DeLONGH1 ECM550.75 PRIMADONNA CLASS Repair

 

 Above is a picture of the coffee machine as it was when brand new ... not actually the example I'm repairing but let me describe its problem. It faultlessly makes a jolly good cup of coffee, but if the steam feature is used, the thing trips the mains supply. A bit annoying so I suggested that I fix it.

Getting it apart is straightforward if you have a Torx T20 bit with a security hole down the middle plus a Philips T1 screwdriver and sufficient confidence in unclipping exceedingly stiff plastic brackets.

First remove the rear cover then the two side panels to see a mish-mash of parts, wiring and plastic tubes. At the left side (looking from the front) you can see the steam producer carried under that perforated black plastic cover.

 The cause of the tripping was already pretty well established and is fairly common... being associated with a water leak at the boiler assembly used for producing steam. Sure enough there was lots of corrosion and traces of limescale or otherwise calcium carbonate at the end of the boiler tubes. Removing the boiler assembly looked complicated but turned out to be fairly straightforward. A set of very tight plastic clips hold in place the lid of the boiler assembly. There are also two pairs of push-on mains clips which need to be removed. Once I'd moved the assembly clear of its position I also found two ground wires. These have ring terminals secured to threaded parts of the mains tags (to which the pairs of clips are connected) by 7mm nuts. By carefully pulling tubes and wiring away from the boiler it can be detached.. or almost detached... because this model of machine has a temperature sensor clipped onto the boiler. This wasn't easy to detach in situ due to limescale so to free this means getting access to the controller circuit board, locating the two pin temperature sensor plug and pulling this off. It can then be jiggled free from assorted clips and tie-wraps so the boiler can be removed. The temperature sensor can then be detached (it was completely coated in limescale) 

 

 

 Once the boiler assembly has been detached the spring clip can be levered away from its mounting position allowing the temperature sensor to be pulled away from the boiler tubes. Here you can see lots of corrosion and this probably affects the operation of the sensor slightly,.. allowing the temperature to rise higher than planned due to poor adherence to the metal.

Once the sensor has been detached the boiler tubes can be separated from its plastic cover, but due to the corrosion I decided to initially unplug the far end of the sensor cable.

 The sensor cable comprises two pretty thin leads about 3 feet in length and threads its way to the control circuit board mounted on the opposite side of the coffee maker.

 

The sensor plug connection is at the bottom left corner of the control board as seen below and can be readily pulled off.

 These pictures show the clean state of the end furthest from the leak, the heavy corrosion at the leaky end and a view inside the end of one of the pair of boiler tubes which clearly shows a partial blockage from chalk build up.

At this point I guessed the boiler assembly just needed cleaning.

 

 I detached the plastic covers at the ends of the assembly and pulled out the plastic input and output tubes. One was nice and shiny but the top one showed some tarnishing and limescale which allowed steam or water to escape under pressure.

The picture was taken after treating the collar with brass brush held in a high speed drill.

 

I then carried out some electrical leakage tests which proved puzzling.

 

 First of all I measured the resistance between the mains wires and the metal body. Bearing in mind the residue present at one end of the boiler is bone dry I didn't expect to see much leakage. In fact the leakage varied quite a bit between about 7 and 14Mohm, using a standard multimeter but, if a water was present when the steam feature was turned on, the leakage would be a lot lower ( measured about 2Mohm when the terminals were damp). However as I made measurements something seemed wrong as the resistance between the metal casing and the mains wires sometimes dropped to 1Mohm so I immersed each end in white vinegar to dissolve the chalk residue. This worked fine but the leak was still present so I decided to test the thing using a high voltage power supply instead of my low voltage multimeter. See pictures below.

Much to my surprise, connecting 400 volts between the casing and one of the element mains connections drew 500mA. I reduced the supply voltage and eventually the leak reduced intermittently somewhat to about 25mA. Clearly there's a problem. I found that the pair of elements both ran hot with HT plus at one of the mains wires and HT minus at the casing so this leak is surely what was tripping the mains supply. My guess is that the tubes carrying the heating elements are not hermetically sealed and water ingress has resulted in shorts inside the tubes. Looking into the end that wasn't at the leaky plastic tube it appears that this indeed is not sealed so water probably penetrated the end at the water leak causing corrosion which bridged the elements and the metal casing.

The next morning I checked the leakage with my multimeter again and got a very puzzling result. One end shows 11Mohm to case and the other was 3Mohm to case with the element reading 44 ohms (ie. two 88 ohm elements in parallel) which I initially couldn't explain, but after a few moments I realized that what I'm seeing is diode action from limescale and sure enough reversing the multimeter leads reversed the readings.

I've ordered a new heater assembly.

 

 Summarising... The coffee machine has been run from particularly hard water during its life. This has resulted in deposits of limescale within the steam boiler assembly, especially around the plastic end caps. This has had two effects. Firstly the limescale deposits resulted in increased water pressure which strained the leakproofness of the joints. Secondly the deposits infiltrated the joints. The input and output joints each use a tapered brass bush which at the ouput end has reacted to the chalk and steam resulting in heavy tarnishing which compromised the joint seal. Steam was then allowed past the seal and settled on the area carrying the electrical connections. Additionally steam or water penetrated the end seals and ruined the insulation between the heater elements and the surrounding metalwork once the electrical leakage was high enough it was detected by mains leakage detector which tripped. During this process the temperature sensor element lost its adherence to the boiler assembly resulting in poor temperature regulation probably speeding up the failure of the boiler assembly.

 First, left is the new boiler assembly whose measurement shows perfect isolation between the mains connections and the casing.

Then, below is the old boiler assembly showing the the pair of readings which are indicative of diode action. I then applied 400 volts across the mains wiring and the case and saw the HT supply initially folding from the initial voltage but indicating a leak of 500mA with the casing running fairly warm. The current gradually fell but settled around 25mA with the HT voltage rising back to normal. A mains supply would have tripped because of potential safety ground current.

I inspected the sealing (or lack of hermetic sealing) of the mains connections of the new boiler assembly and found that it was very likely that water could penetrate into the winding space.

 

 

 

 

 Two pictures showing the fitting of the new boiler assembly. The connections from the temperature sensor were then secured in place using tie-wraps en-route to the controller circut board, then all the panels were clipped back in place and secured with their fixing screws.

 

 Switching on revealed everything seemed to be working normally although a descaling message suggested this needed to be done. In fact the message had appeared some time ago but as the descaling involved turning on the steam boiler the process had never been completed because the mains supply tripped once the boiler supply was switched on.

 Return to Reception