Those nasty little pushbuttons
LoopyByteloose
Posts: 12,537
It is July in Taiwan with lots of heat and humidity. Life here is heavily dependent upon a good air conditioner and yet mine has begun to show signs of old age.
For a year or two now, I have been ignoring the fact that the ON/OFF button didn't work as I have had the remote control to depend on. But the last week or so, the crud the fan has been blowing out finally made me decide to open it up and clean the interior.
After cleaning, the machine got unpredictable. Though I am not sure if I just had been ignoring the fact that it wouldn't cool under 27 degrees Centigrade as I have been running it 24/7 at that setting since the beginning of this year's hot season.
But now, I can't even be sure that it will turn on and off the output fan and when the fan is off, the cooling fins quickly ice up and the compressor runs heavy.
After much ado and poking at it for 3 days, I finally decided to take apart the LED and button console and check the push buttons. After all, I knew the ON/OFF didn't work.
The results are interesting -- all 7 of these buttons are defective in one way or another. And they did not fail open or closed, they provided various degrees of resistance. It is no small wonder the micro-controller on the machine is behaving oddly and the remote control could not respond properly.
So I replace all 7 with ones I keep on hand for reset buttons. Total cost of less that $1USD and suddenly the A/C seems back to normal. It still is an old machine (maybe 7 years or so), but it can now chug along until something significant fails.
What I am wondering is how many expensive devices get prematurely retired because of bad push buttons. This is a Sanyo A/C and much better than my previous Westinghouse that got very noisy at this point in its useful life. I doubt that I could have found a repairman that would have resolved the problem for less than $100USD.
So beware of those tiny buttons. They seem to get dirty and wear out even without much use. I've been using the remote for nearly all the control over the 7 years.
For a year or two now, I have been ignoring the fact that the ON/OFF button didn't work as I have had the remote control to depend on. But the last week or so, the crud the fan has been blowing out finally made me decide to open it up and clean the interior.
After cleaning, the machine got unpredictable. Though I am not sure if I just had been ignoring the fact that it wouldn't cool under 27 degrees Centigrade as I have been running it 24/7 at that setting since the beginning of this year's hot season.
But now, I can't even be sure that it will turn on and off the output fan and when the fan is off, the cooling fins quickly ice up and the compressor runs heavy.
After much ado and poking at it for 3 days, I finally decided to take apart the LED and button console and check the push buttons. After all, I knew the ON/OFF didn't work.
The results are interesting -- all 7 of these buttons are defective in one way or another. And they did not fail open or closed, they provided various degrees of resistance. It is no small wonder the micro-controller on the machine is behaving oddly and the remote control could not respond properly.
So I replace all 7 with ones I keep on hand for reset buttons. Total cost of less that $1USD and suddenly the A/C seems back to normal. It still is an old machine (maybe 7 years or so), but it can now chug along until something significant fails.
What I am wondering is how many expensive devices get prematurely retired because of bad push buttons. This is a Sanyo A/C and much better than my previous Westinghouse that got very noisy at this point in its useful life. I doubt that I could have found a repairman that would have resolved the problem for less than $100USD.
So beware of those tiny buttons. They seem to get dirty and wear out even without much use. I've been using the remote for nearly all the control over the 7 years.
Comments
This unit has the IR sensor and temperature readout in the button panel - just just unplugging the unit was never an option.
Still, I am happy as a clam that I could resolve the problem with cheap parts and 20 minutes of soldering - hobby skills do have their rewards.
Good on ya, Loopy. Mad hobby skilz are very valuable in a world rife with unscrupulous "repair" people who show up, fiddle around, do nothing, then leave with payment. There's no shortage of Youtube videos for "air conditioner scam".
Good story on the buttons.
"Contents: 1,1,1,2 - Tetrafluoroethane (811-97-2), Cyclomethicone (541-02-6), Mineral Oil (64742-21-8), Fragrance, Dye"
Dye must be a code name for "misc contaminants", the stuff is clear. Mineral Oil I think I get. None of the rest makes any sense to me, and it might have come out before the CFC rules were in place. But I only use a tiny bit at a time, I don't identify any smell with it and the only side affect has been the enlarged head as you see in my picture. The stuff sprays out, foams up, and makes crackly sounds for about 30 seconds. I imagine there is some mechanical action associated with the crackly sounds that aids in cleaning, but don't know for sure. I just like it cause its fizzy.
