Eliane and Silvia PID Modifications: notes and photos

[russell]

These notes will not cover details covered in descriptions of PIDing linked in this website. (In the case of the Eliane the notes are on virtually identical Isomac Zaffiros or Amicas.)

I am still testing these machines and the photos show the machines with the wiring on the PIDs uncovered, although you can see the split black tubing that I use to cover them in some photos. I have not aimed at aesthetics but functionality. Some conversions, such as Marshall Fuss's Amica PID done by Michael Teahan, show how nice the final result can be. But this requires a difficult to cut rectangular hole in the casing to accomodate the PID controller inside the machine. I have left them outside, and have not bothered to put them in "project boxes" from Radio Shack as others have. It's my personal taste, but I find the glowing numbers and blinking lights off-putting. If I wanted something beautiful I'd have a La Pavoni or Micro Casa a Leva but I don't have the patience to learn to use one of these to produce consistent espresso. Also I wanted a bigger PF and basket so I could make espresso for two in one shot. However, the machine stays in the kitchen where I don't worry much about looks. Also the machines are pretty dirty from testing. It's a problem that the shiny chrome of the Eliane, in particular, shows every water spot.

The first photo shows the control switches and lights on the Eliane after modification. Note the second switch from the left, between the green and red light. It's from Radio Shack, and permits turning off the heater by interrupting one of the DC inputs to the SSR from the PID. This was Michael Teahan's idea and Marshall has such a switch on his machine. Dan, in his PIDing of an Amica (I think it was an Amica) recommended turning off the heater during brew and for a time afterwards, but a recent posting here offered the view that that was pointless. I'm not sure. I do not it is very disconcerting to see the needled on the presure gauge bouncing up and down about 0.5 to 1 bar during brew, and I think those who are so worried about vibratory pumps versus rotary pumps would be discouraged also. One big problem is remembering to turn it back on after the shot; if you forgot you've got a long climb back for steaming. The other use of the switch, as recommended by Michael, is during cleaning using the blind filter. He recommended running the machine without heat (I believe, but ask him to make sure). Also turning off the PID control probably helps cool down the machine quicker after steaming, but the best way to speed this up is to run water through the steam wand.

The next photo shows the front on the machine with the 1/32 Din Fuji controller from TTI Global (about $129) mounted in a bracket on the front left. The wiring is uncovered at this point. All leads have quick disconnects so the controller may be removed in a few seconds. The bracket is held by one of the existing mounting screws.

There's not much more to say about the PIDing. I used 16 gauge wire throughout which is more than enough. The thermocouple is a washer type J just like recommended on Murph's page on the Silvia PID. I first mounted it under a screw on the top of the boiler (like the Silvia PID) but moved it after getting some instability which someone suggested was due to rising steam. Taking his advice, I moved it down the side of the boiler under a hose clamp with Silicone heat t transfer paste between the washer and the boiler. The green insultation covers the outside of the boiler. This is shown in the photo posted on the pressure mod, but I'll repeat it here:



Note the SSR, also per Murph's specs, from TTI Global. It is secured to the inside side by stainless screws and nuts, with the side serving as a heat sink. Again, I used Silicone heat transfer compound.

The wires are led through the space at the left-rear bottom corner. I ending up filing out the space a little so the wires would be less constricted, although this is not strictly necessary.

Unlike Marshall's, machine, I did not use the existing thermostat to control the steam temperature, but left the original (cheap) control in place. I think this permits a higher temp, but I'm not sure.

The next post will be about the Silvia PID.

Russell

[russell]

The Silvia PID:

This was much easier for three reasons: (1) it's the second one I did (2) the inside of the Silvia is a lot more accessible with just the top to be removed and (3) I did not bother to install a switch to disconnect the PID from the heater. The principal difficulty was tightening the nuts on the SSR relay attached to the inner wall. The relay was mounted similarly to the one in the Eliane, but there was less space to get at the nuts inside. The mounting screws and nuts served to hold the two brackets holding the PID, and you'll note in the first photo that I notched these out to avoid having to remove the inner nuts and replace them.

The posting yesterday on the pressure modication shows some of the insides. I first mounted the (type J) washer probe on the thermocouple under one of the screws on the boiler top as Murph did, but after noting some instability and overshoot, I relocated it under a hose clamp down on the boiler similar to the mounting on the Eliane. I think this is a more temperature-stable sampling point, but I don't have much objective evidence that that's so. I'm not getting much overshoot now, however.
Here's a photo of the mounting:

My guess, and some others concur, is that the probe location probably is not critical. The temperature displayed may differ but there is no functionally important difference that would affect the final brew. Reason? Because you have to do taste and visual tests to determine the optimal PID temp reading. It's always going to be a lot higher than the brew temp, and it doesn't matter if one location means the PID is optimally set at 228F and another location is 218F as long as there is consistency. In fact, the optimal temperature on the Eliane so far in testing seems to be 228F and on the Silvia 223F (both with the same quantity, grind, and packing of Black Cat). I think Murph was a bit optimistic (or is that putting it too mildly?) when he said just set the SV (target value) at 230F and you're done. The settting will vary with the probe position, coffee type, grind, amount, packing, your taste, straight shot or latte, etc., etc. I have found trying to identify the idea temp the hardest part of the process. Moreover, you have to be conistent about machine warm-up times (to control the group temperature), how long the PF is out of the machine to load the coffee, and probably other factors I don't even recognize. We're trying to get better coffee; for most of us, perfection is beyond our capability.

Note that the PID controller is a much larger Fuji PYX4 which sells for considerably more from TTI Global but which I bought on eBay for $35. This was another good idea from Michael Teahan (he bought his PID and the one for Marshall on eBay). Even for eBay the price was low, perhaps because it was listed as "temperature controller" rather than "PID." I have a third PID I bought on eBay for $40 but it is not as good because it doesn't have fuzzy logic, and it overshoots more. If you're going for an external mounting you probably don't need a 1/32 din size and you'll find lots of choices on eBay. If anyone needs the parameter settings I ended up with, let me know. I spend many hours trying to get this to work before I succeeded--it's very different in its settings from the 1/32 din unit.

Here's a final photo of the Silvia, with the top off, the wires to the PID uncovered (note the black cover tube to the right):

I'll discuss rationale for PIDing and test results in future postings.

Russell

[miKe mcKoffee]

FYI FWIW, if you know your boiler to actual shot temperature differential you can set your PID to display start of shot temp rather than boiler temp. On the Fuji PXR3 I'm using it's by adjusting PVOF setting. For instance, if a 227°f boiler temp reading yields measured 200°f at the puck start of shot, a 27° difference, set PVOF to -27. Then PV will display boiler temp minus 27. You will also need to set SV to what is now shot temp, 200 in this example.

BTW, have you done a float-valve water auto fill mod on Missy? If you haven't and decide to keep her it's a must do IMO, and only about $30 for parts, less if you use plastic water lines. I've only done it on Missy, but any machine with a decent size and shape reservoir should be a candidate I would think.