

I was very encouraged by that. However, looking at the group cap's underside was a decidedly different experience. The chrome on the part is flaking off, as seems to always happen on machines where manufacturers choose to chrome water-facing surfaces. I think this is kind of ridiculous. In any case, I found some verdigris (or something like it) on the cap, apparently formed on the fresh brass that was exposed as the chrome flaked, i.e. not on the previously exposed, oxidized faces of the brass. I brushed this buildup off easily with just a rag, and it looked like this:


Silly flaking of chrome on the brass group cap; note that the light-reddish areas interspersed with the chrome on the upper portion were where the verdigris had formed (flickr).


Detail of the buildup near the gasket groove (flickr).

I guess it's good it's coming off now. Why they didn't turn these out of stainless kind of baffles me. These things would be plenty easy to make on a CNC machine. I am sure it would cost more, but if it were done, I wouldn't have found chrome flakes on the stopper for the boiler drain. I could easily see this being one more thing to come off and clog a gicleur; unlike SS cuttings, chrome flakes are super-thin, and they also periodically separate and seem to travel downward. This kind of thing could catch it in a current that could suck it into the gicleur... maybe? Just a random thought.

In any case, on my machine I decided to get as much chrome as I could off, so I scoured it for a long time with a metal bristle brush and some sand paper. It's back now, and I'll be sure and look in on it soon. Here's what it looked like afterwards:


Clean-ish after a heavy brushing with the machine's included bristle brush (flickr).

Oddly, I couldn't get the last bits of chrome off. Maybe next time. For now, I am crossing my fingers that the parts will just oxidize and that this will prevent the Galvanic corrosion (or whatever it may be) causing this verdigris.

[Edit] Here is an article on galvanic corrosion in SS. It sounds like it should be extremely minimal at best with copper, but that it could happen, in which case the SS would be the anode and copper/brass the cathode. That would fit with buildup on the copper (flow of buildup is anode to cathode), but still seems hard to believe. Maybe it's just something in the water?

Nicholas,

Great post.

Didn't you find the verdigris buildup during the period when you were using cation-softened water? Could it be that the copper tube is clean now because you switched to the Claris system?

I think Ken postulated that the alkalinity in his cation-softened water may be responsible for the verdigris, though he said his level wasn't very high. Do you have alkalinity numbers for the cation-softened water you were using? My alkalinity level is around 150 ppm.

Is it your theory that the verdigris on the underside of the cap has caused the chrome to flake, or could it it have formed there after the chrome flaked off on it's own?

FWIW, I wasn't using cationed-softend water when my gicleur clogged. I was using zero-water spiked with tap to produce 30-70 ppm (I was experimenting with levels.) My machine clogged after less than three weeks of use, and during that time I was using water with hardness and alakalinity around 30-50 ppm. If high alkalinity causes verdigris, I doubt it was forming in my first machine. My bet is that some of the chrome came off the underside of the cap on its own.

Here's another clue: As you may recall, at one point I found small metal flakes in the dispersion screw that were partially blocking the holes. When I tried to dig them out, they seemed fused to the inside of the screw, and I had to use a hand-held drill bit to scrape them out. It seemed odd to me that the metal could fuse at hot water temperatures, but maybe thin chrome flakes can adhere in some way to the screw material.

shadowfax wrote:I think this is a tricky idea. Putting the gicleur between the group and the solenoid valve means that pressure relief will be flow-restricted.

Yes that was what I wanted, to achieve a lever group gentle decline. Pulled some awesome shots today which made me switch back to espresso after a 6 month break

shadowfax wrote:I think that's a bad idea that will expose the gicleur to coffee dirt, making it probably even more prone to giving trouble, and also will likely make cleaning backflushes less effective at cleaning the feed tube/3-way solenoid valve chamber.

Yes which is why I reckon we don't see it. It is asking for trouble, 0.6mm will clog with almost anything. So far so good on my machine though and you can backflush without the screw in place.

