I've posted some stuff about my Z11 LE project in the Z9 thread but I thought I would start a new thread as this may be of interest to others using commercial levers at home.

I had made great strides recently, to the point where I was producing coffee with the machine but tonight I ran in to electrical problems, burning out 2 pressurestats and making the garage go dark. I am taking the machine to a guy who actually knows what he's doing to get this sorted, but it happens that this guy knows his onions when it comes to installing PIDs. He's not trying to sell me this idea, I'm just weighing up my options.

So given that a traditional pressurised lever machine controls its brew temperature by dissipating heat through the group and therefore brew temp depends not only on boiler temp but also workload and, to some extent, ambient temperature, would there be any advantage in fitting a PID to control the boiler temperature?

As my machine will be used at home in my garage, largely for my own entertainment until such a time that my ambitions bear fruit and it gets pressed into commercial service, I hope that it may allow me to adjust the temperature to warm the groups more quickly and fine tune according to ambient temperature. As an example, last night I was making coffee like a kid in a sweetshop as I'd just got it going but the air temp was hovering just above freezing and I could not get enough heat into the group to get anything but sour shots and a water temp boost for flushing could have been just the ticket.

Is there any validity to my pontificating or would this be a largely pointless exercise?

Others have PID'd lever espresso machines, so why not? Digital temperature controller on a lever machine and Has anyone ever PID'd a lever machine? discuss the whys and hows. On the other hand, a pressurestat reacts more quickly to temperature (pressure) changes and placement is irrelevant versus a thermocouple that can measure only one point.

Cheers for the links Dan, really should remember to UTFS. Hadn't considered the advantage of pressurestat vs thermocouple, but then a pressurestat is not really practical to adjust on the fly. I shall give this much thought I'm sure.

PID hasn't got diaphragms and contacts to wear out. And it is silent.

frustrated_uk wrote:Hadn't considered the advantage of pressurestat vs thermocouple, but then a pressurestat is not really practical to adjust on the fly.

You could use a pressure transducer and have the best of both worlds, though it could be cost prohibitive. See PID with pressure sensor instead of thermocouple and Controlling steam boiler with PID and pressure transducer? for previous discussions.

I have been thinking about this idea all night as I worked on a frame off restoration of an Olympia Caffarex....not as much fun as a lever but interesting as well. Anyway, you have shown that the big group works both ways, depending on ambient temperature it can get too cold. But aside from burning up pstats it seems that the issue is convenience of adjustment as much as anything since an increase in the pstat setting would serve as well as a pid, if only it was simple to adjust it on the fly. Very few machines have a readily adjustable pstat with the case on the machine....now why is that?
I can't speak to the theory of putting a pid on your Z but I put a makeshift pid in a commercial lever to see what would happen and it nearly drove me crazy with the continual on and off of the element as it worked to maintain that exact temperature setting. Every 5 seconds or so the element would turn on and then turn off right away .....it was a pretty high wattage for our kitchen circuit and with each turn on the lights would dim and then brighten up when the element turned off. It eventually maintained a very even temperature of the boiler but it does seem that it would take a lot of experimentation with a pid on the boiler to get the group temp right.
I prefer the pstat, and a warm room.

You can also use a PID as, or install, an electronic thermostat. Set a narrow band and it mimics a pstat.

Sensor placement is important. It's not a bad idea to sketch out the boiler and heating element location with the expected thermal eddy pattern.

Shadowfax, you are right about the wiring. The effect was not that bad in our workshop which has a 20 amp circuit.
I agree also, the element does not care what is telling it to turn on and off. The machine I was experimenting with has no pressure gauge so I had overlooked the idea that you would be setting temp and monitoring both the pressure and the temperature. Does seem like a nice advantage. With no pstat a lot of the dialing in of the temp behavior when pulling a shot has been worked out with group flush times etc, so I have gotten more or less used to the idea of working these things out without a lot of sensors on the machine in question.
I still don't see the reason for burying the settable pstat so deeply in a machine that it is inconvenient to change the setting.

Hi,
having had a very comfortable PID-ed double boiler machine in the past, yet I wouldn't want a PID on my commercial lever machine.

Why? From my point of view a 100% accurate boiler water temperature is not that important as it isn't responsible alone - it is just the starting point of the extraction process.

The extraction temperature varies throughout the shot - at first, the boiler water is still much too high, then the temperature drops during the preinfusion of, in my case, about 18 seconds, while the grounds and the metal absorb heat from the water.
Then there is the surrounding temperature in the kitchen (winter, summer...).

