Here's a picture that illustrates how the Speedster routes water through the grouphead (cover removed for photo), and how it helps achieve excellent intrashot temperature stability:



When you pull a shot, water flows from the boiler (underneath) up through the aqueduct "A" into the grouphead "B". It then goes back through the gicleur "C", through a copper tube to the three-way valve "D", through another copper tube up the aqueduct and to the dispersion block below "E".

In all this back and forth movement the brew water equilibrates in temperature with all the water in the aqueduct and the group. Because of their exposed position, the aqueduct and group run about 4F lower in temp than the boiler.

Cleaning/changing the gicleur is pretty easy, you just have to remove the group cap and unscrew the gicleur with a 3mm hex wrench. Please don't ask -- duh -- how long it took for me to figure out where the gicleur was.

So how are the shots? Seems they have a little better mouthfeel and a little clearer flavor profile compared with my venerable Tricked-out Silvia. I haven't seen a huge difference yet, and it's tricky setting things up so that one is really comparing apples to apples (eg, making sure temperatures, pump pressures, preinfusion, etc are reasonably equal). Taste testing continues....

Excellent, thanks for the great post. Out of interest, does anyone know if this back and forth arrangement is how it works with other self bleeding saturated groups also (Synesso in particular) or is it uniquely Kees?

Cheers and Merry Xmas to all.

AndyS wrote:Here's a picture that illustrates how the Speedster routes water through the grouphead (cover removed for photo), and how it helps achieve excellent intrashot temperature stability:

When you pull a shot, water flows from the boiler (underneath) up through the aqueduct "A" into the grouphead "B". It then goes back through the gicleur "C", through a copper tube to the three-way valve "D", through another copper tube up the aqueduct and to the dispersion block below "E".

In all this back and forth movement the brew water pretty much equilibrates in temperature with all the water in the aqueduct and the group. Because of their exposed position, the aqueduct and group run about 4F lower in temp than the boiler.

This is really cool, Andy. It reminds me of another aqueduct I visited recently in France, le pont du gard:

ken

gyro wrote:does anyone know if this back and forth arrangement is how it works with other self bleeding saturated groups also (Synesso in particular) or is it uniquely Kees?

Didn't seem to be that way on the GS3 prototype that I tested (photo below). I believe the water pickup point was down near the 3-way valve. And if I remember what Bill Crossland told me, the 3-way had a very small orifice (~0.6mm), so a separate gicleur wasn't used.

When you say "self bleeding saturated groups," are you referring to the owner's bank account?

Ken Fox wrote:This is really cool, Andy. It reminds me of another aqueduct I visited recently in France, le pont du gard:<image>

Looks pretty, but can it make good coffee?

Neat setup -- a perfect picture book counter flow heat exchanger. This means the temperature throughout the shot will pretty much stick to the initial group head temperature no matter what the brew boiler is doing. The brew boiler temperature only matters to set the initial group head temperature at idle.

The whole thing can drift if you make shot after shot unless that aspect has been calibrated as well. But still, this seems a much better piece of engineering than the initial LM version of the group.

As impressive as that group is, the welding of the group head to the stem is pretty impressive too. That's a pretty confined area to get a tig in there.

The water path of the GS3 is indeed different, also the serial models have the same setup:



I am wondering about the temp. stability and the location of the water inlet. I would expect it to be lower, is there a significant offset between coffee boiler temp. and output? Any chance to apply the Scace?

Have you connected it to direct water?

Best,
- Malte

the_deal_maker wrote: is there a significant offset between coffee boiler temp. and output? Any chance to apply the Scace?

Have you connected it to direct water?

Hi Malte:

The Scace says that the brew water temp is about 4F lower than the boiler temp. But I adjusted the PID controller to compensate for that. So now the display reads very close to the brew water temperature.

Yes, the machine is connected to direct water (that's the only way it operates).

Thanks,

very nice picture - seems that you are very talented in Photoshop. I am still undecided towards the Speedster vs. the GS3 paddle. I love the design and outer appearance of the Speedster and Kees is a guarantor in terms of quality. On the other hand the GS3 paddle is a full automatic, fully integrated machine, that gives you manual control over the brew procedure. Haven't figured whether it is possible to control the pressure itself step-less through the paddle. It might be similar to the Speedster off/infuse/brew.

The Speedster is such a beauty... I should go for it... however the GS3 paddle is also very tempting.

Hi all, thanks for all the info. Malte, I take it you didn't plumb in your GS3 at all. As you know the Speedster requires a plumbed in line. Whilst it may be a small PITA to set it up initially, I would never go back to a pourover. I believe Kees includes a fitting to attach to a washing machine water outlet if that helps in your setup? My plumbing situation consists of a washing machine outlet feeding a 'spaghetti junction' which splits it to the washing machine, dishwasher and coffee machine. This was a real pain to sort out since I don't speak Chinese and have no idea about plumbing! However, I really do think it was time well spent. (My previous machine was ok for pourover or plumbed in).

I'm guessing as well that the GS3 paddle pre-infusion would be line pressure, which would require a direct plumb in as well, but I could well be wrong on that one. Thanks for the Nino info as well.

