Auto-fill for boiler on M20 Cimbali Junior

Equipment doesn't work? Troubleshooting? If you're handy, members can help.
Posts: 55
Joined: 1 year ago

#1: Post by Shawn- »

This is a continuation from my restoration project of a 40+ year old La Cimbali Junior
I restored a 40 year old Cimbali Junior

In that thread I described how I took a pour-over machine and connected to water mains.

I mentioned there that I am not a fan of manual boiler fill because of the concern of protecting the heating element. Additionally I would not consistently have enough water above the end of the tea-water dip tube to be able to pull a cup of water before getting nothing but steam.

I started with this...

My original thought was that it would be a cheaper alternative to a Gicar controller, but I did have concerns about it short-cycling when the water reaches the bottom of the probe (these fears were realized)
The data sheet is here... ... 3_csm6.pdf

The normal mode of operation utilizes 3 probes.
E3 connection is to boiler ground
E2 is the low level probe
E1 is the high level probe

If the water level drops below E2 then fill starts and will continue until the water level gets to E1

This configuration would not work in a boiler with a single probe, but with a jumper wire between E2 and E1 it can be made to operate with a single probe, albeit with zero hysteresis.
If the surface of the water is turbulent (boiling) then it would frequently break and make contact with the probe and cause the relay to cycle (on-off-on-off-..,)

I tested it first with just the level relay.
Installed in the 'brain box'

For my connection to boiler ground E3 I made a modified copper sealing washer that I could attach a spade connector (green wire). E1/E2 goes to the level probe (blue wire).

Power was fed to the fill solenoid but not the vibe pump.
The connection went to the switched contact on the manual fill button switch (red wire). I left the original circuitry in place so I can still manually top up the boiler ( could be useful for descaling).

I checked fill function and found that it did work, but as expected it was short cycling the relay. Also my probe depth was too low because I was only getting steam from my tea-water valve.

I removed the probe and trimmed about 12mm off the end and tested again. This gave me about 120ml of water before steam, but from this point I can adjust water level by different means.

I added a 3sec to 10sec adjustable delay-on-break (DOB) relay to the fill solenoid's circuit.


Water level drops below the probe.
Fill relay will cut power to the Pressurestat turning off the SSR for the element, and initiate fill by turning on the fill solenoid.

Water rises until it touches the probe.
Relay switches to provide power to Pressurestat and heat circuit is live. Signal to DOB relay stops (break).

After 'break' the DOB relay continues to provide power to fill solenoid for the time duration set on it's potentiometer.

I have this set so that I have about 275ml of water available from the tea-water valve before it has to re-fill

Now I can be confident that the water level will always have the heating element fully immersed, and if a circumstance occurs the the boiler cannot fill, then the heat circuit will turn off.

Cost $?
Well... with the addition of the DOB relay it did bring the cost up, but it worked out to be about Can$35 cheaper than the best sale price I could find for a basic Gicar with shipping ... JpEALw_wcB

Was it worth the extra effort?
For me it was. I learned a lot about my machine. I enjoyed the process and the project. That's enough for me. Whether it makes sense for someone else depends on their sensibilities.