A Simple Washing Machine Repair and a Magical World

[For The Popcorn]

A couple nights ago, I was moving clothes from the washer to the dryer before bed. Oh crap was my immediate thought when my sock was suddenly soaking wet with cold detergent-laden water. The washer had leaked and everything nearby was soaking wet. Bedtime was considerably delayed while doing what I could to contain the mess.

Investigating, the front-loading washer was running the fill continuously and reporting a low water level. Starting from empty, the water was running a considerable time before dripping from underneath could be heard. This was in contrast to the normal water flowing out the door when the level is too high.

Water level is measured by a pressure sensor. Thinking through the two observations, one fault could account for both – a hole or crack in the tube to the pressure sensor would prevent the water level being measured and would leak when the water reached that level.

Today, I took the washer apart. It was pretty simple. Two screws to remove the top, which revealed nothing, and 4 more screws to remove the back. The problem was instantly clear. The tube from the pressure sensor had popped off its fitting on the drum! The diagnosis was essentially correct, and was easily fixed by reattaching the hose. Trivial problem to find and fix. Big mess to clean up!

I remarked In a Facebook post that I was happy for a quick, easy, zero-cost fix for the washer, but disappointed I wouldn't get to dissect the pressure sensor to see how it works. I had seen comments that it was a frequency output rather than an analog voltage or digital signal.

My friend Clive commented that I didn't need to worry. He had already taken one apart and made a video about it. What an amazing piece of engineering! Extremely low cost, with very little to fail. The technique could also be adapted to a low cost linear displacement sensor.

 

[rjenkinsgb]

Those washing machine pressure switches are impressive.

Some years back I had a machine tool customer with a fault on a massive multi-axis machine tool, that used a complex coolant system with multiple tanks & transfer pumps for filtering etc.

A level detecting pressure switch had failed; a strange custom made thing with several different level outputs, in a tiny enclosure - and about £1000 plus six weeks delivery from the machine maker.

I found a nearby appliance parts distributer and got them to go through the washing machine pressure switches they had, and managed to find one with four or five different switching points.

Also got a plastic junction box to hide it in & some plastic tube to connect things up, and got the machine going again the same day. The washing machine switch worked exactly as the £1000 custom made one, for about £45.

(That was a diaphragm type switch, around 4" diameter).

 

[upand_at_them]

I was impressed when I had mine apart some years ago. It's even simpler! No pressure switch. The tube at the bottom of the drum (top-loader) goes right to a diaphragm. A finger on the diaphragm makes contact with the water level setting wheel and shuts off the water. I think this machine is almost 40 years old!

 

[upand_at_them]

And my clothes dryer doesn't have a temp sensor for setting the heat. No, the thermostat has a built-in heater. The heat setting knob simply alters how much the thermostat heater is on and, hence, controls how much further the thermostat needs to go before triggering.

 

[Externet]

I have a modern clotheswasher used perhaps a dozen times, shiny clean pristine; parked in my basement. Developed an erratic intermittent water level failure, starting the wash cycle with a few litres of water, or overflowing to the floor, or not starting, or not stopping.
All its brains are in a single printed circuit board that includes a surface mount 15mm pressure sensor that connects the level sensing hose. Seems an atmospheric pressure sensor used in automobiles, like this without mounting holes and connector :

Production cost of the board may be a couple of dollars. Everything aimed to ultracheap parts, but when you want to buy the spare board, they want $89.9999 plus $19.99999 for shipping, plus tax plus whatever they can charge for handling; that is choosing the least pricey vendors... Am sure the complete washer costs less to build at factory.
The ripoff, total ripoff selling unreliable appliances for 6, 7+ hundred dollars built for a tenth of the selling profit.

As the failure is intermittent, have not diagnosed it, and refuse to pay the board price. So it is dormant that I may dump at some point, not kept for spares. Because I will not buy same brand nor electronic controlled. Calling a $ervice technician out of this picture. Next they will build crap that calls you on the phone to tell you the laundry is done for $199.9999 option.

