Miele Softtronic W451-E repair

[axero]

I have an old washing machine with a broken control module (EDPL126-B). The service technician found it not worth fixing. A new control module would cost as much as a whole new machine. So I figured perhaps it is possible to repair the module myself.

The module has been replaced once before during warranty and it was the same fault that occurred; the drain pump is stuck at "on"-state and I cannot get into any of the washing programs. I don't have the exact clue as to what is causing this error but the technician hinted at water condensation and mentioned something about "short-circuit".

So removed the module from the machine and took it apart. In the enclosed picture it seems that the plastic support that keeps the PCB in place inside the closure of the module has rubbed off some of that solder mask over that trace. The crevice between the support and the trace has gathered condensed moisture which in turn has caused the short circuit. This is what I believe has happened, perhaps this may have caused other failures as well. Perhaps someone here knows more about this issue. I think this must be a very common issue with this generation of Miele washing machines.

It is a good machine and it wasn't cheap when it was new, it would be a shame to throw it away because of this issue.

[ebastler]

Yes, those two traces have obviously suffered some damage. Hard to say from the photos whether it is mainly corrosion, or a short triggered by humidity. Or it could be burnt traces caused by some component malfunctioning elsewhere in the circuit. (Maybe the plastic tab which apparently covered the traces in that area has prevented convective cooling, causing the traces to overheat just there.)

So, have you fixed it? Cutting the offending traces and adding two patch wires between the nearest vias, should be the simplest solution. (Although those vias are not all that near...) When the machine works reliably after that fix, you can be sure that the problem did not start with some component malfunction elsewhere.

[axero]

I have not yet done anything because I was advised to check for other failures before dealing with this issue. When conducting measurements, I have conduction on both traces and no measurable short circuit. At least not now that it is dry.

[axero]

Cleaned out the corroded traces, before and after shots are supplied below. Then I tested the module on the machine and it seems to work fine now.

I think the cause of rubbed of solder mask goes beyond simple vibrations. That switch above which is a part of a button that is pushed frequently probably added to the stress exerted on the PCB that broke the solder mask.

I will appreciate suggestions on how to seal those cleaned traces. Should I use epoxy or something else?

[Ysjoelfir]

That looks like there is hardly any trace left of the bottom, thicker one, at some places. are you sure the machine works COMPLETELY? Not that the pump now doesn't work at all anymore while it was stuck on before? I would definitely grind away all of this corroded mess of traces and reroute new ones beginning at the vias.

[ebastler]

That looks like there is hardly any trace left of the bottom, thicker one, at some places. are you sure the machine works COMPLETELY? Not that the pump now doesn't work at all anymore while it was stuck on before? I would definitely grind away all of this corroded mess of traces and reroute new ones beginning at the vias.

Yes, I would also stand by my earlier recommendation to use patch wires instead. The "nearest" vias are almost at opposite sides of the board, but two long wires secured to the PCB with hot glue should work nicely. I wouldn't trust the traces as they look now, and they might easily corrode further despite any efforts to seal them.

[axero]

I have located the vias for both traces one of which is a connector pin to the small daughterboard with the 8-segment LED-display. I don't know if the drain pump doesn't run at all, I successfully managed to start a washing program and I couldn't see any erroneous behaviour.

When conducting measurements over both traces, there seems to be sufficient conduction on both of them, i.e. they don't appear to be severed but I will follow your advice thank you for that. How about using Dow Corning 3145 as a fastener instead of hot glue?

[ebastler]

I think there is a pair of vias a bit nearer by, very close to the bottom of the PCB and near its left side. (Referring to the orientation in your second image in the original post.)

Other glue should work as well to secure the wires; it's not critical. Hot glue is my preferred choice just for convenience, because it sets quickly. It's neither nice-looking nor particularly strong. 

[Brutte]

Do not use a permanent coating because you won't be able to fix that if anything goes wrong.
As for the glue - use somethng sticky but do not use epoxy.
Of course try to avoid soldering to traces - use SMD pads or vias instead (for mechanical stability).
Do not put wires over damaged traces as it makes no sense to prettyfy it. Just align them for conveniency and serviceability.
As for the corroded part - cut the traces at clear location so that nothing gets shorted.

[axero]

Thanks a lot for your help, I have patched the traces with wires and removed the broken part of the traces and sealed with Dow Corning 3145 which reminds me of silicone. I also patched another trace on the component side of the PCB just in case, it wasn't as rotten but it looked worse for wear.

When putting the repaired module back into the machine, it turns out that it is still malfunctioning but now I get a "check inlet" error. So what happens when starting a program is that the drum fills up normally, then when the water fill makes a pause and turns to pour water through the washing powder dispenser which it does fine without problems, the "check inlet" light starts blinking and the machine cancels the program.

So I took the entire machine apart and found the problem as shown in the picture below. So there is some kind of a sensor that uses a chamber that is located at the front of the drum beneath the opening. It seems like it uses a pressure sensor to gauge the water level inside the drum. This chamber is not connected directly to the pressure sensor on the controller board. Instead it is connected trough this canister as shown in the picture. What this canister does is a mystery to me. Inside is a rubber membrane with a small hole in the middle that is surrounded by some kind of a weight. There are no electrical connections to this canister and no special active components. Perhaps it serves to dampen the air-flow as the drum is filled or drained. Whichever is the case, the hose connector at the bottom of this plastic canister has snapped off severing the connection between the chamber and the pressure sensor. I think I can forgive Miele for the terrible PCB design if they have fixed these issues on ensuing generations of washing machines, but this?!?! This is really embarassing, I cannot wrap my head as to their poor choice of low quality plastic components.

