Controller part 2 - nearly boiling water

Today I (nearly) boiled water.

The result of the squirrelly form factor of the thermocouple is this ridiculous adaptor:

I put a drop of heat-transfer grease on the tip of the TC to aid the conduction of the heat and attached the whole lot to the boiler. Then I did a quick test of the controller using a lamp in place of the heating element.

Now for the real thing. I put the multimeter TC probe directly into the boiler so that I can compare it with the controller TC. Time to 'put the kettle on'. 

As I feared, the lag created by the thermal mass of the lump of brass in the adapter and the less than ideal position in which it is installed (close to the bottom of the boiler) is significant. After about 8 minutes, the water got pretty close to boiling but the controller probe was nearly 30 degrees behind. After another ten minutes or so, the gap had narrowed to about 9 degrees, but this is far from ideal. Quite possibly, everything will stabilise once the heat is completely diffused through the entire system and in theory, with some calibration of the set point and proper PID coefficients, the overshoot can be reined in. However, a twenty minute plus sample rate just isn't acceptable. To remedy the situation, I am going to move the TC to the top of the boiler and look for a better form-factor - one that can be installed through the wall of the boiler.

Controller part 1 - DIN rail mock-up

I think we've turned a corner because it definitely seems like it's time to start on the other kind of plumbing! 

Here's one I made earlier...   
(Not sure how many here will get that reference).

For those who haven't come across them before, these are modular DIN rail industrial control components - aka Lego for electrical engineers. From left to right, AC distribution, 24v DC (blue) and 5V DC (black) power supplies, DC fuse blocks, DC distribution, 12V (white) power supply in the middle, unconnected terminal blocks and a magnetic contactor for switching large loads. The grey terminal blocks come in various flavours (1x1, 1x2, 2&2 etc) and can be jumpered together, converting them into bus bars for power or signal distribution. The metal DIN rail serves as a ground bar linking all the green terminal blocks together - the equivalent of chassis ground. I put this together a while ago for another project and kept it assembled as it often comes in handy.

The 12v supply wont be necessary and neither will the contactor, but the rest will be useful. The first order of business is the temperature control. I have this nice Watlow controller which I am borrowing from another project. It has a number of features that I want to take advantage of. The first is simply speed - this thing just works: five minutes of mandatory RTFM, specify the thermocouple type, change the set point and away you go. The other advantage is that it has a PID auto tune mode which should obviate the need to empirically determine the coefficients for the three terms of the control expression.

I ordered a few of these $2 thermocouples from China a few months ago (K type if I remember correctly). They are advertised as 6mm but are in fact 1/4-20 (but the nut is metric - what a world). They also don't seal, which is a pain as it means that they can be put directly inside the boiler.

A quick side-by-side test with the thermocouple that came with my multimeter shows that all is well in this particular tiny corner of the world.

The TC is hooked to the input of the Watlow and the output is sent to a solid state relay (SSR) that will control the heating element. 

The setup so far, from left to right:

Arduino and protoboard (top left)
Unused contactor (bottom left)
Autofill probe connections
24V DC / 120V low amperage control relay (light blue)
Bidoowee brand 5V to 24V and 24V to 5V optoisolated signal convertor (green IC board)
DC distribution and fuseblocks
DC power supplies
AC distribution
SSR (top right)
GFCI outlet (to the left of and connected to the SSR) (powering the heating element) 

A couple of words of warning here. The wire gauge used is technically not adequate for the amperage drawn by the heating element. Also, we are playing with 15amps and lots of water, thus the GFCI (Ground Fault Circuit Interrupter). Out of frame, the whole mess is plugged into another GFCI on a 12 gauge extension cable for belt and suspenders safety.

Handle search

I have been having quite a time finding the right set of steam and water taps for the build. Although I know that lever style controls are preferred by some over traditional taps, I would really like to preserve the look of the original design. Personally, I find the lever valves are akin to forelegs on a Tyrannosaurus Rex - out of proportion with the rest of the beast. 



Alas, the Brugnetti valve is out of production (although there is still a little new-old stock to be found) and I have not come across any of the handles at all. I combed the databases of a number of suppliers to find a likely candidate and found one from Pavoni:



I ordered the valve body and the handle separately and was surprised to find that the handle wasn't the one in the image (it is actually nicer imho). Also, not such a nice surprise, the outlet of the of valve I received is angled back towards the body of the machine. This makes it unusable in the design and I have to find something else. I am guessing that as the Pavoni Bar 2L is still in production that the straight valves can still be found. In an ideal world however, the valve outlet would angle away from the body so that the assembly can be kept close to the backsplash. The Pavoni actually uses an additional spacer to solve this problem. Any suggestions for something closer to the original from another machine in production would be greatly appreciated. Otherwise, I will just have to sort out a right-angle Pavoni valve I suppose.

I had to install the tap that I received temporarily so that I can go on to other things.

Not right at all, but it will have to do for now.

Plumbing part 6 - corrections, P stat and drain lines

There were a couple of missteps in the last round of plumbing that I had/wanted to fix. The first had to do with the way that the parts were supported when they were brazed. The intense heat from the oxy-acetylene torch combined with the pressure of the vise caused the brass parts to soften and deform. The result was a complete set of oval compression fittings - sniff. 

I made a couple of compression dies from some scrap aluminum to try to salvage the work.

