Decent Espresso Machines (DE1) - Any thoughts?

[TampIt]

<grin> That is EXACTLY the approach we are taking. :-)

The v1.1 Espresso machine models are able to measure current draw, so this is now a possibility.


Yes, of course, that's a fallback. However if it's just another weeks work to get voltage autosensing, then that's my preference. That's we're trying to make work.

-john

G'day John

Thanks for the fallback on voltage.

I will happily wait another week if needed. If it can be done via a software update, all the better...

TampIt
Do it once, do it right.

[MrJack]

<grin> That is EXACTLY the approach we are taking. :-)

The v1.1 Espresso machine models are able to measure current draw, so this is now a possibility.


Yes, of course, that's a fallback. However if it's just another weeks work to get voltage autosensing, then that's my preference. That's we're trying to make work.

-john

Hi John,

You mentioned this data was generated using a known diameter single hole basket. Do you provide these to customers with the machine?

If so, it might offer another way to do a manual calibration of the machine, since the relationship between pressure and flow through that orifice will be the same regardless of the pump behaviour (and this could be measured in factory)?

Plotting the data with flowrate on the x axis and pressure on the y axis would show the difference voltage makes much clearer.
I did this with the data scraped from the 230V chart you posted, the result was quite interesting!

[>decentespresso]

G'day John Thanks for the fallback on voltage. I will happily wait another week if needed. If it can be done via a software update, all the better... TampIt
Do it once, do it right.

Actually, I decided that there's no reason to hold up sending 240V machines for this improvement, since we're frequently coming out with firmware updates that improve things. So: five machines at 220V are going out today.

The next 220 firmware update will be in a week or two and is simply a tap on the tablet to update the software, and then another tap to update the firmware. Easy squeezy cheesy peasy.

If you are running 220V or 240V., all this means is that your flow estimates will be off by as much as 10% when you are at eight bars or higher pressure,

However, Tampit if you are in Australia my understanding is that 230V is standard
https://www.tripadvisor.com/Travel-g...ppliances.html

In which case our current firmware works perfectly, as that's what we calibrated our flow mathematics for.

-john

[>decentespresso]

Hi John, You mentioned this data was generated using a known diameter single hole basket. Do you provide these to customers with the machine?

Yes, that is 100% the idea. I posted about the "puck simulator" accessory here on this forum some time ago:
https://coffeesnobs.com.au/brewing-e...tml#post631924

Armed with a Bluetooth scale, a 0.2mm basket (will be available for sale soon ~AUD$30) and the espresso calibration program (which is simply an advanced shot), you'd be able to re-create what we have done.

Plotting the data with flowrate on the x axis and pressure on the y axis would show the difference voltage makes much clearer.
I did this with the data scraped from the 230V chart you posted, the result was quite interesting!

That's a good insight.

However, I didn't mean to make you scrape the graphics for the data. Just ask: I'm happy to share the data itself on anything I post.

Here are three CSV documents showing the three calibration tests at different voltages:
http://magnatune.com/p/presstests_csv.zip

Note that these CSV export documents are now created automatically when you save a "god shot", as part of a recent tablet software upgrade I put out.

-john

[>decentespresso]

The first manufacturing prototype of the decent scale arrived on Friday.

I am mostly quite happy though of course, I have about 20 defects and mistakes I cataloged for our manufacturing partner. Not least of which is the red numeric LED (!!!!???!!!).

Most importantly, the really hard things they nailed ( The plastic is high quality, doesn't smell, and the mold errors are tiny and repairable), and their attitude to my change list was very positive, so in a few weeks, I should have the second iteration. And I should also get Bluetooth functionality so I can program my tablet to talk to it.

You can also see the shiny new electroplated drip tray covers, which all of you will be soon receiving.

-john

[level3ninja]

However, Tampit if you are in Australia my understanding is that 230V is standard
https://www.tripadvisor.com/Travel-g...ppliances.html

In which case our current firmware works perfectly, as that's what we calibrated our flow mathematics for.

-john

230 is the "standard" as of a few years ago. They changed the "standard" to 230 from 240 but they also changed the allowable variance to the nominal voltage at the same time. 230 V as the nominal voltage with a +10% to –6% variation at the point of supply. (253 V to 216.2 V). 253 was the maximum back when it was nominal 240, but the minimum at nominal 240 was something like 228(?).

So what changed at the outlets in our walls when they changed the nominal voltage? Sweet FA, literally nothing. It still depends on how far you are from the nearest transformer. I've had friends who kept blowing appliances after a few months, CFL lightbulbs you could count in weeks. I measured the voltage at the outlet and it was 248V. Outside the "preferred range" of +6% to –2%. (243.8 V to 225.4 V) but well within the allowable limits. The lowest I've every measured in a house is 236V, which was such a surprise I measured a few other outlets and the voltage coming in on the mains to make sure there wasn't a high resistance fault somewhere in the switchboard. I've seen 238 a few times, 239 a handful of times, but 240+ is by far the norm (in Sydney at least).

TampIt is near Perth, I replaced some lights for a friend a few KMs from him last year and I think it measured low 240s. His house could be different though depending on the distance from the Tx.

[>decentespresso]

So what changed at the outlets in our walls when they changed the nominal voltage? Sweet FA, literally nothing. It still depends on how far you are from the nearest transformer. I've had friends who kept blowing appliances after a few months, CFL lightbulbs you could count in weeks. I measured the voltage at the outlet and it was 248V. Outside the "preferred range" of +6% to –2%. (243.8 V to 225.4 V) but well within the allowable limits. The lowest I've every measured in a house is 236V, which was such a surprise I measured a few other outlets and the voltage coming in on the mains to make sure there wasn't a high resistance fault somewhere in the switchboard. I've seen 238 a few times, 239 a handful of times, but 240+ is by far the norm (in Sydney at least).

Very helpful info. So, obviously, every machine we ship has to self-adjust to whatever it's given. No "I live in Australia" setting would tell us much.

-john

[level3ninja]

I'll have to see if I can find an old copy of AS60038, and also look through the old copies of AS3000 & AS3008 at work. It was the drop in voltage that caused headaches for us (and fatter wallets for the cable manufacturers) by requiring us to calculate voltage drop on 230 instead of 240, pushing a lot of formerly normal circuits over the threshold to the next size up cable. From memory the 10% had been dropped prior to 230 coming in and bumping it back up again.

[Dimal]

We've got a very responsive energy service group out our way, luckily.
I charted the voltage variation in our neighbourhood over a two month period to prove a point.
Voltage used to peak at 262V AC and bottom out at 238V AC for a nominal 230V AC supply. Requested that they check out the local kiosks, sub's etc... and within a week, the service crew turned up at our house to verify my instruments against their own. They were satisfied with the accuracy and method used then indicated that they would be happy to do some measurements of their own around the local district.

The upshot was, that they adjusted the tap settings on a main distribution txfr and then did the same to a number of kiosks/sub's as well. Our local voltage now sits reliably between 228 - 242V AC so not too bad an outcome in the end. Anyone whose local supply voltage is significantly out of spec' might benefit from completing a similar exercise, or engage someone who can do it for you so that your case can be stated and hopefully addressed.

Mal.

[TampIt]

Very helpful info. So, obviously, every machine we ship has to self-adjust to whatever it's given. No "I live in Australia" setting would tell us much.

-john

G'day John

It actually is worse than that here in Oz in terms of power. Read on if you dare / care.

Rumour from when my family moved from NSW* to West Oz in 1964: SECWA, our local electricity supplier, had recently bought a pile of secondhand transformers from South Africa which were 260V and they could only be modified down to 250V. Fact: Vic Park was 258V to 264V at the time. Most of our precision gear blew within a couple of months so my dad actually fitted a voltage reducer to our Vic Park house switchboard to take it down to 230V. Most Perth suburbs at the time were still measuring around 255V, and Osram must have made a fortune because they actually made a 260V light bulb for sale in West Oz "at a price" - the only bulbs that did not destroy themselves in a few days. About 10 moves later (civil / safety engineer dad on mining / oil rig projects) in 1972 we moved to Mackay in Qld. We were the only house at one end of a brand new suburb - voltage swings from 450 (yep - they screwed the phases up the day we arrived and started to unpack in our brand new house) down to 245V. Dad bought a large Variac whilst the Qld electricity board processed a several thousand dollar claim for a brand new Wurlitzer electronic organ which happened to be the first thing we plugged into the new house (music while we unpacked was S.O.P.). One huge bang and it launched my mum a few metres backwards. I have seen less electronic damage from a lightning strike (the 450V was constant at a high amperage, lightning is a low current, rapid, one shot event). The electricity board also covered an additional three weeks at a hotel whilst they sorted the power out. An "after complaint / regulation" 160V to 245V random cycle (no joking!) was their best effort until we moved again a year or so later.

I have also read a couple of technical articles which claim that tinkering with voltage can save the electricity suppliers some money as the house meters actually read more than the provided power. Not my professional area, however I would not be surprised at any power company dirty tricks here in Oz.

Back to the west. Over the last year in the West I have measured voltages from 210 to 255 at various metro / outer metro (i.e. half of greater Perth including Rockingham is outer metro) locations. Most of them also have cyclic voltage variations over each day however the frequency is very stable. No problem for cheap & nasty switch mode power supplies but heaps of grief for a lot of more expensive gear using better quality power supplies.

Your "I live in Oz" would only forewarn you that the voltage regulation at too many places is fairly notional. Perhaps a bit of software which senses the voltage at the beginning of the shot and modulates accordingly would be a good move? Not everyone has a variac.


TampIt


NSW*: i.e. level3ninja's home turf nowadays where they have been slowly (erratically?) working down from 240V since the '50's. Cannot rush these things whilst the overcharging by privatised power companies keeps their profits up...

[MrJack]

Yes, that is 100% the idea. I posted about the "puck simulator" accessory here on this forum some time ago:
https://coffeesnobs.com.au/brewing-e...tml#post631924

Armed with a Bluetooth scale, a 0.2mm basket (will be available for sale soon ~AUD$30) and the espresso calibration program (which is simply an advanced shot), you'd be able to re-create what we have done.


That's a good insight.

However, I didn't mean to make you scrape the graphics for the data. Just ask: I'm happy to share the data itself on anything I post.

Here are three CSV documents showing the three calibration tests at different voltages:
http://magnatune.com/p/presstests_csv.zip

Note that these CSV export documents are now created automatically when you save a "god shot", as part of a recent tablet software upgrade I put out.

-john

[ATTACH=CONFIG]21091[/ATTACH]

Thanks John! I would have asked for the data at some point, I was just looking for something to do on the train ride home!

Great that you've incorporated the csv export functionality. Much easier to parse than the old format. I would note that having spaces between the values makes it more readable in text format, but it would be better without them from a data analysis perspective (less steps to clean up the data).

I knocked together a few plots (colour coded by nominal voltage), one with the raw data and one with a crude calibration (which actually seems quite good).

[>decentespresso]

would note that having spaces between the values makes it more readable in text format, but it would be better without them from a data analysis perspective (less steps to clean up the data).

Good suggestion. I've now gone ahead and done that both for the standard CSV export and for the new EEX CSV format that we have developed in partnership with SEP https://www.naked-portafilter.com/sm...esso-profiler/

was:

espresso_elapsed, espresso_pressure, espresso_flow, espresso_flow_weight, espresso_temperature_basket, espresso_temperature_mix, espresso_weight
0.0, 0.0, 0.0, 0.0, 88.0, 88.0, 0.0

is now:

espresso_elapsed,espresso_pressure,espresso_flow,e spresso_flow_weight,espresso_temperature_basket,es presso_temperature_mix,espresso_weight
0.0,0.0,0.0,0.0,88.0,88.0,0.0

I knocked together a few plots (colour coded by nominal voltage), one with the raw data and one with a crude calibration (which actually seems quite good).

That's great news.

Ray was hoping that a simple calibration based on these charts would work and was about to apply it in a few days.

He's working on a phase locked loop as part of determining voltage accurately, and once he has that working will be applying a correction in much the same way as you have.

-john

[>decentespresso]

Refill kit video

After nine months of work, we finally have the three components of our refill kit finished and in stock. I made a video yesterday showing how it all works together.

About half our current customers have purchased the Pro version of our Espresso machine and have been very patient in waiting for us to finally ship these three parts which will allow them to fully plumb their machines in. I'm hoping to start contact next week with all those Pro customers, verifying their current postal address, and also talk to them about the free update pack they should be receiving (of things we've improved since releasing our version 1.0 machine).

-john

[>decentespresso]

Calibrating to reality



Before we ship an espresso machine to a customer, we always calibrate temperature and pressure at the puck, uniquely for that machine.

Small physical variation in pressure sensors and temperature sensors are a real thing, which is why we use a Scace 2 https://www.espressoparts.com/scace-...ressure-device to verify that our puck measurements are correct to within 0.1ºC.

I asked Parry to keep track of the calibration numbers that he puts into the firmware for the first 19 machines we shipped of our v1.1 model. I've charted and attached the data here.

You can see that our pressure sensors don't have much variation, whereas the temperature sensors do need typically 1/2°C of calibration from each other.

Also note that all temperature and pressure sensors are slightly not as the manufacturer has stated (~1.69ºC for temperature, and 1.23 bar for pressure), which is why we calibrated to the Scace.

I prefer to never trust a manufacturer's specification sheet when we can measure against a known accurate benchmark.

-john

[>decentespresso]

Ben's DE1CAFE research setup

Decent engineer Ben Champion shows us a tour of his working setup.

He’s aiming to increase the power, control the dry/wet mix and controllability of steam for our upcoming DE1CAFE espresso machine.