La Pavoni + PID = better temperature control?

Equipment doesn't work? Troubleshooting? If you're handy, members can help.
User avatar

#1: Post by RayJohns »

On this thread, I'm going to detail the modification of my La Pavoni Europiccola (Pre-Millenium) into a PID controlled lever machine.

In learning about and modifying my La Pavoni over the last year or so, it seems that the last great hurdle to better espresso is being able to properly control the water temperature in the boiler.

Don't get me wrong... I can pull a great shot of espresso on the machine now; but I know there is more out there as far as the "sweet spot" and consistency goes. My current method involves heating up the machine on the "II" setting, then switching it down to the "I" setting. Once it levels out a bit (on the low setting), I then turn the machine off and allow it to cool down for a minute or so, while grinding the coffee and preparing the portafilter basket. In most cases, I can still catch enough pressure in the boiler to pull a shot, although typically I'm just guessing at the temperature going into the group head.

After doing a little research (as far as maybe buying a PID controlled boiler machine or a heater exchanger machine), I decided I would just modify the La Pavoni instead. With the La Pavoni, it's simple, robust and provides excellent control via the manual lever. Its main short coming - at least in my eyes - is the lack of any sort of suitable temperature control.

So to make a long story short, I've decided to follow the path a few other people have gone down and add a PID controller to my machine. This will allow me to control the boiler temperature and keep it down more around 200 degrees (or at whatever level great shots of espresso result). Right now, I don't know what temperature that is, but I do know that it seems to be below the boiling point of water.

Of course (and this has been covered in some of the other La Pavoni PID threads), this will not result in the production of steam (thus pressure) inside the boiler. One solution is to kick the machine on to the high setting for a second - to produce some pressure - and then flip it back down the low setting (where the PID controller will take over control again). This might work, but it feels a bit kludgy in some ways.

The other solution would be to adapt a small hand pump to pressurize the boiler; or perhaps even bypass the siphon tube system completely and setup some sort of small pump that just draws water from the boiler and then pumps it directly into the group head. Right now, I'm still not sure exactly which route to use, but I'm leaning a bit more to the "use a hand pump to boost the pressure in the boiler", as opposed to installing some sort of pump and bypassing the siphon tube.

Along with the PID controller, I'm going to also have to install either a thermocouple or RTD (resistance temperature detector) into the boiler base of course. This will keep tabs on the boiler water temperature and provide feedback to the PID controller. The PID controller, itself, will only be wired into the lower power circuit (i.e. the "I" switch setting), so this will still allow for steaming via the higher (i.e. "II" switch) setting.

I will probably also be installing a pressure gauge, so that I can monitor the boiler pressure (just waiting to hear back from Doug @ OE with regard to the adapter they sell).

If anyone is curious, here are some of the PID controllers I'm considering:

Image ... _Series%29 ... db531c10a2 ... r-P44.aspx

The first one is nice, from the standpoint that is has a dual display (which would show the set point temperature and also the actual current temperature in the boiler), but it's a little longer [in length] than the other two and this may cause some problems as far as fitting it into the base of the La Pavoni. I really do like the dual display however, so it may be worth the extra cost and effort. Not sure yet. The biggest issue is that, in order to mount the longer PID controller in the front of the base, it would interfere with the screw that holds on the base.

Anyway, I'll post updates as things progress. In the meantime, here's the current state of the machine:

Looks like no espresso for a while! :-)

Wish me luck.


User avatar
RayJohns (original poster)

#2: Post by RayJohns (original poster) »

BTW, I meant to add that I was able to remove the threaded boiler ring by just putting some WD-40 into the threads and tapping it a bit with a small hammer (in the counter clockwise direction). I also threaded one of the allen head screws into the hole and was able to stick an allen wrench in there and use that for leverage. If anyone is facing having to remove that ring down there, it's worth trying the WD-40 trick and seeing if it frees things up. I'm not sure if mine wasn't on that tight or what, but it came loose without a lot of problems and then I was able to unthread it little by little (until it came off).