Your group looks nice and clean and you did a great job on removing the plating. Had a poke around the office and have found a few different group covers. Here's a brief summary

From 70's up until late 90's we had all brass caps without any plating.
From late 90's onwards standard AV and EE groups used the plated caps.
The new style Piero caps for AV and EE are stainless steel.
New paddle groups are brass without plated.

The tubes AFAIK have always been copper with brass fittings. Maybe they kept with copper and brass because it is a compression fitting and easier to get a good seal that way.

A catalog of precision orifices with M4 x 0.70 threads is here: http://www.okcc.com/PDF/Metric%20MD-1.pdf . Gives you a chance to use that 2 mm hex key wrench after you tap the dispersion screw as per Paul Pratt's idea here : Coffee side flow restriction .

Slowly relieving the pressure just seems like a brilliant idea.

Peppersass wrote:Didn't you find the verdigris buildup during the period when you were using cation-softened water? Could it be that the copper tube is clean now because you switched to the Claris system?

That's not true, and the whole truth is way more complicated. I'd used Cirqua AB Formulator packets for a few weeks, 'regular' salt-based cation softening, but mainly Everpure Claris. Since then I've only used Claris. So it's possible that Claris is non-offending, but I actually can't rule out the possibility that the verdigris formation was from, say, leaving the machine with a drained boiler exposed to damp air for over a year. It occured to me that maybe what I saw on the top of the cap might also be from some tiny pocket of air that didn't bleed out when I bled the group last time.

Peppersass wrote:I think Ken postulated that the alkalinity in his cation-softened water may be responsible for the verdigris, though he said his level wasn't very high. Do you have alkalinity numbers for the cation-softened water you were using? My alkalinity level is around 150 ppm.

Is it your theory that the verdigris on the underside of the cap has caused the chrome to flake, or could it it have formed there after the chrome flaked off on it's own?

You can find my water test results published here. The alkalinity I think was even lower than yours (100 ppm) for the salt-based ion exchange softener. My Claris numbers are very different, I believe at setting 3 (51 ppm hardness, 35 ppm alkalinity).

I really have no idea what makes chrome flake off brass when it's exposed to water, whether it's the scale/verdigris that separates the two, or some other process (and then the scale/verdigris forms on the newly exposed brass).

Peppersass wrote:FWIW, I wasn't using cationed-softend water when my gicleur clogged. I was using zero-water spiked with tap to produce 30-70 ppm (I was experimenting with levels.) My machine clogged after less than three weeks of use, and during that time I was using water with hardness and alakalinity around 30-50 ppm. If high alkalinity causes verdigris, I doubt it was forming in my first machine. My bet is that some of the chrome came off the underside of the cap on its own.

Here's another clue: As you may recall, at one point I found small metal flakes in the dispersion screw that were partially blocking the holes. When I tried to dig them out, they seemed fused to the inside of the screw, and I had to use a hand-held drill bit to scrape them out. It seemed odd to me that the metal could fuse at hot water temperatures, but maybe thin chrome flakes can adhere in some way to the screw material.

Well, it's too bad you didn't have a look in your boiler, or we'd know a bit more. I imagine you're right about the flaking chrome (doing it on its own), but who knows. And if verdigris is mainly formed by air exposure from having an empty boiler left in a warehouse (or a tiny unbled air pocket under the cap), then you're just as likely a 'victim' (whatever that actually means) of it as Ken and me.

Paul, thanks for the historical data! It's interesting how many things they've played around with. I guess cosmetics wins on an exposed group, even though the cap is mostly hidden by the plastic group cover. Too bad we didn't luck out with SS like the Piero caps...

And that's sure the truth about the feed tube. I'd honestly hate to have to re-seal a SS compression fitting, so I can't complain that much. On the other hand, they could use a viton o-ring or something like that to seal the group side of the feed tube; they sure do on the back side (not sure if it's viton, but it is an o-ring).

Rats, Paul! You're really tempting me. I am getting quickly attached to the idea of having the gicleur come out with the dispersion screw (great for backflushing as you say, and I already backflush without it). Time to order an M4 tap... I even have spare dispersion screws.