Fairly straight forward to use: drop them into the vise and crush the living *&^#&@#$ out of them. 

We will find out if these are round enough when I fill the boiler!

I used North American 1/4" tubing for very first run that I made from the boiler to the manometer. The 1/4" is really close to 6mm, but the fittings, well, don't really fit. I was also not particularly happy with the stack of adaptors on the back of the meter, so I swapped it all out for a new 4mm line with a coil to catch the condensation. The nice thing about the 4mm is that it can be bent easily and without much fear of the wall collapsing. I made the coil by winding the tubing around a cylinder by hand.

Here is the line, with the fittings brazed in place (round ones this time) and ready to be installed.

I also wanted to add a tank drain to the plumbing system. I don't really understand why this isn't a more common feature. The HX setup means that the mineral concentration of the water in the boiler is constantly increasing as pure water is drawn off in the form of steam. On the Aurora, a proper boiler flush can only be done by removing the case and disconnecting the lowest convenient fitting, which is the line to the level gauge. What this actually means in practice is that boiler flushes don't get done nearly as often as they should. I built the stainless boiler as a (nearly) drop-in replacement for the copper original. However, I don't intend to build a level gauge as I think that they are vestigial and appendix-like. This means that the ports can be used for something else - e.g. a drain!

That is an ooooold valve. I'll have to find something a little better eventually.

That, I think, is it for now for the plumbing. Here are a few shots of the finished setup.

Plumbing part 5 - HX inlet and solenoid

Pluuuumbing! 

First order of today's business is the line from the inlet to the HX - this one is always open but requires an additional connection to the expansion valve at the sump. 

I found that the tubing bender has a tendency to drag the section around the mandrel which screws both the tube and the placement of the bend. In order to get them in exactly the right place, I am using a clamp collar as a stopper. The marks on the tubing are either the start, middle or end of the bend and are aligned with the degree marks on the bender where they will end up when the bend is made; by measuring or simply rolling the straight tube around the mandrel (if that is even sort of clear). 

Here it is in the bender.

This creates a bend with a radius of about 2.75", but I need this one to be smaller so I tightened it up with the Alien-critter-tool afterwards. Two more bends added and one to go.

This is a skill that needs a fair amount of practice in order to deliver good (and precise) results. I haven't done this for a while, so the first one took quite a bit of time (and plenty of adjustment). I have lots of admiration for the electrical conduit trade (ever looked at a manual for one of those deceptively simple looking tubing benders they use?). There is no correcting bends in that stuff with your bare hands!

The glands are pulled into axial alignment with the fitting when the nut is tightened, so there can't be too much fudging. Next is the auto-fill for the boiler with the solenoid in the line...

...and the connection from the solenoid to the inlet.  

This one was done using 8mm line, which was far easier and more pleasant to work with than the 10mm. Next time, I will use only 8mm 'cause life is too short. 

Now the really fun part :)

After cleaning the parts, I put little flux on the joint and a ring of the filler metal...

...and use a reducing flame (less oxygen than gas to avoid oxidizing the metals) to heat up the parts. Once the flux melts completely showing that the parts have reached temperature, the silver filler material flows into the joint perfectly. Mustn't forget to thread both nuts on the line because otherwise you are just starting over.

Plumbing part 4 - fill reservoir and water inlet

Cutting the copper tubing can be done with a pipe cutter, but this pinches the annealed tube too much as these small diameters.
I prefer to cut with a mitre and an x-acto micro saw. This method allows you to cut extremely precisely and to remove a tiny section from the end, which is difficult to do with a regular pipe cutter.

I mounted the custom bending dies I made on a cheapo 4" machinist vise. Somehow it reminds me of the critter from Alien. 




























Starting with a roughly 2.5" radius bend made with the regular tool, the custom bender can tighten up the radius to about 1.5".

This image of the line from the exit of the HX to the brew reservoir shows both a large bend used where there is space and a tighter one where there isn't.



























I got caught up in the process of making this bracket and it was all done before I realized that I hadn't taken any process shots:



























I cut off a 3" section of 1/8" x 2 1/4" cold-rolled steel and then cut that in two lengthways. Mark and centre punch the hole position, then drill using first a small bit to locate, a couple of intermediate sizes and then the finished sized bit. Then I did a "don't try this at home" by putting the 1/8" stock in my new 16 gauge rated (i.e. half the thickness) V-jaw metal brake. It is [i]just [/i]strong enough to bend the inch-wide piece of material - any wider and it would be hammer time. The V-jaw makes a nice sharp bend - close to the material thickness in radius.

Here is the bracket with the 3/8 BSPP parts it will support: the line in and the two feeds for the HX and boiler. 

Frame part 3 - LRFs and more plumbing

Before I go on with the plumbing, I think we need to get LRF equipped. I haven't sourced the final LRFs yet, so I looked through the LRF drawer and found some temporary ones. First, drill and tap some holes in the corner plates.

I have a little set of blocks with holes sized exactly for the tap that serve as guides to keep the holder perpendicular to the workpiece.

Not the best photo, but we are now officially Little Rubber Feet equipped. 

First up for plumbing is the sump. This plastic part is a huge improvement over the cast-metal one in both of my machines. Both of those are tiny and often silt up with grounds. Appart from being perhaps eights times the size, this one also has a large outflow.

A quick hole in the side and the expansion valve can be attached. I'm not going to bolt the sump in place until I have the rest of the lines sorted out.