ITO/Leva! Controller - Q&A + Experience - Page 12

Need help with equipment usage or want to share your latest discovery?

#111: Post by Jali »

I know there are forwarding services to buy in Europe and get it sent to the US. Check out Easy Delivery, that's the one I used to get the ITO to Florida. Link below.


#112: Post by Fafou34 »

Just a short video of the ITO/Leva on a Rancilio Silvia to give an idea to the people that are considering buying it


#113: Post by sandc »

Here is another youtube video, from Kilian, with a nice view of the filter during a 1bar preinfusion, via pressure profiling:


#114: Post by Fafou34 »


I just thought that a possible upgrade for the temperature sensor positioning would be the following high temperature zip ties. I am not a huge fan of the thermal conductive epoxy. A thermal paste with a zip tie (after a good cleaning of the boiler surface with a scratching sponge + drying) could avoid poorly mixed epoxy and would give a better thermal contact. ... 46&sr=8-10


#115: Post by sandc »

> A thermal paste ... would give a better thermal contact

Not in my experience; I would not recommend that.


#116: Post by Fafou34 »

Did you try? I am a bit surprised. All the processors are clipped on the motherboard using thermal paste. It is a standard in the semiconductor industry.
Just to stress that it was just a proposition, my sensor is working fine now. The first time I glue it though it looked a bit fragile, so I added more.


#117: Post by sandc »

> Did you try?

Yes. A dia-cast boiler has a fairly rough surface, it is not necessarily flat (often round) and the TSic sensor has a depression. A nylon tie will relax over time. Thermal epoxy provides a more reliable thermal interface under such conditions.


#118: Post by sandc »

The leva! firmware for ito has gained support for the new Eureka Precisa Bluetooth scale. That scale has about the same dimensions as an Acaia Lunar but costs only about a fifth (59€ incl. 19% VAT). It is made from aluminum. It is rechargeable, protected from splash water and offers a shot timer. It is of course a simpler scale in some regards: Less polished looks (mostly down to the backlit LCD with a certain viewing angle dependency), a bit slower than my Felicita Arc and Skale 2, no auto timer and no auto tare. On the other hand, such advanced functions are not needed in scales connected to leva!. leva! provides auto-timer and auto-tare. This scale is a fairly inexpensive path to under-the-cup gravimetric dosing.


#119: Post by jaibm »

Hi all

After my ito arrived in Aus the other week, I got the chance to install it in my Silvia over the weekend. It all went super smoothly thanks to the great manuals supplied and forums such as this. Worst part was having to wait for the adhesives to dry.

Just wondering if anyone has run into problems with the reliability of the pressure sensor? I used the one supplied with the kit in the same install location, and after 1 day it seems to be broken. Up until then the values all seemed consistent and correct with not too much noise. This happened after I steamed milk for the first time. I noted that when using the steam thermostat that boiler reaches 136C, whereas the sensor is rated to 120C. Given it is reasonably close to the boiler this is my best guess as to what would have caused it. Saying that, after scrolling through forums online I hadn't seen anyone facing a similar issue.

Hoping someone might have an idea/fix before I replace the sensor.



#120: Post by sandc »

> wondering if anyone has run into problems with the reliability of the pressure sensor?

Unfortunately, you are not alone. Quite a few users have recently reported failures of Chinese-made 1/8" NPT pressure sensors. The failure mode usually is constant output of the maximum value of the sensor. Something bad has happened to the quality of these sensors. Up until the end of 2021, they were ok. The failure rate after 2 years was below 5%. Sensors sourced after late 2021 fail much too quickly, sometimes within a day. For some users, the replacement of a replacement, sourced from Amazon, has failed. I don't know if these failures are related to the temperature or something else (seal, quality of soldering, shock tolerance...). My personal suspicion is that the current global supply chain issues have lead to the manufacturer using unsuitable components.

- If you have a high tolerance for trial and error, you could buy a replacement from Amazon and keep returning it until you get a good one. I believe that sensors with a 7mm thick hexagon part were produced before those with a 10mm thick hexagon part. They could be of better quality.

- If not, I would recommend a brand name sensor, e.g. from Honeywell (80US$ - 160US$). Electronics distributors such as,, etc. have a parametric search functions where you can enter the parameters ...

... which must be met:
* 5V supply voltage
* Measuring range min. 200PSI/13,8bar, max. 500psi. If possible not more than 300psi/20,7bar.
* Ratiometric output, usually 0.5V - 4.5V
* Operating temperature range min. 120°C

... which should be met:
* Type "gage", sometimes called "gauge". In other words, a sensor which measures ambient pressure as 0bar.
* 3-pin Delphi Metripack 150 connector with the same pinout as the supplied pressure sensor

... which you would be nice to have:
* 1/8" NPT threads

A fairly inexpensive one would be Honeywell's PX3AN2BS200PSAAX, item no. 785-PX3AN2BS200PSAAX at That's a "sealed gage" sensor, not "gage", but that should work, too (I didn't test it). A sealed gage sensor probably outputs 0.0 bar to 0.15 bar at ambient pressure / at 0m NN - instead of always 0bar. This can be corrected with a -0.1bar offset in the sensor menu.

> Given it is ... close to the boiler

In a Rancilio Silvia, the recommended position at the cold water inlet (OPV) it is not too close to the boiler - as long as the sensor is rated for 120°C+. The recommended position has worked for years with various pressure sensors.