My mom's washer lasted 31 years since the sixties...

 

[Nigel Goodwin]

I have a modern clotheswasher used perhaps a dozen times, shiny clean pristine; parked in my basement. Developed an erratic intermittent water level failure, starting the wash cycle with a few litres of water, or overflowing to the floor, or not starting, or not stopping.
All its brains are in a single printed circuit board that includes a surface mount 15mm pressure sensor that connects the level sensing hose. Seems an atmospheric pressure sensor used in automobiles, like this without mounting holes and connector :
View attachment 137561

Production cost of the board may be a couple of dollars. Everything aimed to ultracheap parts, but when you want to buy the spare board, they want $89.9999 plus $19.99999 for shipping, plus tax plus whatever they can charge for handling; that is choosing the least pricey vendors... Am sure the complete washer costs less to build at factory.
The ripoff, total ripoff selling unreliable appliances for 6, 7+ hundred dollars built for a tenth of the selling profit.

I'm afraid you have no idea how the consumer electronics business works, generally profit margins are extremely low, and spare parts supply and distribution is expensive, because service and spares is a VERY expensive thing to do.

As for having to buy a complete board?, that's always been the way it works - washer repairs are done more by mechanical/electrical engineers than electronics engineers - so wouldn't be capable of repairing a PCB. Same with microwave ovens, although often at least there was a circuit in the manual - even if separate parts weren't available.

Even worse, pretty well all TV's have now gone (or are going) the same way - with no circuits in the manuals, and only complete PCB's available for repair purposes (at ludicrous prices). Again, the same applies, it's EXTREMELY expensive to run a spares and service distribution service, and it's only because the new equipment is pretty well given away that makes the PCB costs look extortionate.

Check what what your mum's (we don't use mom over here ) 31 year old washer cost when she bought it - then convert that price for inflation - I bet you only paid a small fraction of that for your new washing machine.

 

[For The Popcorn]

Yes. Modern appliances have the last cent engineered out of them, to the point where lasting for 30 years is the exception rather than the rule. I was not sure that I wanted to give up a 30 year old work horse for one of these new-fangled gadgets, but the new-fangled gadget has worked reliably for 10 years now except for this very minor, easy to fix problem. Gaining access to reconnect the tubing only required removing six screws total. The benefit of the new-fangled gadget is that clothes are almost dry when they come up; the time required in the dryer has been reduced by 50% or more, and only a fraction of the water is used.

But the point of my post is the simplicity and ingenuity of the pressure sensor / water level sensor. In searching the web for a replacement, I saw it had a frequency or pulse output rather than an analog voltage output as might be expected. I was expecting a traditional pressure sensor and some complex circuit, perhaps including a microcontroller, to convert from voltage to frequency.

If you watch Clive's video (which is quite informative), you'll see how wrong I was. This sensor design is cheap (as in inexpensive, not as in junk), without any high-precision parts (no metal pressure diaphragm), and should last virtually forever (no contact points to burn or corrode). The sensor has a flexible plastic diaphragm with a ferrite slug connected to it, a coil of wire the ferrite core moves within, and a dirt-common CD4060 Ripple Counter/Oscillator driver. Given the age of most people here, I suspect you learned about this chip in The CMOS Cookbook, published 45 years ago.



The circuit is the ultimate in simplicity. Pressure on the diaphragm moves the ferrite slug in the coil. This changes the frequency of the oscillator created by the CD4060, which is divided down by the ripple counter to a 50% duty-cycle square wave of around 10 Hz that the washer's control board translates into a water level. This from one chip and a couple caps costing a few pennies. Watch the video and you'll marvel at the simplicity.

​

Speaking of the good ole daysI, here's a good example!

 

[Nigel Goodwin]

The circuit is the ultimate in simplicity. Pressure on the diaphragm moves the ferrite slug in the coil. This changes the frequency of the oscillator created by the CD4060, which is divided down by the ripple counter to a 50% duty-cycle square wave of around 10 Hz that the washer's control board translates into a water level. This from one chip and a couple caps costing a few pennies. Watch the video and you'll marvel at the simplicity.