Also, if anyone is working down in that area, the 3 allen screws hold the bottom plate (with the heating elements) into the boiler. If you remove those 3 allen head screws, the bottom of the boiler will pop off (there's an O-ring in there to seal it water tight). This allows for some nice cleaning inside the boiler.

The threaded ring holds the boiler to the base and there's a little hard gasket that takes up the extra space (it goes between the bottom of the base and the threaded ring that threads onto the bottom of the boiler itself.


da gino

#3: Post by da gino »

Good luck indeed. You also might want to read

Has anyone ever PID'd a lever machine?


Digital temperature controller on a lever machine

Dave put a PID on his lever (I think it is a Gaggia, which is essentially the same as a Pavoni). In the end he decided it wasn't worth it.

One problem is that the temperature in the boiler of a Pavoni and the temperature of the water at the coffee can be very different - ie with a fixed boiler temperature you can pull coffee that is too cold and coffee that is too hot.

That said I don't think the PID can hurt and it will be fun to do.


User avatar
RayJohns (original poster)

#4: Post by RayJohns (original poster) »

Thanks for the link Hugh.

I've read through several of the threads on here regarding the La Pavoni and PID, including the one you have referenced above. For me, at the end of the day, I just feel like having control over the temperature in the boiler is something I want. I realize that some change in temp will take place when the water enters the group head, but that's okay. In my experience with the machine thus far, having water that is too hot (and too steamy) really affects the shot in a negative way. I think being able to keep the water hot, but below the boiling point, will help improve things.

Will it be as nice as a machine with an E61 brew head? Probably not. However, I do think it will be a big improvement from having zero control over the temperature (which is the case now). Either way, it will be a fun project and an interesting learning experience. I have learned a great deal about espresso with this machine and - no matter what - I feel that will serve me well in the future (e.g. if I do upgrade to a semi-automatic machine one day).


User avatar
RayJohns (original poster)

#5: Post by RayJohns (original poster) »

Just a quick update...

I have located the temperature sensor I'm going to use. It's actually not a thermocouple, but rather an RTD sensor. Here's the link, as well as a photo: ... cts_id=193

Here is the write up from the company website:
This new RTD sensor is specially developed for measuring the temperature of espresso machine boiler. The sensor has a metric screw tip. The thread of the tip is M4x0.7 and is 6 mm long, identical to the thermostat tip used on Rancilio and Gaggia machines. This sensor can be directly mounted to the hole underneath the thermostat of Rancilio Silvia, Gaggia Classic or Gaggia Deluxe. Compared with sensor mounted on the surface of boiler, this sensor will respond to the temperature change sooner. A faster response will lead to a more stable temperature control.

The sensor is also more accurate than commonly used thermocouple sensors. When using the thermocouple sensor with an industrial PID controller, the temperature displayed on the LED screen can be one to two degree off the actual temperature during the warming up period. This is because industrial controller normally uses very simple cold junction compensation circuit as long as it does not affect the accuracy of the working range. For K type thermocouple, 0.2% accuracy of working range is about 5 F. The RTD sensor (Pt100) is more accurate than thermocouple because does not have the cold junction compensation issue.

Please note that this kit will not work for Rancilio Silvia manufactured before May 2000 and after July 2006. The manufacture date of Silvia is printed in the last row of the label on the bottom of the machine. If it says 0506, it means it is made on May 2006. ine. The newer machine does not have the threaded thermowell.

Specifications: Sensor type, Platinum RTD, 100 ohm (Pt100). Tip thread, M4x0.7. Length, 6 mm. Housing of the probe: Nickel plated brass. Cable: 45 cm long TEFLON insulated. Maximum working temperature: 350C (650F) Accuracy: Class A. +/-0.15C at 0.0 C. Alpha: 0.00385
I like that this sensor is so small. From what I can see, I'll only need to drill a small 4mm hole into the heating element base and then thread this in. I may have to capture it with a nut on the other side (and perhaps a copper washer), but I'm not sure yet. Either that or fill the threads with some JB weld perhaps.

