In this very long post, I describe my process of repairing a kitchen appliance.
I explain in excessive narrative detail (a) to provide an amateur how-to guide for someone else who might have the exact same issue with this exact same model and (b) to generally encourage people: try and fix your broken stuff! It can sometimes actually be done, and you really don't need to be an expert or even particularly skilled! To be real for a second: if in a few decades we want to be living in a world where we can have things like kitchen hot water dispensers ordered from the internet, we're going to have to get comfortable fixing them now because as I've said previously, extractive growth capitalism in 2022's current form is simply incompatible with that future. But we can probably still have nice things if we make them to last, make them repairable (ideally user-repairable), and make them sustainably.
I'm not an expert, so probably did this repair in an unnecessarily difficult and painful way. But the important thing is, that doesn't matter! My intended audience for this is someone who maybe has never attempted to repair anything electrical before, beyond perhaps changing a fuse.
First, a disclaimer: As we all know, electricity is dangerous, and electricity mixed with water is double-dangerous. I'm a cautious person and am confident I did this in a safe way. But please, don't touch metal parts inside an appliance even when its unplugged (capacitors hold charge when not plugged in), and don't plug something into mains electricity if you're not happy with it.
Ok, on with the post.
Instead of a kettle in our kitchen, we have a hot water dispenser. Specifically a Morphy Richards 131004 Redefine Hot Water Dispenser. Instead of filling a kettle with a tap, boiling it, and pouring out, you use a dial to select a ml quantity of water, a temperature, and press a button.
I highly recommend replacing a broken kettle with a hot water dispenser. We got it because it's more energy efficient than a kettle: if you want to make a small cup of tea — 220ml — you only need to boil 220ml of water. You can boil as little as 150ml (enough to warm a baby's bottle) without reboiling the excess amount in the kettle over and over again. Also the heating chamber is sealed, so heats faster, quieter and presumably more efficiently than a kettle. Turns out it's way more convenient too: one button and your drink is made or your bottle is warm when you walk back into the room.
However I don't necessarily recommend this Morphy Richards one in particular. I got it second-hand on eBay, about £60, which is already more than a new kettle. New, they're £170. It came highly reviewed, and indeed it's super easy to use, higher capacity and more featured than many others. But after about 3 months of owning it (and, based on its condition, I'd say maybe 1 year of life), the dial on the front broke in such a way that one could only select the maximum possible quantity of water: 1.5l. Everything else about it still worked, but the dial did nothing.
This didn't render it useless, as you can still finely control how much is dispensed using a button, but it totally negated any energy savings: reboiling 1.5 litres for every 220ml cup of tea is a lot more wasteful than even a bad kettle.
I emailed Morphy Richards to ask if they would repair it out of guarantee, and they won't. They offered a discounted new one, but that doesn't help if it also breaks after a year. Even if I could afford to do that (I can't), it also again somewhat negates the energy savings if I need an entire new appliance manufactured for me every year. I looked for another used one on eBay, to discover if you look at the "spares and repairs" ones there, they all have exactly the same "1.5l issue".
I went to Amazon to look for "users also bought" recommendations (with the intention of doing what I usually do after looking for things on Amazon: buying it used elsewhere). While I was there I glanced at the reviews, and something caught my eye.
Someone had photographed a circuit board in their review, including what looks to my untrained eye like a rotary variable resistor — the kind of thing a dial might use.
I found the actual review, written by an Amazon user named Michel harrington, hereafter "Saint Michel".
Back in March an annoying problem started, kept randomly dispensing 1.5L of water into our mug where we normally leave it on 340ml! hot water everywhere.
wasnt going to mess around during covid trying to get it fixed/replaced.
cracked it open..(fun there too!) to find there is no mechanical relief for the start/stop button, despite every plug/connector being dobbed with glue gun to stop them moving/coming undone, they haven't put anything on the little preset which selects to dispense quantity which the start/stop boil button activation rod travels through.
so when you hit the start/stop boil button it distorts the circuit board breaking the tracks to the preset.
quick bridge with a bit of wire and glue gun around the preset to stop it moving ...sorted..bit of a bad design their MR.
