<<  <  Page 2/3  >  >>
Larry M
User Rank
Another washing machine debug
Larry M   6/22/2014 8:49:13 PM
(This should really be a Sherlock Ohms submission.)

I'm still using the washing machine my late wife and I bought in 1984. It was one of the first ones with solid-state control. We selected it because we were tired of repeated failures with mechanical timers. Either the motor or contacts would fail and replacements were outrageously expensive. (GE/Hotpoint had come up with a timer with replaceable contact points by then, but the timer motor was still a weak link.)

The first problem with the washer, a Whirlpool, occured in 2007. The outer tub rusted through near the outlet and had to be replaced. After I did so, the machine failed to operate properly. It would start, fill, and then agitate continuously, never proceeding to a spin (and pump), fill, rinse, and subsequent cycles.

The electronic control was straightforward with pin-through-hole construction. It had what looked like an 8041 microcontroller with sensor inputs (door, water level sensor NO and NC contacts) and outputs which drove the machine (motor on/off, two solenoids on the transmission) through simple diac/triac circuits. All the components and circuits were clearly silk-screened on the board and Whirlpool had helpfully provided an inserted pamphlet with a block diagram.

I first discovered that I could turn the washer off, and then start it in the spin cycle and it would properly empty the machine of water. I also discovered that if I let the machine fill and agitate beyond the expected time and then removed the pressure hose from the water level sensor, that the machine would continue on to spin, the next cycle portion.

Consultation with the experts on the forums gave only the following unhelpful information:
  • It could be either the water level sensor ($50 or so) or the controller board ($350)
  • Both components were no longer available

so I knew I would have to figure it out myself.

Now the problem with debugging with an electronic timer rather than a manual one is that the only way to advance the timer is to wait for it--there's no knob to rotate. So after some head-scratching I decided to follow the awful and hard-to-read timing chart on the back of the washing machine.

I chose to trace the Gentle cycle because the wash cycle was only eight minutes; the other two were twelve and fourteen minutes. It was fortuitous that I did so.

Crouching behind the washing machine studying the timing chart and watching the works as it filled, then the water level sensor caused the microcontroller to start the agitator as the motor ran at the slower, gentle speed. After eight minutes the motor suddenly advanced to the faster speed as agitation continued. This told me that the microcontroller was working. After thinking about it for a bit, I disconnected the water level sensor hose and the drum began to spin as water was pumped out. But the timing chart said the water was to be pumped out and then spin would begin.

Time for more thinking. The transmission has two solenoids. That means it has three or four states. The identified states are:
  1. Agitate (no pump, no spin)
  2. Pump
  3. Spin and pump

The state was never changing from Agitate to Pump. Watching the solenoids as I went through the sequences of Off-->Spin and the long Off-->fill-->Agitate-->Pump let me identify that the Agitate solenoid wasn't releasing at the end of the wash portion of the cycle. I poked at it and determined that it was still energized. When turned off, it moved freely so it wasn't mechanically stuck. And now the problem became very simple.

Early in my career I had worked for a few years in reliability physics and one of the components I had evaluated was SCRs. Triacs are similar to SCRs. One of their well-known failure modes is latch-up. When an SCR or Triac suffered this failure it would remain off when biased (as it should), and turn on when triggered (as it should), but it would not turn off at the next zero-crossing when the trigger was removed (as it was supposed to).

And this was exactly what was happening. The triac turned on for the agitation and wouldn't turn off to shift the transmission so the pump would run. The microcontroller was waiting (forever) for the non-running pump to drain the tub. If I removed the sensor hose, the controller would start the pump running and spin while there was still water in the tub, instead of pumping without spinning until the tub was empty, then spinning. If I turned the machine completely off and then started it directly in the spin cycle, the Agitate triac was never triggered so the pump worked fine.

Of course the original triac part number was nowhere to be had, but I was able to order a triac with the same case and pinout from Mouser Electronics. They had units rated at 110v for 35 cents, 220v for 36 cents, and 440v for 37 cents. I thought the higher breakdown voltage might reflect a more reliable part so I splurged on the expensive one. 

