Posted Sat Mar 9, 2013, 11:47pm Subject: Re: Custom Thermocouple for Gaggia Classic
I had about an hour to make a mount for my V2 temp probe. It didn't turn out exactly as I expected, but It turns out I cut the ceramic tube too long so I had to accommodate for that, as you can see in the picture.
There isn't anything special going on here, just a block of T6061 aluminum that I sized up, drilled, and machined out. Then pressed the bushing into the mount. It got stuck before I was able to any sealant in the hole, but it was a REALLY tight fit, so I'm hoping it will seal ok. We'll see in a few hours when I install it into my Classic. I wanted to press the bushing into the mount a little more, since there is still about 1/8" at the top, but I was afraid to snap the tube. Although the ceramic tube was NEAR impossible to cut, it is also brittle. I used about 100lb of pressure to get the bushing in there... that should explain it. It's also the weekend and I don't want to make another probe. The alumina ceramic tube was a bit on the expensive side, so I wouldn't want to make another one anyway.
Posted Sun Mar 10, 2013, 1:05am Subject: Re: Custom Thermocouple for Gaggia Classic
OK. Version 2.0 is installed on the boiler. There appears to be a leak, but not in the probe like I was anticipating. I guess the opening and closing of the steam valve had worn out the o-ring. I guess I'll be at the hardware store, looking for an o-ring. I'll replace it with a high-temp o-ring later. The leak isn't too bad. Just a little sizzling at the top, but nothing major. The 100lb press fit is A-OK. Phew.
Down to the nitty-gritty. The probe is sunk and taking temps. As I mentioned, the thermocouple itself is short 1.8F, reading a steady 210.2F at boiling for 15+ minutes so we know we have to offset by just less than 1%. Watching the temps for about 15 min, the monitor is reading up to 219.2F. The ceramic is not doing as good of a job as I had hoped but still pretty good, compared to a straight brass tube which read at about 235F.
Technical stuff without getting scientific, the algorithm will look like: water_temp = ((measured_temp * 0.99) - 5).
If anyone is attempting what I'm doing, the 0.99 is the accuracy of the thermocouple. Each thermocouple will have a different value. You'll have to boil yours (a known, constant value... also altitude will be a factor but for most of us, I believe it to be 212). The 5 (5F) is extra thermal transfer through the ceramic tube. This will vary as well, depending on the composition of the tube as well as length.
So just it's clear, the total temperature is multiplied by the accuracy percentage first to remove the thermocouple's tolerance from the equation. What's left above 212F is coming from the boiler wall. I know the equations are much more complex than this, but I don't think the Arduino is powerful enough to handle accurate scientific equations... and controlling all the aspects of the espresso machine all at the same time. So, I think that simplified algorithm is accurate enough for this application.
Now to make adjustments to the firmware and we'll start seeing "closer to actual temps"! I'll log some temps and post the plots soon.
Posted Tue Mar 12, 2013, 11:37pm Subject: Re: Custom Thermocouple for Gaggia Classic
Alright, I finally had some time to work on the machine and the firmware.
Now that the new Version 2.0 probe is in the water and I took out the R-SSR and replaced it with a 12V transformer that I stepped down to just below 5V to get my "logic-high" when the thermostat kicks in, otherwise, my input on the Arduino is pulled low via pull-down resistor. This is the chart that I got from the datalogger.
There is a glitch, which I believe is in the wiring. Nothing is hardwired, dangerously wirewrapped at the moment. Don't try this at home... or anywhere else. This is a temporary setup and will be replaced by hard wires soon. I believe on the second recovery, the tstat kicked hard enough that the wire caused a fault reading, but you can make out what is happening there. It is very consistent about kicking on at the low of ~202F and turning back off at ~203F. That's not a lot of ON time. As you can see, Gaggia designed the boiler and chose a tstat that gets the water temp JUST there (~212F). Sometimes (depending on the situation I suppose), it JUST doesn't make it to the top. In my plots as well as my every day monitoring, a lot of times it just makes it to about 206F and starts dropping.
My next step, set up my firmware to kick on at 202F and off at 203F and see if my data comes close to this chart. From there, I can have some fun with temp-shaping.
Posted Wed Mar 13, 2013, 4:55pm Subject: Re: Custom Thermocouple for Gaggia Classic
That test will be done. I just need to be able to connect a bunch of thermocouple amps together. The Arduino doesn't have enough I/O pins to connect more than... maybe two at a time, but the portable Arduino dev board I made was compact and can't fit more devices on it.
I'll have to wait until I have more time to put a bridge together to hook up more thermocouples.
I think the probe will definitely be staying in the boiler, but perhaps Version 3 or 4. The ultimate version of the probe will have provisions for steam pass-through so I can have the probe and steam. Right now, it's one or the other. When the system runs dry, I have to take the probe out, reinstall the steam valve to prime the system and... you get the picture.
The Version 3.0 will definitely have a shorter probe, whether standalone or in addition to the longer probe. Somehow, I knew I should have used the shorter piece of ceramic tube. It's so freaking hard to cut, even carbide won't cut it. At least I still have the short piece to make a short probe. :D
I also need more material and another basket to make a new thermofilter. So many things to do. But I'm definitely going to add SSR support so I can disconnect the thermostat and drive the SSR to replicate the thermostat condition. It might take a few days though. I've been working 18 hour days again... for the last couple weeks. I might need a break.
Posted Tue Mar 19, 2013, 10:52pm Subject: Re: Custom Thermocouple for Gaggia Classic
OK, I'm too lazy to start a new thread for my thermobasket (version 2.0) so here it is.
I didn't spend as much time and effort on this as I should have/could have, but I just wanted to throw something together so I can get some data. The way I made this thermobasket is a way anyone here can, except boring out the 1/8" copper tubing.
First, about the tubing, just to get it out of the way first, so we don't have to talk about it anymore. I was able to get my hands on some 1/8" copper tubing with ~0.064" ID. This is WAY too small to feed a couple fiber-jacketed thermocouple wires so I bored the hole out to about 0.100", which leaves a wall thickness of about 0.0125". Is this significant? No. I just didn't have any suitable capillary tubing. Anyone doing this, just use some cap tubing and be done with it.
Oh, I also threaded the end with a M3-0.8 thread so I could cap the end. I don't believe in soldering (potential of lead and flux contamination) any food-related part. The end-screw was a modified M3 phillips head screw, turned down to 0.125" to match the tube. Other than that, everything was easy to do. Anyone could probably solder, braze, or plug the end any way they can. This would make it liquid-tight, if done properly.
The copper tube was cut and bent to... well, the appropriate shape. and fed through a rivet (I believe a 5/32, which gives an approximate 1/8" hole, once the pin is removed). Obviously, an old double-shot basket was sacrificed and a hole drilled towards the bottom wall. This was just quickly sealed with thermo-epoxy. Any high-temp epoxy should do, but something food-safe would be best. I didn't check if my epoxy is or not, but exposure should be very minimal. The copper tube was also epoxied in place to secure it.
Of course, the thermocouple element was inserted before any of these steps were done. I know, these steps are backwards, but this is how my brain works sometimes. The rest is just standard wire stripping, crimping, and cleaning up. Please let me know if this sounds like mud.
Anyway, as always, everything is sealed, except maybe the side of the tube where the thermocouple wires enter, but I figured it is outside and above the bottom, so it wouldn't get wet anyway.
Testing: Again with the boil test - it turns out, even through my quickie job proved to seal the entire application very well. There were no short circuits or anything. This was just a quicky until I made a more permanent one. I would hermetically seal the new one... but whatever. I have a thermobasket I can use to get data. I was very surprised that this one is approximately 0.5% optimistic, which makes my calibration value 0.9957, as opposed to 1.0. I work in the 0.0001" at work, so excuse the 4 digits. This is better than my probe, which was calibrated at roughly 0.9900.
How it works: Ok, here's the beauty of this basket. It doesn't look like much. It doesn't look very elegant, but the thermocouple is placed approximately where the coffee puck would sit, approximately in the center of a double-shot puck. I assume most people here are pulling doubles, if not triples. The flow media is... espresso grounds. Yes, I figure there is no better medium that has the same density and thermal coefficient than coffee grounds.
So, a double is dosed into the basket. I have to WDT this basket since my grinder clumps, taking care not to knock the probe around too much, but making sure I pack under the probe the best I can. Tamp with my calibrated tamp, and pull the shot.
Chart explanation: The beginning 85F is approximately the temperature of the PF, radiating heat to the probe. I normally keep the PF in the grouphead, but I had to pull out the basket and install the thermobasket. The temp dropped, but that doesn't matter at all. This is also after the doze was WDT and tamped.
At approximately 64 is when I hit the brew button. It did take some time for the water to come down, soak the puck. I assume full saturation happened at approximately 118 @ ~194F and maintained a good window between 194-187F throughout the shot.
At about 244 is when I dumped the puck and stopped the datalogger.
Results: Everything on my Classic is stock, except the internal probe. We see that stock, out of the box (this machine is roughly 6-10 months old with duties of pulling 3-4 shots daily, and occasionally 12-15 shots for house parties), the brew temp is quite low for a "text book" espresso shot.
Next steps (if time permits): - Add additional thermocouple amplifiers to the Arduino to take multiple readings simultaneously. - Gut the darn thing already so I can get the temps up so I can finally have a GOOD espresso!
Well, my Handspresso from Whole Latte Love came today, so I'll be playing with the the next few days too.
Posted Wed Mar 20, 2013, 8:43am Subject: Re: Custom Thermocouple for Gaggia Classic
I think that I understand what you did, coffee as the puck and then thermocouple in the tube. I am not as sure that I follow the time sequence. If a pull is about 25 seconds, how does that figure in? Did the heater click on during the pull, or did you surf to keep it on? I ask because I do not see a drop with cold water inflow. What was the water volume?
Not a criticism, just a note for anyone else who wants to try this. KS copper tube available in hobby and craft stores is 0.125 OD and ID is 0.096" without boring. It will easily accept this k Thermocouple.
The seller mentions plastic and teflon as the coating. On the coper tubing, you can solder the tip and not drink the brew from that run, or get realatively safe solder. Copper water pipes are soldered without lead.
Symbols: = New Posts since your last visit = No New Posts since last visit = Newest post
Forum Rules: No profanity, illegal acts or personal attacks will be tolerated in these discussion boards. No commercial posting of any nature will be tolerated; only private sales by private individuals, in the "Buy and Sell" forum. No SEO style postings will be tolerated. SEO related posts will result in immediate ban from CoffeeGeek. No cross posting allowed - do not post your topic to more than one forum, nor repost a topic to the same forum. Who Can Read The Forum? Anyone can read posts in these discussion boards. Who Can Post New Topics? Any registered CoffeeGeek member can post new topics. Who Can Post Replies? Any registered CoffeeGeek member can post replies. Can Photos be posted? Anyone can post photos in their new topics or replies. Who can change or delete posts? Any CoffeeGeek member can edit their own posts. Only moderators can delete posts. Probationary Period: If you are a new signup for CoffeeGeek, you cannot promote, endorse, criticise or otherwise post an unsolicited endorsement for any company, product or service in your first five postings.