I still cannot get the A/C to go below 26 degrees Centigrade, but the micro-controller no longer seems confused. My cleaning just may have been the final blow for the push buttons.
I now see one output pin on the UNL2003 is corroded clean thru, so tomorrow I will pull the whole control board and solder a new one in place. My guess is that a relay is not getting power to engage.
At least the temperature is comfortable enough while I am reverse engineering the problems. I don't see any other obvious failures, but I cannot see the backside of the control board until a pull it out and it is a single sided printed circuit board that has been getting drips of water from the cooling coils above for seven years. A tray is suppose to catch and remove all the water, but dirt and rust had clogged its drain hole.
The main point is that the failure mode has NOT seemed to harm any of the impossible to replace logic. And I am fortunate enough to have the damaged chip generically labeled.
I don't think anyone here will replace or repair the board - so this is all worthwhile as my only other option is to buy a new A/C.
I've repaired several household appliances found with bad buttons (or switches). Now my family thinks I can repair anything. My 3-1/2 year-old grandson told my wife, "Papa will fix it" after his older brother popped his balloon. I had to burst his bubble a second time and explain that I couldn't fix it.
I had a mother that would let her 10 year old son try to fix the washing machine and dryer. The amazing thing was that I could look at the mechanics and the schematics and actually do so. So I adopted the habit of opening up anything that was near junk and broken. If it is man-made, it can be man-repaired. Not always cost effective though....
When a grandchild asks you to make a sound like a frog, don't ask why. His mother or father probably said, "We will as take a trip to Hawaii when grandpa croaks."
I pulled the control board and replaced the damaged UNL2003 with a socket and a new chip. That got my beep in the unit back when I hit a remote button. That beep has been missing so long, I didn't know I had it.
So it appears that the UNL2003 was defective and driving the beeper as well as input and output fan relays. The one line of the UNL2003 that was obviously disconnected ran to the 'intake fan' and that would explain not getting down to a lower range of temperature. The output fan (into the room) has been fine.
It is back to running and testing again. I still may have a hung relay, a broken intake fan, or a micro-controller that has lost its way. I just have to wait and see. I've confirmed it reads temperature well by spraying the sensor with some cleaner that dropped it from 26 to 14 degrees Centigrade. And the unit does shut down when it reaches the target thermostat setting.
I just suspect that it can't go as low as it should with a target and continues to run with less that full cooling. This isn't that bad because I run the A/C at 26 or 27 degrees, a lower target is not really that important. But there is the challenge to understand and to do a complete repair rather than accept limited function.
I'll look around town for replacement relays before I go any deeper into the controller board. But at the very least, I can use the air conditioner for this season and buy one on sale in the off season.
I live 'a stone throw' from the arctic circle, and here we love this kind of equipment...
Of course, ours are built to 'run in reverse' and are used to heat our homes.
(Mine draws about 1KW at full power, but can deliver heat equivalent to 3.7KW if the outside temperature isn't too low.)
This technology is just getting better and better year by year. The units they sell now are way better than those from three years ago.
Unfortunately there was an installation problem with mine so after a few months all the coolant vented out.. during a cold spell, and when I wasn't home. Fortunately my brother-in-law handled it for me and got it all fixed under warranty.
It's also possible to use it in reverse (during summertime), which, in this case, means as a cooler.. but that's such a waste, I simply open the veranda door instead!
You can get it from places like Halfords (but don't use halfords because they are rob dogs) or any motor spares shop
I still cannot get below 26 degrees Centigrade and am convinced the outside fan motor is not working. Since I already repaired the UNL2003, I can check the voltages to the fan on the other side of the relays. It is a five wire setup with H, M, L, CX1, and an un-labled red wire. These would likely be something like High, Medium, Low, Current or Clock, and a common power line. I suspect it is all 110VAC as the unit is 110VAC and their are protection capacitors for the relay points that are rated at twice that one each fan motor.
Still, at this point I am more concerned about the mechanical state of the outside fan motor because of several things. The first is that i have used the A/C at 26 to 27 degrees for years, I just don't need any cooler. So it may have frozen up long ago and be beyond repair. Second, it seems that testing voltages and repairing relays when their might be a frozen motor might be a waste of time.
As long as the micro-controller is not in jeopardy of failure, I suspect it is best to work on this from both ends towards the middle.