Looking closer at the photos, we can see that Saint Michel has shown exactly the repair they made.
To me there were two particularly noteworthy parts of this review. First:
quick bridge with a bit of wire and glue gun around the preset to stop it moving ...sorted
That kind of sounds like something I could maybe actually... do. Assuming I have the same issue, or at least very similar (which seems highly plausible, given what Saint Michel describes), that sounds like something I could maybe attempt. But second:
cracked it open..(fun there too!)
…that's a little ominous.
Anyway, I tried it, and to cut to the punchline, it worked. Here's how I did it.
Here's how my appliance looks from the front, with the water reservoir tank and drip tray removed.
Even after I emptied the tank, as I carried the appliance over to my kitchen table, turning it over to look at all sides, a lot more water dripped out of the two nozzles at the front, the water intake at the back (white ring you can see here), and a little hole in the back-left of the "roof" of the container, which I think is a pressure-release valve from the heating tank. Make sure that you get as much water out as you can before opening it up. As we all know, water and electricity should be kept far apart, and we don't know if there are charged capacitors inside here, and we don't want water getting somewhere dry when we eventually plug it in again. I had lots of paper towel on hand to clean up the drips.
At first, recalling the "fun there too", I was expecting to have to pry something open. But turning it around to look at the back, we can see six recessed screw holes.
Five of these are standard cross-head screws. (Top two, bottom two, right-hand middle one.)
But one of them (the left-hand middle one) is like a miniature version of one of those cursed screws they use to fix arm rests on public benches to stop homeless people sleeping on them.
Luckily, I have a multi-bit screwdriver set from back when I replaced a broken spring on my AA battery recharger.
With the screws out, we can see that there are a total of 3 different kinds of screws holding the back on, which is completely unnecessary.
I jammed a screwdriver into the gap between the back and the side and twisted to try and get the back to pop off, or see if there were plastic clips holding it on — there aren't. But it still wouldn't come off easily at all. Maybe there are more screws somewhere? A closer look around the bottom reveals some hiding.
These ones are of the small self-tapping variety. When I remove both of them, a little strip of black plastic falls out — remember to keep it!
Now the back pops off partially, enough at least for me to see that the sides are just plastic panels which slide into groves in the case, but are otherwise not secured. They slide out with a bit of joggling. Now we can see what's inside.
At this point I'll mention something which breaks the continuity: there are actually some more secret screws which continue to tether the back in place, but I don't discover them until later. That's why the back is still pretty much in place for the next few photos. The whole case is plastic and flexes a bit, so I was thinking I was going to have to attempt the whole repair with the case under flexing pressure to make enough room. In fact it's not quite as bad as that. Anyway, back to the repair.
Now let's see what's just behind the broken dial.
We can see the edge of something with cables coming out.
Ok, yes, there's a circuit board back there. That must be the one revealed to me by Saint Michel. And indeed , though it's not very clear in that photo, the plugs are glued into place using what looks like hot glue. Normally I'd be very grumpy about something being glued into place, but in this case I think St. Michel is right, and it's there to prevent the vibration of the pump or the pressure of button pushes from working the plugs free. I didn't photo this, but some delicate work with a utility knife and needle-nose pliers let me peel the glue away without damaging anything.
But the board is still jammed right up next to the white plastic heating chamber, and seems to be surrounded on all sides by a black plastic shroud.
More inspection reveals that the plastic heating chamber is secured to the case with two metal straps, which are screwed into the top of the case. There's one on each side. Perhaps if I remove them it will give me more room to flex the case and get at the board? Worth a try.
It kind-of works, though I can also see that the heating tank is still attached to the front case through an inflexible pipe join: this is where the freshly-boiled water flows out of the nozzles on the front, presumably through an electronically controlled valve. Pictures of that later.
With a bit more room to manoeuvre I can see that the black plastic surround of the offending circuit board is secured by four screws.