Of course, the grind, the amount of coffee meal, the shape/size of the basket, the tamp (soaking speed), all influence the heat absorption during the extraction.

Then there is the decreasing brewing pressure, after a full start, also influencing the flavour of the shot.

The temperature of the water in the boiler to me is only one parameter of many influences and thus I wouldn't compare the effect of using PID-ed boiler water has in a double boiler machine with the situation in a commercial (i.e. big boiler/massive group, thick tubes..) lever machine.

I've tested so many pstat settings throughout the last months, that my husband drilled a hole into the cup tray to make the pstat easily accessible and now there's a screwdriver sticking permanently in the pstat screw, so I can adjust it anytime

Of course, I still don't have a temperature display. But having measured the brewing water temperature at different pstat settings, I can roughly estimate the boiler water temperature from the boiler pressure indicated on the pressure gauge.

When changing the boiler pressure/water temperature often, I've learned that the boiler water temperature isn't all.

When I lowered the pstat setting as much as possible - and steaming milk over a longer time period had already become difficult - I could measure a low brewing temperature (with a K thermocouple, under the basket, in which lay a one time used coffee puck) but still a shot can taste burnt (not totally, but a little ashy or even having a burnt undertone), depending on the roasted beans used.

Apparently, the boiling water temperature in the beginning of the preinfusion, when the water streams into the chamber, can be too hot for a coffee blend.

Surprisingly, I currently use a blend to be brewed quite cool and despite a higher pstat setting (to be able to better steam milk) it doesn't have a burnt touch in the overall taste (I usually make a double ristretto out of 14,5g of coffee meal in a 14g double basket).

The pstat reacts quickly and is already very accurate. Depending on the sensitivity set, it could take some time until a PID has finally approached the desired boiler water temperature. Now I'm often pulling one shot after another. Then again, the group heating up furthermore influences the brewing temperature of the extraction...

And I don't change the setting often, anymore - it's now set as low as possible to be still able to steam milk well enough.

To sum it up - I don't miss a PID and I don't see a genuine advantage.

It was very useful in my former double boiler machine as it influenced the extraction temperature clearly, controlling a single boiler with cooler, non steaming water (as there was a second steam boiler operating - via pstat, by the way - at a much higher temperature), bringing it to the ideal brewing temperature so no cooling flush is needed as in HX machines.

On the other hand, a commercial lever is a single boiler machine operating with constantly boiling, steaming water in the kettle.

Apart from warping the striking simplicity of an almost non-electronic machine a bit - in my opinion controlling the boiler water via PID is only responsible to some extent for a good shot. Still the used blend must correspond to the machine.

I've retried many blends I knew from my former non-lever machines - and I think trying is inevitable - the lever shot is so different from a pump driven machine, it's like a concentrate, emphasizing the body, acidity and all flavours (also the less pleasant ones).

And when I've found a blend (or single origin roast, needless to say) that goes with my pstat setting (which cannot be lowered anymore in favour of a sufficient steam capacity), I'm all happy.
Only in rare cases I've stumbled over a roast that could be brewed at an even higher temperature.
So I don't alter the pstat setting often anymore, the testing period being over. If I wanted to, I could simply turn the screwdriver handle sticking out of the machine behind the cups. No temp. display, of course. But do I really need to know the current boiling water temperature that is still much too high?

I learned that the commercial lever machine types vary quite a bit. I can only speak on behalf of my Rancilio.
And I'm just a simple user

Cheers
Tanja

Good to hear from someone who REALLY knows their machine, on an intuitive level.

We use our Faema Velox wall machine constantly and get the best shots from it. It has only a thermostat with a wide deadband. The group must be cooled as the thermostat setting drifts upward over time but the temperature is determined by feeling the group during the cooling flush. Very satisfying to have no gauges or temperature readouts but just to proceed with the right feeling and timing, dose and grind. It is very low tech and very satisfying.

Well done, Tanja!

Using both a PIDed Silvia and a Pstat commercial lever I cannot see any advantage in PIDing a 5l boiler machine. While the replacement of Silvias thermostatic control with its wide deadband really helped a lot to get repeatable shots, the thermal mass of both the boiler water and the grouphead and the thermostat's narrow deadband result in very stable temperatures - furthermore my cup tray is made of perforated sheet, so there is always a hole for the screwdriver to change temperature within 1°C

The thing I like most about the lever apart from the unique taste of it's coffee is the KIS (you cannot say you keep it small) which doesn't support a PID inside the machine.

Manual filling, manual pulling ...

Cheers Emil

@Doug
I understand what you mean