La Pavoni + PID = better temperature control? - Page 21

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

#201: Post by drgary »

RayJohns wrote:What's the current state of your Prestina any way? Do you have any recent photos of that project?
In pieces, parts collected (except bolts and such hardware), panels painted, a couple of steel panels to be brushed, parts out of acid/descaling bath and rinsed that need hand cleaning and polishing, drip tray and grate to be fabricated, lever assembly to be serviced and possibly spring replaced because it's been compressed for a year. Also valves need rebuilding. No pix at the moment, no time.

What I WOULD do for a good cup of coffee!

User avatar

#202: Post by KurtAugust »

KurtAugust wrote: And now I think:
Someone with moderate machining skills could make non invasive pid kits that you can use with several lever machines and several types of machines: pavoni, olympia, peppina etc.
Put the pid controller, ssd and heatsink in a sturdy box that you can put under the machine. You plug the machine into this box. There would be room for storing the excessive wiring of both machines. The box has a switch to use the machine as is, or with the pid taking over. All home levers have roughly the same dimensions, so one box could fit many machines. Only thing you have to design per machine is the probe: pavoni - long probe through boiler cap, different cap for cremina, something flexible for the open boilers. If I'm right it could be universal 110-220V?

Is this a stupid idea? Would it work? Who feels like doing it?
Seems like Auber already has this. ... ucts_id=48

Now I just have to machine a probe. Although I agree with Rays point that it should be 'invisible' in the end, because this is so unelegant. Still... temping...
LMWDP #325


#203: Post by Imre »

Hello guys,

would it be possible to attach the temperature sensor of the pid to the outside of the kettle? Perhaps to the bottom? Or would it react to slow to temperature changes?

Best wishes from Germany,


User avatar
RayJohns (original poster)

#204: Post by RayJohns (original poster) replying to Imre »

Hi David,

Sure, that can be done. In fact, I think that's about how some of the PID retrofit kits out there work. If you take a look at this RTD sensor at Auber: ... ucts_id=21

That's the one they use in some of their PID kits. If I'm not mistaken, it does not make contact with the water inside the boiler, but rather just threads into a screw hole on the thermostat housing of the machine. The page lists which machines it works with.

In fact, I used that same RTD sensor on my machine for a while and it worked fine. If you go back through this thread, you will see where my probe sensor failed and so I had to make a brass fitting which this smaller thread RTD sensor threaded into. I ran that configuration for quite a while (during testing) and only later upgraded to the probe type sensor, which actually extends into the boiler itself (and thus touches the water).

Frankly, I didn't see a whole lot of difference between the two methods. I think maybe there was a couple of degree shift when I changed (due to the direct reading), but I couldn't really detect a whole lot of improvement in accuracy of holding the set value. Maybe it went from .5 degrees to more like .1 degree control, but don't quote me on that (it's been a while since I was using the smaller RTD sensor).

Anyway, the bottom line is this: a PID controller which makes use of a TC or RTD sensor that is mounted on the boiler housing is not going to be as ideal, nor as accurate and/or quick to response, as one which is in direct contact with the water. However, it will still be pretty darn close. At the end of the day, having some form of PID temperature control is far and away better than having none at all. Even if you are taking your readings from the boiler housing, that's better than not having any sort of control.

As mentioned, the readings you see will be slightly delayed (based on the thermal properties of the boiler material, etc.) and they will also be off from the true temperature of the water slightly. However, once you zero in where the machine makes good espresso, then you can basically tie into that boiler housing reading and it will give the PID controller enough to go off of in order to provide quite a bit of stabilization for the brew temperature. Also, once the machine does stabilize itself, that tends to help the PID controller keep things under control better. If I had to make a guess, I would say the biggest impact of not being in direct contact with the water might just be that it takes a little longer for the PID controller to stabilize the machine around the set temperature value.

Naturally, the closer you can get to being in contact with metal that is in direct contact with water, the better.

Also, the method of attachment can make a difference. If you can thread the sensor into a hole, that would be best. If you are just basically taping it to the side of the housing, that might not be as ideal, but I would say it probably would still function (especially if you could put some heat insulation - such as a tiny bit of Silica cloth or something - over the top of it). In fact, I think if you want to just give it a try, I would say buy maybe a 1" wide strip of Silica cloth and wrap it around the boiler, then stick your TC or RTD sensor between that and the boiler. That would probably function half way decent. Again, threading into a hole near the boiler would be better. Do you have any mounting bolts or anything close by the boiler?

Like I say, something is better than nothing.

Also, what is the wattage of the heating element that the PID controller will modulate? That can have a factor here also, due to the cycle times that most PID controller employ, etc.