Whatever the case, the small size will be helpful. Additionally, from the write up above, it appears this provides slightly better temperature control over a traditional J or K type thermocouple.

As far as the PID controller itself, I'm leaning towards just going with the same company (i.e. Auber Instruments). While I do like the dual display unit pictured in my first post, the length (99 mm) is just a bit too much. I really want to install the PID controller into the base if possible and 75 mm (which is the depth of the Auber one) vs. 99 mm is a big factor. The Auber controller is also less than half the cost, which is nice.

So we'll see.

As far as providing pressure to the boiler (after it's up to temp), I'm thinking about ordering a gauge adapter from Doug at OE and then installing a little hand pump. The only other option would be to tap into the bottom of the boiler and then run a line over to a pump. Then run the water back up throught he boiler and directly into the group head. This would be okay (and I even found a mini pump), but it would require a push button setup to spin up the pump. Right now, given the manual nature of the machine, I think I'm leaning towards just using a little bulb type pump that hangs off the side of the machine.

Going to order the PID and RTD sensor here in a minute.



#6: Post by JimG »

While it may not be the best solution for your machine's "geography", a controller identical to the Solo is available from They sell it as the Delta DTB series, and their version is a little cheaper.


User avatar
RayJohns (original poster)

#7: Post by RayJohns (original poster) »

More parts - this time from

316 Stainless M4 thin hex nuts (to go on inside of boiler and secure the RTD sensor) --> Part # 93935A325

Copper flat washer M4 size (to produce water tight seal around the RTD threads) --> Part # 98044A216


User avatar
RayJohns (original poster)

#8: Post by RayJohns (original poster) »

JimG wrote:While it may not be the best solution for your machine's "geography", a controller identical to the Solo is available from They sell it as the Delta DTB series, and their version is a little cheaper.

Thanks very much Jim! I see the one you are talking about here: ... category=4

Unfortunately, it suffers from the same issue as the Solo one: a little too long in the depth department.

Nevertheless, thanks for the info. I'm probably going to go with Auber, since it's a bit more compact. I've written to Auber asking if they can provide one with a green or amber LED.



#9: Post by gamgra »

I agree, the auberin PID is a good choise, many a Silvia is fitted with their PID's.
They also sell a nice little box to fit the unit.
Don't forget to order a SSR relay, you should be able to hide the SSR in the baseplate, they may get pretty warm, if you fit it thightly to the metal, the heat will dissipate quite well.

If the bottom electrical element plate material is thick enough >4mm - 3/16", you could easily drill and tap a 4mm thread in there, for the best possible seal put the copper washer on the outside, between the bottom plate and the nut.

User avatar
RayJohns (original poster)

#10: Post by RayJohns (original poster) »

Don't forget to order a SSR relay, you should be able to hide the SSR in the baseplate, they may get pretty warm, if you fit it thightly to the metal, the heat will dissipate quite well.
Thanks for the heads up on the SSR :-)

However, the PID controller has a 3A 250v relay built right into it (i.e. it can handle a circuit that draws up to about 750 watts). On this machine, since it operates on 120 volts, that should translate into roughly 6 amps of current.

On the La Pavoni, the smaller of the two heating elements (the one for the "I" switch setting) only draws about 2 amps, so I should be able to run it directly off the PID relay on the back without needing a SSR involved. The steaming element draws about 10 amps (at 120 volts), but it will be wired outside of the PID. In other words, if you turn the machine to "II" to heat it up, the power will bypass the PID altogether. Then when you flip it down to the "I" setting, the PID controller will power up and take over control of the lower watt element (thus controlling the temperature in the boiler).

As far as I can see, the only time a SSR would be called for is when you need to control a circuit that pulls more than about 750 watts of power.