A few home espresso enthusiasts love their nectar in a demitasse so much that they
rarely drink it with any sort of milk. When they choose an espresso machine, temperature stability matters much more than
the ability to simultaneously brew and steam.
For such people the Isomac Zaffiro (or the similar Amica), with its nearly 800 ml.
boiler, E-61 brewhead and vapor bulb thermostat, has been among the top choices in home espresso machines. Professional and
consumer reviews alike have heaped considerable praise on these machines Zaffiro. [See link below]
But the Zaffiro has its shortcomings. All thermostats have a "deadband," the range
between where they turn on and turn off. Although the Zaffiro's has a much narrower deadband than simple "click" thermostats
do, an even smaller deadband would be better. And, while the thermostat can be adjusted, to do so you must remove the entire
steel case and fiddle with a stem that has no dial markings.
Enter the "PID." Introduced to espresso machines by then-hobbyists (now professionals),
Andy Schecter and Greg Scace, these devices offer several improvements over a thermostat. They greatly narrow the deadband.
They minimize boiler "overshoot," which happens when a regular thermostat cuts power to the heating element, but the energy
already in the element continues to add heat to the water. PID's are easily adjustable, when different coffee blends call
for hotter or cooler water, and they have digital displays that monitor temperature changes in your machine's boiler.
I am not a handy person with tools and a workbench, even if I had them (which I don't).
But I was starting to give serious consideration to doing a basic PID installation myself, attaching a thermocouple to the
outside of the boiler and hanging the PID and its relay off the side of the Zaffiro. Then I had a better idea. Espresso machine
writer, Michael Teahan, co-owns a wholesale espresso parts company (Espresso Parts Source [See link below]) not far from my
home, and I have known Michael for years. Might he be interested in taking on this project? Yes, he was. And he had more than
a few ideas of his own. Little did he know the project would take nearly a month of his spare time.
The photos on this page show the results of Michael's ideas and his work. I take
next to zero credit.
We did, however, discuss each aspect of the project before he went ahead, and I gave
him two guidelines at the outset:
1. The finished product should look as much as possible like a factory-built machine.
The Zaffiro is a beautiful object, and I did not want wires and boxes hanging out of it, if they could be avoided. 2. If any
part of the project proved to be a bad idea, we could go back to the original functionality of the machine. I use those words
instead of "reversible," because I knew "we" would be doing some cutting that clearly could not be undone.
My machine, which I now call my "Bionic Zaffiro" has had the following modifications:
1. Vapor bulb thermostat replaced with a Eurotherm 2132 PID controller. The PID was
installed through a rectangle Michael cut through the middle of the faceplate, previously occupied by the pump light and switch.
The original vapor bulb thermostat was re-assigned by Michael to steam control, but later failed. Steam is now controlled
by the original steam switch and click thermostat.
2. Solid state controller attached by a long aluminum bracket to the stainless steel
wall in front of the water reservoir to dissipate heat (a "heatsink"). Internal computer fan aimed at the PID to blow away
heat rising from the boiler directly below it. [Three years later the PID and the internal wiring are still performing and
looking like new.]
3. Thermal probe installed through a hole drilled in the top of the boiler, to reach
just above the midpoint between the E-61 thermosyphon inlet and outlet, on the opposite side of the boiler from the reservoir
inlet. The idea was to get immediate feedback from direct contact with the brewing water, but in a sweet spot where there
would be the least short-term fluctuation from incoming reservoir water. Others have shied away from this approach, prefering
an "averaging" (but delayed) reading from an external thermocouple. We still have this option, since the hole can be sealed,
Michael originally hoped to be able to just unscrew the thermal well for the vapor
bulb, insert a new fitting and drop in the probe. Unfortunately, the well turned out to be welded in. So, he drilled a new
hole and left the vapor bulb in its well to act as the steam thermostat.
4. Adding a separate pump switch that operates completely independently of the main
switch. This allows me to prime the pump or flush the boiler for descaling with the heat off.
Testing with a Scace Thermofilter (December 2007) shows nearly flat-line temperature
stability during each shot and quick recovery between shots.
February 7, 2005
July 2008 Rebuild
Well, the water flow started getting weaker, and the lever a little stiffer. So, I took the machine back
to Michael for a look-see. After five years of use (including two before the PID installation), some chrome plating had worn
off inside the E-61 brewhead and clogged the water pathway. So, Michael:
1. Removed the internal chrome plating on the main stem (which is unnecessary).
2. Replaced the small internal gaskets.
3. Replaced the brass internal lift rod and springs.
4. Lubricated the internals.
5. And just to be sure, installed a new 52w Ulka pump with brass connector.
Ol' Bio' is now quieter and smoother than ever.
[Updated March, 2005, February, 2006, May 2008 and July, 2008]
Zaffiro review on CoffeeGeek: http://coffeegeek.com/proreviews/detailed/isomaczaffiro
Link to Michael Teahan's Espresso Part Source: http://www.espressopartsource.com