It took just a few minutes to unsolder the defective triac and install the new one and the washer has worked fine ever since. I did let the forum experts know that the washer had been fixed for 37 cents, but never received a response. I guess they were more expert at replacing modules than working at the component level.

User Rank
Re: A great story, thanks.
Frogman842   6/18/2014 3:02:18 PM
My Fisher and Paykel washer, dryer and dishwasher have excellent troubleshooting guides and modes. 

You can exercise all the valves, lights and motors, test the switches and read the sensors. All from the front panel display.

My wife's washer had a problem that was easy to diagnose - AND REPAIR. The offending sump pump had a decent connector for the motor and the assembly twisted out by hand with a coarse threaded fitting. And the part was not expensive or hard to find.

Well done. They have me as a customer for life.


User Rank
Re: Wrong problem
JimT@Future-Product-Innovations   6/18/2014 12:01:33 PM
Bob, that's funny. "disconnect the power cord, replace the device and reconnect cord".

User Rank
Re: Wrong problem
JimT@Future-Product-Innovations   6/18/2014 11:56:07 AM
I don't have any problems with the manuals getting thinner; less paper wasted in the first place.  Hate to sound arrogant, but most manuals are not written by the design engineers, and the design-intent is never well-communicated.  The tech-publications team is (most times) not an engineering staff and they don't convey the clearest messages. (insert tab-A into slot-B. Gimme a break!) That being said, I repair all my own appliances, and found this web site to be the best work-partner you can have.


Repair-Clinic.com, out of Canton MI, offers parts (with photos) and instructional videos for every brand of every type of appliance, and offers full returns of all parts (even electrical) up to 365 days after purchase,  no questions asked.  I have used them again and again (Washer, dryer, Stove, Microwave, DW, disposal) and cannot praise them enough! (I sound like a salesman.  But they are worthy.)

User Rank
Re: A great story, thanks.
Technophile   6/18/2014 2:30:51 AM
@ab3a, kudos for putting in test modes!  I'd guess that test modes pay off immediately during product development.

For appliances, it seems to me that companies either respond to the market or lose market share.  Contractors building a furnished house:  I'm guesing most buy on price.  Many (most?) consumers buy appliances based primarily on price (some using e.g. Consumer Reports ratings), so corporations have to keep the price down if they want to sell appliances.  Other consumers buy based on bells & whistles or plan to move every few years; they aren't interested in keeping their washer running for 30 years.  If companies spend money making appliances more maintainable, that's an up-front expense.  How do you justify that to management?  How does management justify that to shareholders?

This is in contrast to production lines, where I'd think engineers are usually picking machinery to last, and are (hopefully) well aware that avoiding down-time is worth spending more. 

Perhaps some consumer education, or a "Repairability" rating.  The more consumers buy based on repairability, the more companies will focus on it.  Could be sold as part of a "green" rating.

When a company is building 100,000+ of something, every penny gets shaved.  $0.01 x 100,000 = $1,000.  $1.00 x 100,000 = $100,000.  It's easy for management to see the immediate, up-front benefit of shaving pennies.   Harder to convince them to spend on maintainability unless it clearly and obviously will sell many more appliances.

User Rank
Re: A great story, thanks.
ab3a   6/17/2014 8:35:15 PM

I couldn't agree more.  We have test modes for the systems we design. If a float switch, level indicator, or flow sensor goes bad, our staff have the technology to figure this out.

But for whatever reason, they don't do this for household appliances. I guess they want people to be frustrated when the repairman can't fix it right away so that they  go out and buy a new one instead of fix the existing equipment.

This is how waste happens.

User Rank
A great story, thanks.
jtcronin   6/17/2014 3:45:35 PM
Modern washers and dryers are a great example of decades of optimization.  They have a number of sensors that often serve several roles.  How they can affect the operation of the appliance is often not obvious.  My compliments to your diagnostic skills and persistence.

Many years ago when I designed machines and used PLC's to control them we had very similar problems.  A faulty sensor could trip an important safety interlock.  Finding which sensor could be tedious.  