​

Speaking of the good ole daysI, here's a good example!

For those who remember Betamax?, the tape end detection was done with a piece of metal foil at either end of the tape (VHS used transparent tape and a light beam shining through) - the tape ran past an inductor, which was part of an oscillator circuit, and when the foil crossed it the metal foil damped the coil, and stopped it oscillating. On early Betamax machines there was a preset resistor on each oscillator, which you used to adjust the feedback, until it was 'just' oscillating - too much and it wouldn't stop at the end of the tape.

Again, a simple 'old school' solution.

 

[DrDoggy2]

but there is one thing i dont understand about all this, well kind of, but i just dont get why they dont just release schematics, so that we can trace and repair that 50cent capacitor. i mean lets face it the tech just gets thrown out and replaced with new instead of buying a board so there is no proffit there, since all the replacement boards will just be used for warrenty items. and odds are they are not making any $ selling replacement items since i would say a smart consumer wont go and buy the same brand of piece of junk, just to have it blow again.

so its true that releasing schematics wont gain a proffit, but selling replacements doesnt either and it does contribute to larger land fills and more waste.

 

[Nigel Goodwin]

but there is one thing i dont understand about all this, well kind of, but i just dont get why they dont just release schematics, so that we can trace and repair that 50cent capacitor. i mean lets face it the tech just gets thrown out and replaced with new instead of buying a board so there is no proffit there, since all the replacement boards will just be used for warrenty items. and odds are they are not making any $ selling replacement items since i would say a smart consumer wont go and buy the same brand of piece of junk, just to have it blow again.

so its true that releasing schematics wont gain a proffit, but selling replacements doesnt either and it does contribute to larger land fills and more waste.

I've already explained that - the people who repair washing machines aren't capable of repairing electronic boards, so no point - and often the manufacturer will just be buying the board in from a third party anyway, so even they won't have schematics.

If you're that desperate?, then draw the circuit out, make a nice neat job of it, and post it on-line.

 

[Pommie]

Back to the original topic. Here in Australia, in rural areas especially, there are many people that use rain water tanks. Most are good at tapping the side of the tank to estimate how full it is.
This could act as a very cheap water level indicator that currently isn't available.
My question then becomes, how do you calibrate them?
Are they manufactured extremely accurately so no calibration is necessary?
Wouldn't this make them expensive again?

Mike.

 

[For The Popcorn]

I was thinking about this for an application I have – measuring water and *ahem* tanks in our RV (that's caravan for those who speak the Queen's English). The challenge would be matching the transducer range to the water level range in the vessel you are trying to measure. Modern front-loading washers might have a 4" range between empty and maximum level while a water tank might be 3 or 4 feet tall. The stiffness of the springs and/or diaphragm would need to be increased to make this sensor work.

But I don't think calibration would be a huge factor. If the sensor output is linear or varies in a regular way, take readings when the tank is empty (or before the sensor is plumbed to the tank) and when it is full. Is the tank almost empty (or full depending on WHAT you are measuring), half full, three-quarters full? I don't think a high degree of accuracy would be needed for the application.

 

[Nigel Goodwin]

I was thinking about this for an application I have – measuring water and *ahem* tanks in our RV (that's caravan for those who speak the Queen's English). The challenge would be matching the transducer range to the water level range in the vessel you are trying to measure. Modern front-loading washers might have a 4" range between empty and maximum level while a water tank might be 3 or 4 feet tall. The stiffness of the springs and/or diaphragm would need to be increased to make this sensor work.

But I don't think calibration would be a huge factor. If the sensor output is linear or varies in a regular way, take readings when the tank is empty (or before the sensor is plumbed to the tank) and when it is full. Is the tank almost empty (or full depending on WHAT you are measuring), half full, three-quarters full? I don't think a high degree of accuracy would be needed for the application.

Considering how wildly inaccurate (non-linear) car fuel gauges are, a water gauge probably needn't be any different.