I can probably get all those screws out, but there's no way I'll be able to get the plastic cover off afterwards.
At this point I wondered if maybe I had to remove the metal face panel, so began bending back metal clips holding it in place. Turns out you don't need to do that. But it explains the scratches on the case you might spot later.
There's got to be a way to make more room in here. The entire back of the case is still attached. I can see that it's not a single piece of plastic all the way around, but that there's a complex-looking join at the front of the base.
I gingerly try prying around with a screwdriver to see if it's clipped together, but it's not. I can actually slide a piece of paper into every part of that to see that those jig-sawed bits don't clip at all, they just abut. So what's keeping it in place?
Wait a minute.
Yep, there are extra screws under the glued-in rubber feet. Pry them out, take out some more self-tapping screws (matching the other ones from the base), and we can see how that complicated join works.
Now the entire back case comes off, merely tethered by the power cord. There's a lot more room to see what's going on inside with the water routing, and now I can see that the inflexible pipe join between the heating tank and the dispensing valve is actually just pressure-fit, and will loosen with a bit of shimmying.
Those ribs on the valve assembly feel rubbery, and don't feel greased, so I hope and trust that if I push it back together, it'll remain water-tight.
With the tank partially pulled away, there's now room to pry back the unscrewed black plastic surround of the board, and slide it to the left (facing from the back of the appliance).
And here's the case around the circuit board, with cut-outs for the plugs:
Now we can see the circuit board laid bare.
In that photo you can also see that I've unplugged the cables from the right-hand side. I don't know what these kind of plugs are called, but I've encountered mini versions of them before when repairing or upgrading my laptop. The trick with them is they have a tiny lip you can get a screwdriver or spudger under and carefully lever them out. Just be sure not to pull them by the wires, tempting though it is, as you're likely to sever the wires from the plugs. (I don't actually own a spudger, though I keep meaning to get one. A thin flat-head screwdriver is fine on something reasonably robust like this.)
There is also a red-and-white ribbon cable plugged into the bottom, and a black ribbon cable plugged into the top.
These ribbon cables have a similar, though slightly more delicate, plug. There are lots of tiny pins in the socket soldered to the board. I've found a good way to deal with these is to get one end out a little, then the other end a bit more, and kind of gently rock them back and forth until they come free. Don't use force.
It's worth getting the board unplugged first, before you remove it, because you don't know how easy that will be, and you don't know how much movement these cables can deal with. This is more of an issue in laptops etc where the ribbons are really fine and sometimes printed on brittle plastic film, here they look more robust.
The bottom ribbon plug was also hot-glued in place, but the glue peels off in a satisfying way.
I'll mention at this point that I don't own a hot-glue gun. Because this glue is probably in here for mechanical resilience more than tamper-proofing, I should really be replacing it when I'm done. Saint Michel mentioned using hot glue to bolster the troublesome part, which also sounds like a good idea, and is also something I didn't do. Again, something I need to get because it's super useful for this kind of thing, and will make the repairs last longer.
Ok, the board appears secured in place by three screws (yet another new kind of screw — at least it's a cross-head). Get those out, but the board still doesn't move.
Oh, you've got to be kidding me.
So this screw is really annoying for me, trying to repair this, but it's actually sensibly placed: if the board is really taking all the force of the button pushes, you don't want it to flex, so tethering it at more points, especially in the middle, will help avoid that and help prevent just the kind of fault I'm repairing.
Anyway, my first thought is that I can use this handy bendy screwdriver which came in my set.
And although I can just about feed it in there, the screw is too tight and this bendy friend doesn't have the torque. Luckily past me accidentally left a tiny multitool in my soldering kit, and it does the job.
Ok, with that screw removed, the board is finally free.
And we can see all the screws we needed to remove to get here.
Looking at the beige side of the board we can see some recognisable bits.
The offending component looks to my amateur eye like a kind of rotary variable resistor. St. Michel called it a "preset", which I'm going to guess means it's a variable resistor with discrete preset values. I'm going to call it a variable resistor, because that's what I thought. These are simple questions with simple answers, but I don't know what they are and I haven't checked. It's ok to be wrong or imprecise when it doesn't matter.
Looking closer at the offending component, we can see that sure enough, the appliance's start button is mounted behind the dial.
I had some fun twiddling the dial around with my screwdriver and pushing the button. But as I discovered later, this is a bad idea: at the moment you remove the board, the variable resistor is aligned to the current physical position of the dial, and there's not nearly enough room for me to be able to see that when putting it back in place. Much better to leave it where it is, and know that if I don't touch the dial, it'll line up back where it started. Or do what I did, and refer back to the photo of the component you took to return it to the correct orientation.
By the way, as I removed the board, I heard the sound nobody wants to hear when repairing something: a tiny piece of something falling out of somewhere unseen and bouncing around into a convoluted cavity. Luckily, it was just the plastic buttons on the face of the appliance, which are held in place just by resting against the board: when I removed the board, one fell out. I retrieved it and kept them in place with a bit of tape.
Anyway, now we see the dial — which remember is the only bit which apparently doesn't work (assuming the problem is located with the component itself, and not some transistor or controller chip elsewhere) — we can look on the other side of the board to where it's connected.
Let's look closer at where the variable resister is mounted.
Looking closely at the component itself (not photoed from a good angle), it has three pins going through the board. (So it's probably not a simple variable resistor, I think they have two pins. Never mind.) Anyway, these pins come through the board and are soldered onto contacts on the reverse side, as in the photo above. They're easy to see because the pins are much taller and pointier than the surrounding components.
So either the component itself is broken, or some part of the circuitry is broken. In the words of Saint Michel:
so when you hit the start/stop boil button it distorts the circuit board breaking the tracks to the preset
I don't quite know what this means, but I'm thinking the thing most likely thing to break is a solder joint, rather than the actual printed wires on the board. At any rate, if it's as simple as a broken connection somewhere, I can hunt down where that is using my trusty multimeter which I got when I tried to repair my ancient iPod HiFi. (And failed, since the whole thing is glued together; in the end it fixed itself.)
Using the probes, I can now go from each pin of the variable resister and trace the parts of the circuit it should be connected to, to verify that part of the circuit is not broken. First I'll just check up to the first solder joints in each circuit. If that all looks good, I can think further afield.
Here's the parts of the circuit each of those pins are connected directly to with just wires:
So I use the multimeter to test continuity between all the endpoints of each of those.
Yellow checks out, cyan checks out… orange doesn't. This is new information! And looking at St. Michel's photo, this is the same pin he seems to have made repairs to, though I can't fully make out what he's done.
Now we just have to figure out exactly where the discontinuity is.
The wires printed on the board are sealed in some insulating material, so I can't literally trace along with my multimeter probes to find the break. So I can only use the terminals. Using the multimeter I can see that B, C and D are all connected to each other, and none of them connect to A. So the break is somewhere between A and where the wire from A joins the circuit. I suspect it's just the solder joint on A. If you look closely to the other two pins of the variable resister, there's a little silver dot next to each one. These are presumably there precisely to check the continuity of the solder joints. But there isn't one for the solder join I'm interested in, so I can't test directly.
Looking REALLY closely at the join, I think I can see that yes, the solder attached to the pin has separated from the board contact itself.
So at this point I could just heat up that solder join to attempt to reflow it into the contact. But this is obviously a weak point of the appliance, so there's nothing to stop it cracking again again, and my soldering job will probably be less robust than the current one. So I'm going instead to try and recreate the actions of Saint Michel and bridge this pin to where it's supposed to connect to using wires. (Actually, I tried to reflow it too, but my soldering iron has a hard time melting the solder on that join. Perhaps it's a different kind of industrial solder with a higher melting point? I don't want to roast the surrounding components by holding my iron on there for ages. Eventually I kind of reflow it a little bit.)
Problem: what do I use for wire? I don't have any wire just hanging around. I actually first tried to use some twisty-ties, but the metal was stiff and I don't want the joins to be under elastic tension. I checked the box in which I keep electronic waste, waiting to be taken to the local electronics recycling point. There I found an old mouse into which some AAA batteries had catastrophically leaked. It's fully broken already, and not made in a repairable way, so I ripped it apart to hunt for wires.
Bingo! Two wires connecting the battery compartment to the board. I snipped those out and clipped and stripped them to the right lengths. Now to do the bridging.
I'm not a very good solderer; I've not had much practice. Soldering onto a board like this always feels like you need three hands: one to hold the thing you're trying to solder, one to hold the iron and one to feed the solder wire. This board is probably the fiddliest thing I've ever had to solder, the contacts are really small and (worse) close together for my ancient, blunt iron.
The best thing I found to do here was to attach two wires together first (that's easy because you can twist them before soldering, then hold them in a crocodile clip). Then get solder on the join of the two wires (easier than just one end because there's more room for the solder to be "soaked up" into the twist). Then I can attach the wires to the faulty pin by re-melting the solder already on the twisted wire, so I only need two hands.
Once that's done, I can feed an extra blob of solder onto the join to sure it up.
Now that the wire is fixed to the board, it's easier to do the same thing with the other ends: get a tiny bit of solder on the twisted end, use that to pin it to the desired contact, then add more solder to sure it up. The really important thing is: don't touch any component with the iron's hot tip, and don't let solder spill over a gap.
One wire connects the main pin (A) to the right-hand terminal of the component (resistor?) marked R1 on the board (labelled D in my diagram above).
The other wire connects the pin to the right-hand terminal of component R5 (B above).
While I'm doing the soldering, I keep checking the continuity between the pin and the terminals I'm soldering to. A handy tip here is to attach one multimeter crocodile clip to the pin on the other side of the board; then you only need to mess around with one of the probes to check for continuity.
I'm also checking continuity between both terminals of the components I'm soldering to, to make sure I haven't accidentally bridged a gap. These resistors are really tiny! So both sides of R1 were not detected as continuous when I started soldering, and they weren't continuous when I finished. Probably good!
At this point I have good continuity between the pin and the three other points it should be wired to. I could add a third wire from A to C but I have found soldering such tiny parts to be difficult, and it's also getting late, so I decide to quit while I'm ahead. I think I'm… done?
A sensible thing to do here would be to hot-glue the wires to the board so they can't move around. Adding some dots of hot glue around to brace against the black plastic shroud and prevent the board flexing might also help, but I don't own a hot-glue gun, so I just tape it.
Time to reassemble. Board back on:
I put the board back in before I re-attached the plugs, because I was mentally locked-in to reversing my steps. But that was really fiddly to do, and as I pushed in the top ribbon cable I felt a "crunch" which felt an awful lot like lots of tiny metal pins getting bent. I mention this for two reasons: first, having seen the robustness of the cables and glue-reinforced sockets, I should have thought to put the plugs in before the screws; second, it turns out that some plugs just make a crunch when you plug them in. The whole time I was doing the rest of the reassembly I was convinced I'd just destroyed it, but actually no!
As I was trying to get the plastic shroud back on the board the heating tank completely detached from the valve assembly. That made it really easy to get the shroud back on. And the tank squeezed back on without apparent issue. I should have been more assertive during disassembly! I did have to tape some paper towel to a screwdriver to mop up water before it trickled down to the motor, though.
Heating tank straps re-attached. The screws partially bite into the plastic tank holes, so it's hard to tell when they're actually going into the front case. They should screw fairly tight.
Plastic join at the base of the case reassembled, side panels slotted in. Top realigned, back screws in (well five of them, I'll see that security screw in hell).
If these are the arrangement of the holes in the back of the appliance…
…then these are the screws that go in them...
Finally the little plastic panel under the bottom.
And now, a moment of anticipation as I walk it back over to its home on the kitchen counter and re-fit the water tank and drip tray…
IT WORKS! I'm only annoyed I didn't take a video of the original fault for comparison.
Time for my reward, and about 24 hours of insufferable smugness.