Cheers, Chris

And re the photos, I concur. Andy, you've taken and tweaked some pretty cool snaps...

You are right, Chris. I should spend some time to figure out how to realize direct water. I have to integrate some kind of stop valve that is easily accessible. I am living in a flat that is owned by Germany's largest insurance company, as a result I have to pay attention to avoid any damage by water.

And yes, the GS3 paddle needs to be plumbed as well. I have asked Kees concerning the direct water and he answered the following:

"No, it is progressive pre-infusion. This means, after wetting the coffee, it absorbs the pump pressure and then slowly releases this pressures. So, this system also works with drawing water from a container."

Hope it is okay with Kees to post his statement on that topic. Running from a water container/gallon would be the best option for my home setup.

Cheers,
- Malte

the_deal_maker wrote: I have asked Kees concerning the direct water and he answered the following:

"No, it is progressive pre-infusion. This means, after wetting the coffee, it absorbs the pump pressure and then slowly releases this pressure. So, this system also works with drawing water from a container."

In espresso, the word "preinfusion" is used in a lot of different ways. But they all basically mean wetting the coffee completely before applying full pump pressure.

In the Speedster this process is accomplished in three ways:
1. A gicleur (orifice) in the brew water path slows the initial buildup of water pressure in the group.
2. The spring-loaded progressive preinfusion cylinder releases to slow the initial buildup of pressure.
3. Using the shift lever, the operator can saturate the coffee with line pressure (3-4 bar) before turning on the pump (8-9 bar).

This results in some pretty freaking gorgeous extractions -- and they taste good, too!

Malte, I use an EVERPURE filter and some of these come with a shut off valve. They seem to be more common in the States but I am sure you'd easily be able to get your hands on them in Europe.

http://www.instawares.com/ql3-sin...p-ev925914.0.7.htm

If you were concerned about a leak, you could easily isolate the pressure to the machine using this. This would give a fair amount of back up in case of a leaky speedster! You can pretty much locate it where you want. Alternatively, I have a pressure regulator which can be easily adjusted to zero and an inline valve downstream of that with a simple on/off. Thats where my setup currently ends (ie no machine) and it hasn't flooded my kitchen yet (although I haven't been home in 3 weeks!).

There are a few other posts I have seen here as well regarding water protection, some have very flash setups in this respect.

http://www.home-barista.com/espre...achines-t7521.html

I figure either machine you are considering is worth the effort, it really is great having it all plumbed in.

Cheers, Chris

If you ever get bored, you could mount a small faux supercharger and air intake on the cup warmer. Hook the charger compressor to the pump so when you shift into gear, the supercharger spins and intake butterflies open just to add that extra vroom the speedster, or not.

Any way, that group water path is very interesting. The gicleur location would make it easy to change if you were so inclined. They sure have a long water path, down side would be more tubing to descale but that is a small price to pay for one of the most tricked eye candy machines out there.

cannonfodder wrote:If you ever get bored, you could mount a small foe supercharger and air intake on the cup warmer. Hook the charger compressor to the pump so when you shift into gear, the supercharger spins and intake butterflies open just to add that extra vroom the speedster, or not.

Hi Dave:

I really like the idea of extra vroom. But what's a "small foe supercharger?"

It was supposed to be faux, but Word decided to auto correct it to something entirely different. I once misspelled categories and Word decided what I meant to say was cat orgies, figure that one out.

How do the shots between the speedster and GS3 compare? Is there a noticeable difference in the cup or is the speedsters extra cost simply a matter of art and exclusivity? That is the only box I would trade my A3 in for.

AndyS wrote: Taste testing continues....

Yesterday Owen O'Neill ("Phaelon56") was over to help me pull shots. And today John Gant from Gimme Coffee did the same. Silvia was adjusted to duplicate as much as possible the Speedster temperature and pressure profiles.

Silvia shots generally seemed to have a high note "glare," while the Speedster shots included a mid and bass note register that made them better balanced and more interesting.

Yesterday Owen brought over a coffee (roasted by a third party) that had been slightly scorched in the roaster. The ground beans displayed a carbony note that reminded me of my old Starbucks days. Pulled on Silvia this carbon note was very obvious. On the Speedster it was evident, but other flavors were there to partially mask it and the shot was clearly superior.

Today John brought some Gimme French Roast blend which was excellent on both machines, but particularly rich and buttery on the Speedster. He also brought some Kenya which made a surprisingly fruity, balanced shot on the Speedster. On Silvia the body was lacking and the flavor was rather one-dimensional on the high note lemony side.

These differences weren't real subtle, they seemed to be pretty consistent over quite a few shots. The shots were not blind-tasted, however. So we could have been fooling ourselves, but I'd be shocked if these differences couldn't have been repeated in blind tasting.

Of course you'd have to be blind not to notice how the Speedster is just a wee-bit better looking than my Tricked-out Silvia, and a lot more fun to use! Below is a photo of Owen showing how the typical espresso geek feels after spending a couple hours with the Speedster: