No big deal right?
http://www.amazon.com/LG-WM3997HWA-White-Electric-Washer/dp/B00LIAP2ZY
But wait a minute. Hold up. It’s only 110v. Some reviews say it takes 6 HOURS to dry a load! That won’t do. I can do better.
….
To understand the final outcome, it’s important to know that I’ve been fascinated with washing machines for a long time:
(movie of me building a washer as a kid)
Yes. This was recorded on VHS tape. That’s how long ago it was. The composite video digitizer we have only works with windows XP so yes this is a video of a video. This was a nice little weekend project. Basically I could wash a months worth of clothes in one go with this. A teenager’s dream. Draining the 40 gallon garbage can used for this was…. difficult. The clothes line also fell over due to the very wet clothes I was trying to hang on it (no spin cycle), and each piece of clothing had to be hand-wrung in the end.
So. Then we had college. Laundromat was three or four blocks away. Unacceptable hardship I tell you! I did what any clever college student who was trying to save money would do: examine the true nature of the problem. What about a washer cleans clothes? is it the soap? the agitation? the spin cycle? The rinse? Only one way to find out.
Try washing socks in all the various combinations, compare the results. Turns out that at least at an empirical level, most of the cleaning action – killing bacteria that make bad smells – is done by the soap and the agitation is really only there to get the dirt. So. minimize dirty-ness and you can get rid of the agitator we saw in the earlier video. Yes. All you have then is a bucket, a hose, and some laundry soap.
And the solution to the clothes line is to just hang the clothes on their final hangars, positioning the washing bucket beneath them so any drips end up back into the bucket. Directing a 1500 watt space heater at the clothes will expedite drying but just like the people with the above mentioned washer, it can take 6+ hours to dry them.
For at least a year in college the bucket-washer was reasonably effective, and spared me the 10 minute trudge through the snow plow berms along route 9 with a laundry tub. However, the water usage to actually get the clothes clean was…. rather high.
The next apartments had a washer/dryer included. Lesson learned.
But, the time came for me to finally purchase my very own washer. And I wanted the best. So I inspected craigslist on a fairly regular basis, until one day a working whirlpool duet front load direct drive washer with stainless steel drum showed up. $100 and a short truck ride later I had my washer. I figured – add a blower, a heating element, a couple relays, maybe worst case a VFD, and I’d be all set to build a washer dryer combo, and no way would I be using a little 110V heating element.
The actual version is alot more messy. This project went so far off the rails that at times I wasn’t sure if I would ever finish it. I knew I was in trouble when I needed a hole for the hot air to enter through the front door. The glass front door. It went downhill from there.
Final numbers:
$100 washer
$100 electronics
$100 unused servo amps & burned up power supply.
$25 furnace blower fan
$100 misc.
Oh yeah – and – waaaay too much time. So much I don’t even want to admit how much. *sigh* Suffice to say that buying a brand new combo machine would have been significantly cheaper if I assign any value to my time at all. Or buying a separate washer dryer and stacking them. Or pretty much ANYTHING else.
How-to technical infos:
3 phase BLDC motor driver basics
Discrete transistor based servo amp:
http://elabz.com/bldc-motor-with-arduino-circuit-and-software/
Seriously, wish I’d found these guys before I soldered up my own servo amp. I ordered one from ebay if/when mine breaks.
http://www.a-m-c.com/products/drives.html
Adding Back EMF detection (without halls) and feedback/detecting stall (awesome article)
Sensorless BLDC Control with Back-EMF Filtering Using a Majority Function
Arudino AC Current transformer (monitor blower and power supplies for shorts/other bad stuff)
https://openenergymonitor.org/emon/buildingblocks/ct-sensors-interface
Final electrical schematic:
*ahem*
How about a cut and paste from the code? Basically 3,5,6,9,10, and 11 drive P or N FET’s that switch the 48V power supply respectively.
pinMode(2, OUTPUT); //Connect our servo amp via relay board. (AND disconnects washer servo amp!)
pinMode(3, OUTPUT); //phase A, pull up
pinMode(4, INPUT_PULLUP); //run-mode timer switch. Note to self: this is active low!
//may want door switch status? Whatever. Won't break anything if you run with door open, will probably just not dry. so no biggie.
pinMode(5, OUTPUT); //Phase A pull down
pinMode(6, OUTPUT); //Phase B pull up
pinMode(8, OUTPUT); //blower + 48V power supplies (two relays)
pinMode(9, OUTPUT); //Phase B pull down
pinMode(10, OUTPUT); //Phase C pull up
pinMode(11, OUTPUT); //phase C pull down
pinMode(12, OUTPUT); //240V heating element contactor.
//A0 => Potentiometer (unused)
//A1 => Input heat temp
//A2 => Output heat temp
//A3 => Humidity sensor
//A4 => Current sensor
Not sure if anybody cares to see the Arduino sketch that runs this. There are a couple of different things I played with getting this to work: software 20khz pwm, using arduino pwm to approximate sine waves. lengthening pulse timing based on back emf signal seemed to make the most difference for top speed and reliability. So of course I never bothered to implement that for the final version as it doesn’t seem to need it.
Other references: