Decent Espresso Machines (DE1) - Any thoughts?

[>decentespresso]

Decent Espresso ❤️ Espresso Forge



Decent Espresso ❤️ Espresso Forge

I've been a big fan of Andre's Espresso Forge https://www.facebook.com/espressoforge/ http://espressoforge.com/ project for some time https://www.home-barista.com/espress...ct-t34309.html and was very flattered when Andre bought a Decent Espresso Machine.

A few years ago I'd pulled shots on a forge, found that they were excellent, but better with our Decent Baskets. Andre agreed and now sells our baskets as an option for the forge.

The Forge can make some truly excellent shots, for two reasons. Firstly, it's a lever machine, so you have manual control of flow and pressure. https://www.youtube.com/watch?time_c...&v=E3vtVYT_rhw

Secondly, because it takes pre-heated (kettle boiled water) it has a natural "temperature profiling" capability, with shots starting high and ending at a lower temperature. A number of coffee experts think that a declining temperature curve makes better coffee.

Indeed, Scott Rao's latest "blooming espresso" discovery on the DE1PRO makes use of a declining temperature curve. https://www.scottrao.com/blog/2018/7...ode-on-the-de1

Andre has developed two advanced profiles for his DE1+ that mimic the best technique he's developed on the Espresso Forge. His idea is that when at home (where he has electricity!) he and his wife can easily have the best espresso they've become used to.

And when they're on vacation or camping, they take their Espresso Forge and recreate this espresso by hand.

Andre's two profiles for the DE1+ take advantage of temperature profiles, low-pressure preinfusion, multiple stages with conditions, flow and pressure profiling. This is fancy stuff, but Andre knows his stuff.

Thankfully, he's created and tested this recipe, so the only thing the rest of us need to do is tap the preset and his START.

These two profiles are now included by default to all our Decent Espresso customers, as a free "App Update"

-john

[>decentespresso]

The End of the Line



Next week, we'll finish building the last of our v1.0 Espresso Machines. We'll have about a week pause, as we await the parts for the v1.1 design to start arriving. The new parts will be arriving over a 2 to 6 week period. We won't be able to ship any v1.1 machines until every last part has arrived, likely around the end of September. But we'll have done a lot of prep work, so they should flow out fairly rapidly then.

Making machines at an increasing speed and good quality has been our focus these past 6 months with v1.0. The pause gives us time to tidy up, document, rethink how we do some things, and (vitally) find space to store 2x as many parts as we've ever had.

I took these two panorama photos early this morning, as I was the 2nd one to arrive, and things were nice and quiet. You can see how our factory, warehouse, and engineering spaces look. That's the final set of v1.0 machines (all 110V) being built, on the main assembly table.

-john

[>decentespresso]

Filter bliss



It took 9 weeks, a lot of negotiating and paperwork, but I today have in my hand 20 precision filter screens by IMS. Made in Italy.

Some Decent customers have put these screens into their machines, and reporting more even water flow, as these screens resist getting "gunked up" better than our current screen.

If my tests over the next 2 weeks find the same benefits, then IMS screens will become what we use in the v1.1 models we're going to soon start building.

-john

[>decentespresso]

The Acid Test



Here are the results from testing both PVD and Electroplating coated drip tray covers, in full strength Rinza Rinza® Acid Formulation Milk Frother Cleaner - Urnex Professional

Note that this stuff is supposed to be 15x diluted, so this is a harsh test.

The PVD did not fare well.

Some of the PVD dissolved in the acid, and when we washed it residue off, you can see where the PVD is gone, leaving shiny electroplating below:

Interestingly, there is no "boundary line" visible, where the air met the Rinza liquid.

With the electroplating, the fully submerged portion suffered no damage. However, a boundary line is clearly visible where the air meets the liquid. Is there a chemist in the room who can explain why?

Of the two approaches, I think I prefer the Electroplating, but neither approach can withstand pure Rinza.

This weekend, for 2 days, we're soaking the samples in 4x diluted Rinza (aka "quadruple strength") as this is likely a more reasonable test than 15x strength pure Rinza.

-john

[Lyrebird]

The answer depends on what the plating and substrate are. Looks like nickel so I assume you have a copper base layer on steel?

A combination of a pH extreme and oxygen will often cause rapid corrosion, for instance titanium is very corrosion resistant, neither sodium hydroxide nor hydrogen peroxide cause any damage at 1M concentration. Combine the two and you have an excellent etchant.

[>decentespresso]

The answer depends on what the plating and substrate are. Looks like nickel so I assume you have a copper base layer on steel?

A combination of a pH extreme and oxygen will often cause rapid corrosion, for instance titanium is very corrosion resistant, neither sodium hydroxide nor hydrogen peroxide cause any damage at 1M concentration. Combine the two and you have an excellent etchant.

We don't have clarity at the moment on the mix, but we're asking now and I'll report when I hear back.

Will also pass on your oxygen/ph tip.

-john

[level3ninja]

Will also pass on your oxygen/ph tip.

I don't think he means during the plating, I think he means while you've dipped in the acid for your testing the air immediately next to the surface provides oxygen for a speedy oxidation thanks to the low pH of the liquid. That is always going to happen.

[Lyrebird]

L3Ninja is correct, sorry if what I posted was misinterpreted.

Actually I am at a loss as to why you are looking to use a plating process for the drip tray cover in the first place, I wouldn't have thought this was a terribly expensive part to have made in SS?

I know of no plating process that will stand up long term to acidic / oxidative conditions and inevitably when the plating peels it looks terrible.

[HBexile2]

[ATTACH=CONFIG]19758[/ATTACH]

Here are the results from testing both PVD and Electroplating coated drip tray covers, in full strength Rinza Rinza® Acid Formulation Milk Frother Cleaner - Urnex Professional

Why use a milk cleaner? The only time I have ever seen it being used is for cleaning super auto machine's milk frothers. You should test with backflush detergent as that is what will be used all the time.

Just go with stainless and be done with it.

[>decentespresso]

Why use a milk cleaner? The only time I have ever seen it being used is for cleaning super auto machine's milk frothers. You should test with backflush detergent as that is what will be used all the time.

I'm testing the drip tray with both a group head detergent, and a steam wand cleaner, because both chemicals will find their way to splash onto the drip tray, and potentially stain it.

Just go with stainless and be done with it.

Unfortunately, manufacturing is always a set of compromises, and this design, which we were told was doable in stainless, we now find that no manufacturer wants to take the risk to do for us.

The stainless casting companies have wanted us to make quite a few modifications to the current design, which would make it a lot less functional and uglier. For example, switching to stamped stainless is doable, but would cause water to bead up on the "wires". Or if we stay with casting, adding crossbars and thickening the wires quite a bit.

Stainless is doable, but it's (a) very time consuming and (b) not assured of success, (c) so far, looks likely to require ugly modifications of the current design.

In the meantime, we need to keep shipping machines.

For me, the current design has the virtue of:
1) it works now
2) it's shipping
3) it's attractive

the only downside is the tarnishing problem, which should be solvable in a few ways.

-john

[>decentespresso]

Don’t go with the flow



Don’t go with the flow

We’ve had a lot of trouble with our flowmeters. So much trouble, in fact, that we’ve invested months of dedicated R&D to work around its defects, and to try to invent our own hybrid flow meter. The new “method” will (theoretically) count pump strokes to be low latency and capable of tracking low flow rates, and then use the physical flow meter. I write “theoretically” because it’s proving to be a hard problem to solve.

We’re currently using a big-name Italian-made flow meter. You’ll find this thing in most commercial machines. It’s ok. Really. But it only delivers 90% accuracy. And its electrical pulses are noisy and need filtering. And sometimes it double-pulses. And other times it skips a pulse. All this means that we currently wait 6 seconds before we can trust the numbers its giving us.

I’m currently looking at a flow meter made by EPT. They make the same model in 98%, 97% and 95% accurate versions.

I asked them what the difference was, and the answer surprised me.

They’re all the same.

But because small differences in tolerances have effects on the accuracy, they test each flow meter, and then sell them at 3 different prices based on how accurate that one was.

I asked them how they thought their flow meter might be better than the big-name Italian one we use today. Three people came to visit us to deliver the answer (the lead engineer, the Big Boss, and the salesperson). We found:

1) they’re using higher grade nylon and a higher tolerance process. Even then, though, there’s import variation between each flowmeter made.
2) their tines are curved slightly, rather than straight, which they feel reacts better to water flow
3) their magnets are larger, and deliver stronger, less noisy electrical pulses
4) their tines are slightly taller and closer to the edges, so less water can slip through the cracks.

I’ve been using EPT’s flow meter in my own personal machine for a few weeks. It definitely works. Over the next few weeks we’ll be testing 20 of them, to see if they are indeed better.

They’re slightly more expensive, but if we can get a better flow meter, the additional cost will be very much worth it.

In the photo, you can see the old vs new flow meter models, and in the espresso machine profile photo, the flow meter can be seen in the top left.

-john

[Lyrebird]

That's been standard practice in electronics manufacturing for years. The devices are fabricated en masse and then sorted into bins on performance. The highest spec units sell for a premium, the lowest spec are sold off to second sources.

Check , for instance, Cree's LEDs. You can buy the same LED at a range of efficiencies and a corresponding range of prices.

[gc]

John, the usual question from V1.0 users - will this be an easy upgrade path? It looks like a simple replacement job, but the V1 firmware would have to support it I guess. Alternatively, is there any way to use the Skale output to manage flow in real time? (I REALLY want to use flow instead of pressure, but it seems a bit variable so far)

[>decentespresso]

John, the usual question from V1.0 users - will this be an easy upgrade path? It looks like a simple replacement job, but the V1 firmware would have to support it I guess. Alternatively, is there any way to use the Skale output to manage flow in real time? (I REALLY want to use flow instead of pressure, but it seems a bit variable so far)

Yes, the new flow meter has identical operating specs, uses the same plug and mounting. It just "works better". No firmware change needed. Two minute swap, you can definitely do it yourself.

-john

[>decentespresso]

That's been standard practice in electronics manufacturing for years. The devices are fabricated en masse and then sorted into bins on performance. The highest spec units sell for a premium, the lowest spec are sold off to second sources. Check , for instance, Cree's LEDs. You can buy the same LED at a range of efficiencies and a corresponding range of prices.

Agreed, but it was news to me that this same rule applied to something mechanical like a flow meter. I didn't realize that (a) the tolerances were so important and (b) that it was so difficult to "nail" the desired tolerances.

-john