[Ysjoelfir]

if it realy is only a passive membrane, I would try to connect both hoses together. Or drill out the hole and try to tap a G1/8" thread into it or glue a hose fitting directly to that mystery box. It doesn't seem to be that critical.

[djacobow]

I think it must transfer the pressure signal from one hose to the other, while keeping them moisture-separate.

I broke a similar plastic pressure sensor part (mine did have wires) while trying to repair my LG machine, and it ruined a couple of days for me, since this was on the bottom unit of a stack and replacing that part could not be done without unstacking and removing the entire outer cabinet -- something very difficult in my tiny kitchen.

[Brutte]

Your washing machine is an electronic one so in case the water level sensing fails (because of various reasons) the algorithm can close the inlet valve and report an error. 

In a old-type washing machine, if water level sensing fails, the water level will eventually overfill the drum and flood the appartment. This pipe and level sensing is a critical component.

If you want to tweak with level sensing - make sure your concept always fails safe.

[axero]

Thank you a lot for your suggestions! The machine is working fine now. It was difficult to find a suitable replacement part since everything is closed on the weekend but I finally did.

So what I did was using a hose connector for a hose with 6 mm inner diameter using 1/2" threads (pics of my, ... I mean Ysjoelfir's solution are supplied below). So I took the canister apart and drilled a hole from the inside starting with a small drill and successively using larger drills until about 10 mm (my toolset was limited). Then I took a threading tap and made 1/2" threads so that I could screw the hose mount right in. Then I used the Dow Corning stuff both on the inside and the outside to ensure that no air will leak past the threads. I think the canister repair will last longer than if I opted for a new one. The broken canister was not my doing. I was nowhere near the area when I was troubleshooting the control module. I didn't even disassemble the machine.

The reassembly was simple except for two things; I forgot to properly align the orange manual door opener latch when putting the front panel back again. But I could get it right after a while. The clamp ring for the door seal was even worse to put back. But after watching a generic YouTube clip on how to remount the clamp ring, I had an idea on what to do and successfully put it back. Another clip that was a good inspiration was this one:



although it is not the same model, the disassembly was quite similar and gave me an idea on what to do. What really struck me was how simple this machine really is. Essentially, you have the following components:

1. The washing drum with a motor, heater, and shock absorbers. This is the actual machine. The rest are merely ... accessories. This is the heart of the machine. Nothing else matters.
2. Electric inlet valve that lets water in
3. Drain pump
4. Door opener
5. Water level sensor comprising a small chamber mounted at the front of the drum underneath the door. It basically monitors the water level via a pressure sensor on the control module
6. Some relays and switches
7. Washing powder dispenser with a control unit that controls the direction of the flow of water into the different slots of the unit.
8. The main control module

That's it. So I think what triggered the inlet error was that as it opened the inlet valve, the unit was expecting a change in pressure of the drum. Since it cannot determine whether that is because the water flow is inadequate or because the water level sensor is faulty, it just stops the program after failing to register a temporal pressure differential in the level sensor. It is not a very advanced algorithm. On the contrary, it is very simple, but it works and does as intended. There is no need to worry about flooding of water. The troubleshooting procedure just gets a bit trickier when the sensor fails.

As for quality of the machine, the drum assembly looks like it can last for a very long time. The outer chamber is made of cast iron which is what Miele brags about. After 15 years of operation, I would expect a lot more rust on the thing when considering that it is made of cast iron which is not stainless. From the looks of it, it looks like it is only a few months old still. The weak points are the plastic components; the water level chamber and the washing powder dispenser. And the other weak spot is the main control module. When this fails beyond repair, perhaps it is time to switch to an Arduino

[Ysjoelfir]

 Nice fix  I really like those old Miele machines for their simplicity and ease of repair. In my apprenticeship I had to repair some of those (but usually older models) and I absolutely loved it for the simplicity of maintenance. Not for carying, though. Those Machines are build for decades. Which, and that comes to no surprise I guess, is also noticeable in the substantial weight. 

Actually, I just got myself a used Novotronic W985 2 weeks ago, one of the finest machines back around 1999. This one got barely used as it stood in a vacation home from some rich banker guy here in Frankfurt. So I guess that this one hasn't had more than some hundret hours of useage. Transported it back to my apartment and lifted that 99 kilo monster up to the third floor - works a charm since then, no problems what so ever.

[Brutte]

As for quality of the machine, the drum assembly looks like it can last for a very long time. The outer chamber is made of cast iron which is what Miele brags about.

From my experience with (cheaper) washing machines which have plastic tub + concrete slab, what typically fails is the drum bearings. These parts are not designed to last infinitely and looks like this is the first predictable point of failure of washing machines. The main trick to extend operating life is to identify this failure and act before other components are destroyed.

So I needed two new bearings and a radial seal. $10 with shipping, 2h job.
This is only possible with older >10yo washing machines where you can disassemble the tub - the new ones have tubs sealed/welded and smaller less durable bearings. Then the replacement is much more complicated and expensive.

Progress.