This forum is now read-only
To login to the new support channel and community forums, go to the Support Portal
Water On The Floor, Smell in the Air
After making my way to the Londinium I for my morning shot arrived to find water all over the floor of the kitchen, a loud hum and the faint smell of something burning. All three lights (white, red, and blue) were lit. No alarm had sounded. I immediately pulled the plug and began some detective work.
The L1 is on a timer and was on for at least 3.5 hours at this point.
Upon examination I found the reservoir empty (contents on the floor). The metal fixture that holds the pump and rubber sling in place was extremely hot (enough to burn my finger when I carelessly touched it). No apparent breaks in tubing to the pump but could not easily examine the portion of the tubing under the tank.
After waiting for cool down I refilled the reservoir and briefly turned the L1 back on. The heating element came on but the pump failed to work.
Opening the side opposite the pump I noticed that the large cap where the wiring for the heating element enters appeared discolored and a burnt substance (cooked scale? grease?) had oozed out from underneath this cap. One of the wires close by was also discolored, the boiler mount (which looks new on one side) was discolored and there were carbonized bits on the inside bottom of the case (photos attached).
I'm guessing the pump failed in an on state or there was a leak in a line, the alarm failed to sound and the heating element did not shut off, potentially overheating the boiler for an extended period of time. Guessing this a pretty significant failure and may need a boiler and pump replacement but not sure how to diagnose further. Any help from Reiss and other members would be appreciated.
Thanks,
Tom
The L1 is on a timer and was on for at least 3.5 hours at this point.
Upon examination I found the reservoir empty (contents on the floor). The metal fixture that holds the pump and rubber sling in place was extremely hot (enough to burn my finger when I carelessly touched it). No apparent breaks in tubing to the pump but could not easily examine the portion of the tubing under the tank.
After waiting for cool down I refilled the reservoir and briefly turned the L1 back on. The heating element came on but the pump failed to work.
Opening the side opposite the pump I noticed that the large cap where the wiring for the heating element enters appeared discolored and a burnt substance (cooked scale? grease?) had oozed out from underneath this cap. One of the wires close by was also discolored, the boiler mount (which looks new on one side) was discolored and there were carbonized bits on the inside bottom of the case (photos attached).
I'm guessing the pump failed in an on state or there was a leak in a line, the alarm failed to sound and the heating element did not shut off, potentially overheating the boiler for an extended period of time. Guessing this a pretty significant failure and may need a boiler and pump replacement but not sure how to diagnose further. Any help from Reiss and other members would be appreciated.
Thanks,
Tom
Comments
Also wonder if scale could have messed with the sensors but I live in San Francisco which has soft water ( measures around 3/3.5 grains per gallon with a hardness kit) and run the oem filters.
thanks for posting - this is an interesting one
if your water is 3 - 3.5 gpg that translates as 51.3-59.9ppm (mg/L)
if you took the water box out of the machine, kept the filter in, filled the box with water, held it over the sink, and depressed the spring loaded black button in the base of the box and collected the outfall and measured the hardness after it had been through the filter that would be interesting to know the result
i don't know, but i suspect it will be below 30ppm and below this level sensor is going to become unreliable, as others have observed from time to time with their machines. my suggestion is that below 70ppm you don't and shouldnt use our water softening cartridges
there is a thermal fuse on the end of the element and the pin looks to have popped out - this is a safety device to cover situations like this, so while the red light may be coming on (an instruction from the Gicar control unit), i doubt it is heating unless you push that reset pin carefully back in on the thermal fuse
i am slightly perplexed about the water however - in a situation like this i would expect the boiler to overfill and water to be on the floor perhaps, but I'm struggling to see how it boiled dry
are you sure there isn't a leak between the water box and the pump, or that the plastic water box wasnt fully seated into the black water box (in which case it would have run out over the floor overnight before the machine turned on)
i think it would be worth filling the water box and putting it back in the machine and leaving it there for 3 or 4 hours to check that there isn't a leak between the water box and the pump - it is possible for the clear silicone tubing to get pinched and a fine hole in it as a result
I'm just struggling to reconcile overfilling with boiled dry - how long since the machine turned on when you got to it?
the pump times out after i think 1 minute of pumping, which could put at worst say 500-600mL outside the boiler if it was all but at the max fill mark when it malfunctioned (and less if the boiler right down at the minimum level) - it wouldn't pump out the full 3 litres in the water tank
kind regards
reiss.
Thanks much for helping me troubleshoot this issue.
In answer to your queries:
1) I'll test the hardness of the water post filter tomorrow. San Francisco publishes a yearly summary of its water supply (http://sfwater.org/index.aspx?page=634). Hardness as defined by CaCO3 ppm has a range of 7-77 with the average being 46.
2) While I see the thermal fuse (white nub sticking out, correct?) it is not full extended, as I can move it out more with a gentle pull. The boiler definitely begins to heat when you power the L1. I don't believe the failsafe tripped in my case.
3) The water box was fully sealed (had not been out of the machine in 2+ months) so it was not mis-seated.
4) I had refilled the reservoir a quarter full while doing initial troubleshooting - no leaks. BUT: I then read your reply and refilled the water box fully. This turned out to be sufficient pressure to load the tubing & pump and I found the vicinity of the leak. It's coming from under the pump, though I also see a slow leak from the end of the pump. Without pulling the pump (it's after midnight here it will need to wait until tomorrow ;-) I cannot tell if it's a tubing leak or the pump itself causing the main leak.
Here's a rather poor video of the leak done in a bit of haste:
https://www.flickr.com/photos/biglebowski/24796047049/in/dateposted-public/
5) The machine had been powered on about 3.5 hours by the time I got to it.