One day inspiration struck.  We were at our favorite watering hole debating a design on some cocktail napkins.  We looked over and saw a repairman working on the establishments pinball machine.  Inside that pinball machine was a single board computer that worked very much like our PLC's.  It had a diagnostic mode.  In that mode you could check and test ever sensor and actuator.  Stepping through the diagnostics the technician quickly found a microswitch behind a bumper was mis-reading.  Its bracket was bent.  With a little bit of finger pressure he was able to reposition it.  With a few drops of glue he was able to secure it better.  

In about 15 minutes that repairman was able to diagnose and fix a machine as complicated as ours.  It took us hours to do the same thing, and we needed a mountain of drawings too.  

Why don't we put in a diagnostic mode program in our PLC systems?  

We did and the folks in the factory loved us for it.  They thought we were genius miracle workers.  All we did was to borrow an idea from a pinball machine.

What troubles me is you don't see something like this in household appliances.  None of my appliances with electronic controls has any sort of diagnostic mode.  It would be trivial to include one.  

I'd bet that Whirlpool washer has a mechanical timer and control system.  Mine does.  The new one I just purchased has a mechanical timer too.  Given the low cost of microcontrollers, why does anyone still use the mechanical models?  

User Rank
Re: Wrong problem
OLD_CURMUDGEON   6/17/2014 3:38:53 PM
Seems to me Cabe that you've hit the nail on the head, BUT not just with your "smart" thermostat.  I think this trend started a while ago as application software on PCs got to be more sophisticated, all-encompassing, and all-powerful.  The producers decided it was far easier to pack HELP files full of "stuff" and to invent the ubiquitous, but despised acronym, FAQ!!!!  No matter how complete these HELP files think they are, most often what you need cannot be isolated easily, OR it was not there to begin with.  And, FAQs are just as despicable.  A person has to wade through a literal mountain of questions to find something close to the answer being searched for.  Personally, I get a charge from some MICROSOFT websites... down at the bottom is the question, "Was this information helpful?"  How many times I wanted to throw my shoe at the PC screen, I can't begin to count!!!!!

I HATE WINDOWS and EVERYTHING It stands for!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

Mr. Fixer
User Rank
Re: Wrong problem
Mr. Fixer   6/4/2014 8:06:34 PM
Oh it was agitating, agitating the owners!  

These days, many of the components of an appliance have markings that indicate what material they are made from, so they can be recycled.  To me, the ultimate in recycling is repairing the appliance and continuing to use it.  That is getting to be more and more difficult with every product cycle, for the very reasons we all read about in "Made by Monkeys".

Cabe Atwell
User Rank
Re: Wrong problem
Cabe Atwell   6/3/2014 7:51:21 PM
I agree, the manual I got for my new programmable thermostat is one page long. I guess when it comes to the devices programming features I should just 'know' how to do it.

<<  <  Page 2/3  >  >>

Partner Zone
Latest Analysis
Dow Chemical and several other companies have launched a program in Omaha, Neb. to divert about 36 tons of plastics from landfills in its first phase, and convert it into energy used for cement production.
Both traditional automation companies and startups are developing technologies to improve processes on the factory floor, while smart sensors and other IoT-related technologies are improving how products are handled during transport and across the supply chain.
Remote monitoring now looks beyond the individual motor or conveyor and instead views the process from end to end.
BASF is partnering with Hyundai on a high-performance concept car that showcases both eco-friendly, sustainable materials and materials to enable lighter weight cars.
In an move to ease the use of multiple product design tools, PLM provider Aras and EDM company Zuken have integrated their products.
Quick Poll
The Continuing Education Center offers engineers an entirely new way to get the education they need to formulate next-generation solutions.
Oct 10 - 14, Embedded System Design Techniques™: Getting Started Developing Professional Embedded Software
SEMESTERS: 1  |  2  |  3  |  4  |  5  |  6 |  7 | 8 | 9 | 10

Focus on Fundamentals consists of 45-minute on-line classes that cover a host of technologies. You learn without leaving the comfort of your desk. All classes are taught by subject-matter experts and all are archived. So if you can't attend live, attend at your convenience.
Next Course September 27-29:
Sponsored by 3M
Learn More   |   Login   |   Archived Classes
Twitter Feed
Design News Twitter Feed
Like Us on Facebook

Technology Marketplace

Copyright © 2016 UBM Canon, A UBM company, All rights reserved. Privacy Policy | Terms of Service