This repair has offered up several interesting design insights. The most important is how one protects mechanical relays at 110VAC with rather large capacitors and a resistor. I am reading more about that. Also, I am not surprised to see the lowly UNL2003 driving the relays. There are 3 relays and with two darlingtons in parallel, one gets 120 ma to each. I know putting darlingtions in parallel is not quite kosher, but in this context it seems to have worked well. The 7th darlington is attached to a 12 volt beeper with 60ma supply - much louder than its 5 volt cousin at 20ma.
The micro-controller is an unknown 40 pin dip. I suspect I could figure out which it is by the power pin placements, but it is rather pointless to do so as their is no external ROM.
If all else fails and the A/C is no good, I think I will pull the electronics for further reverse engineering before pushing the unit down the road. There are several features that add rugged durability to the use of micro-controllers along side three very busy inductive motors.
Text books don't teach such details, but reverse engineering reveals all.
H, M, L appear to be a resistance ladder somewhere (in the motor or on the board). Power is available, so the relays are engaged somehow.
Tomorrow will check the motor bearings.
It seems all the electronics are now working properly and they certainly needed repair. Maybe I was too hopeful that these items were the whole problem.
At some point, one has to test to verify repairs are successful and an old machine may be have more than one problem. I am just in the testing and watching phase. Today, the air conditioner cooled the room to 24 degrees centigrade and I am letting it run to see if it will reach 15 degrees. If that happens, all seems okay.
It doesn't sound like it lacks refrigerant. Usually this causes the compressor to struggle and make a lot of noise.
The blockage you refer to could include a failed fan, and yes, leakage of refrigerant is always another issue -- even a small leak. Both of these are mechanical and take special parts or special tools. If that is the case, I'll just buy a new A/C. The unit was made only for the Taiwan market and generally a machine this old is no longer supported with parts.
SO, I pulled the unit out of the window and stripped all the sheet metal off it. The fan motors turn freely, but the outside cooling fins certainly looked clogged. The next move was to take the whole unit into the bathroom and wash out dirt from the fins - and there was lots and lots of dirt. I used a fine brass wire brush that is normally used for cleaning files and shook loose what appears to be about 5 years worth of carbon dust from the local steel and petrochemical industry.
Tonight the unit is drying out. Tomorrow I am hoping to power it up and have it work down to its minimum of 15 degrees centigrade.
BTW, I had another look at the button controls and it seems that they are set up in resistance ladders that are measured by an ADC. This certainly reduces the number of wires required, but it also is more susceptible to failure caused by the buttons creating their own resistance rather than a full on and full off.
The interior heat exchanger had a filter on it, but of course the exterior one did not. This is its last chance. If it doesn't work, it may end up at the recycler at this point. The only other reason for poor performance is probably a loss of coolant.
If you are wondering why the roundabout repair.... I just started out wanting to clean the interior fan as it was dumping crud on my desk. One thing led to another.
Yes, that is what frequently happens to me in the course of repairs. One thing leads to another (and several trips to Home Depot).
Hoping dirty fins was the problem. I got 3 refrigerators to work (and reputation as genius) just by wiping the dust off the coils in the back.
So now my Great Fear was that I'd used too much water and soap around all the electronics and mechanics. To make a long story short, it is back in the window and running nicely. I guess I will have to wait and see if I can get it down to 15 or 20 degrees centigrade, but I am sure that all the dirt that came out needed to be gotten rid of.
I am not sure one can merely wipe dust off coils. In this case it took a brass wire brush and a lot of soap and water. In other cases, a commercial vacumn and compressed air might be best. It was a very messy dirt job. But I may be able to squeeze a few more years out of this rather rusty machine for $2 USD of replacement parts and a lot of sweaty grimy cleaning.
I like the Sanyo better than my old Westinghouse because it is still running quitely and smoothly after many years of service. Having an old noisy a/c to put up with is a pain in the neck (and other anatomy).
At least the room is back to 26 and in my comfort zone for now. I am hoping that by the end of the day it will be quite chilly.
So it is back to studying the SX28 Assembly as I seem to have about 10 times as many of those chips as I do the Propeller.
Anything that man has made is worth taking apart and reverse engineering. You might even repair the beast.
As it is, I can at least remain comfortable while I shop around for a good price or wait for a Winter sale. Either the coolant has leaked or the compressor is worn out or both. Nothing left to repair, just use it until failure.