Thanks,
Tom
was the water on the floor at room temperature (i.e. as if it had run straight out of the water box) or did it feel warm as though it had once been in a hot boiler (perhaps an unfair question given the time that had transpired, but it may not have failed at start up, it may have occurred 5 minutes before you walked in)
give the thermal fuse a push - if you feel it click in it has popped, if it is firm and can't be depressed then it hasn't popped
my next question would be, do you have the 20A Sirai pressure switch, or the 30A Sirai (take a look in the parts section of the store if you need to see what it looks like - the 20A has it fairly clearly stamped on the top of the black plastic casing off the top of my head, whereas the 30A i think you have to lift the black plastic cover off and it is stamped into the metal frame
I'm still struggling to get my head around what might have happened as the logic of the Gicar control unit is as follows;
1. Power on
2. Nothing happens for about 5 seconds while it performs its internal integrity checks
3. It will check first to the water level sensor and if the boiler is not full the Gicar will activate the pump and open the solenoid
4. The upper fill level sensor should then trigger and send a message to the Gicar, which will turn off the pump and close the inlet solenoid
5. Only then will it allow the element to be turned on
6. If the water tank was empty it would alarm and refuse to turn on the pump
7. If the pump had not managed to trigger the fill level sensor within i think 1 minute it would have automatically been turned off by the Gicar to prevent it from thrashing itself to death
reiss.
The thermal fuse is firm and can't be depressed.
I have the 30A Sirai pressure switch. When I opened the cover to determine the type there is a definite discoloration around the right most terminal when viewed from the top. There is also a screw sitting on the rubber grommet at the base of the pressure stat. I'm not sure if it came from the hole directly above it or was left in there by accident during production. Photos are attached.
Tom
Guesswork on my part, but seems to be for the connection in the middle:
I re-ran the water hardness test with water sourced post filter. The current hardness is about 3.5-4 gpg.
Thanks,
Tom
have you determined where the leak is? you should be able to see if the silicone hose is leaking fairly easily, but my guess is the pump has thrashed itself silly until the internal seals have torn and the leak is from the inside of the pump
i think you need a new pressure switch, thermal fuse, pump and perhaps a new safety valve as i expect the one you have was actuated for some time after the pressure switch failed and the element then ran the boiler dry
if you can confirm that the leak is from within the pump, that would be helpful
kind regards
reiss.
As I was manipulating the pump to get a better look at the silicon hose to the reservoir, the hose / pump coupling pulled free and you can see the damage / burning in the attached photo.
i have added the 120V pump to the parts section of the store
kind regards
reiss.
Had a FaceTime session with Reiss where we walked through what likely happened and how to resolve the issue. It appears that the pressure stat failed and basically welded itself in the "on" position, sending all the available (tank, not plumbed) water supply to the boiler until it overfilled and leaving the pump running dry until it thrashed itself to death (Reiss, pls correct if this isn't a good summary).
While this is a normal failure mode over time, it happened relatively quickly in my case (approx 1.5 years of service at roughly 6 hours a day). Weird as the L1 only sends 12 amps through the 30 amp rated switch. Typical workarounds for this issue that I've seen online are to 1) switch to the second set of contacts in the Sirai when one sees carbon build up or coating damage from arcing or 2) use the pstat to switch a solid state relay in which case no arcing occurs due to the extremely low current.
Service and help here were typical Reiss, which is to say excellent, and replacement parts arrived in CA, USA 3 days after I ordered them from the UK. Very simple to work on and the pstat, pump, and thermal fuse were replaced quickly. For future searchers connector couplings from the pump to the braided hose are a 13mm spanner size and the pstat to boiler connection is a 17mm spanner.
Thanks Reiss for getting me back up and running so quickly! Much appreciated.
Tom
I've just now spoken with the manufacturer about this as i haven't been able to make sense of it
it seems as though you had a level sensor issue which resulted in a boiler over fill situation, which in turn resulted in the pump thrashing itself to death
this is why you discovered the boiler was full after you changed the pump and pressure switch and turned it back on
i have been assured that the element has not run dry as it would have burnt out in 20 seconds if that were the case
it is the view of the manufacturer that your original pressure switch is almost certainly still good as they are absolutely bomb proof (in line with my understanding), so hang onto it as a spare
why the pump didn't time out i don't understand
suggestion has been made that you take out the level probe and give it a clean as insurance against recurrence
kind regards
reiss.
Sadly the pressure switch is making its way to a landfill near San Francisco. I'll try to look at the level probe this week and report back. Any tips or tricks for removal and reinstallation? There was a link to some photos of the operation a while back but it's no longer working (404 error) and didn't turn up in search for me. Do you happen to know the spanner size for the level probe?
The full boiler was indeed a surprise upon restart so the revised failure sequence makes sense, except, as you say why didn't the pump shut off after a minute?
Looking at the original pstat there is a large amount of residue / discoloration on the leftmost contact (when viewed head on). Not sure if this is normal.
Also keep thinking back to the first set of pstat photos I posted. When I opened the pstat a screw was missing from the center unused terminal and was laying at the base in one of the grommets. The contact plate it had been holding had slipped down. Wondering now if that middle contact terminal plate managed to touch something below it, creating a short and effectively leaving the pstat triggered? Not sure if this is even possible or likely. But if so that could explain why the pump was ignoring a timeout from the Gicar?
Will examine the boiler probe and report back.
Thanks,
Tom
ps. 3 great shots this morning, very happy to be up and running again ;-)
reiss.