Subaru Power Windows Saga -Part 3 (Repairs & Failure Analysis)

[@hillbillyrigrepairs1814]

Thanks for not editing out your mistake with the test light-we can all learn from each others screw ups.

[@ufartface]

trump could learn from you

[@vinsitang8067]

Hahahaha you good❤❤

[@ratbag359]

all the zeners will be 5v or 5.1 the big one is power supply in conjunction with the large resistor. the smaller two are to clamp the output voltage. this is a circuit that is only safe for a led test light less than 20ma with no inrush.

[@PineHollowAutoDiagnostics]

Love it. Will take a look. Thank you Josh!

[@Graham_Wideman]

You beat me to it: grahamwideman.wiki.zoho.com/PHAD-Subaru-window-regulator.html At least we more or less agree! The only oddity is that the SMD code on ZD1 appears to indicate it's a 22V zener.

[@TheRetiredtech]

@@Graham_Wideman but it was something in the switch he burned out with the test light

[@BerlinCityTrucker]

There is a small error in the circuit. R2 and R3 are 10 ohms. a 10 with multiplicator 0. A hundred ohms would have 101 on top of it. 10 and one 0.

[@ratbag359]

@@BerlinCityTrucker yes indeed my mind was primed as ivan said 100 ohm did not think about it. no wonder protection zener and resister died.

[@porfiriocastillo4065]

Ivan after you fried that component with your test light I bought an LED one. I have never damaged anything with a test light but these newer cars are something else. Simply Diagnostics did an amperage check on test lights the in rush on incandescent light is up to 1.2 amps. Thanks for the case study keep them coming.

[@engeneer_ru5827]

Yes, yes, Steve did a whole study of the starting current! I overtook Ivan in the laboratory :)

[@fieldsofomagh]

Wow, 1.2 amps is over the top !!

[@ericlisin9509]

WOW. I read a comment from "window saga Part 1" about how you deserve an award!!! I know I usually fast forward thru YT vids to get directly to the final fix, but you gave your audience a theatrical performance that your viewers personally have frustratingly gone through to diagnose something that would "normally" be fixed by a "new part!" Whomever you specifically are or perhaps only known "Pine Hollow Auto Diagnostics," you have a gift! Even though the last post I notice was two years ago (or however posts date themselves) I hope someone influential sees your videos and sees the gift you have.

[@josephdunton3651]

I've been anxiously waiting for this one, thanks for posting! I agree with you-this switch lives in a car door and it is asking for trouble when they are made this delicate, as the age of the car and TSB attest...

[@boomer9900]

You had to do it. My wife started complaining that a few days ago her Outback's (2015 3.6) passenger window was acting funky, going up almost all the way and back down. I was wanting to see the struggle to get that regulator out and back in. I've personally known a few people with window regulator switch problems. Ford Expedition, BMW 325, Hyundai tiburon. Our toaster died a few weeks ago. It's internal power supply is regulated with a 12V zener, fried by electrical storm. I hot wired it to run with a 12V wall wart, just cut the pcb traces and soldered on some wire. Maybe take another road trip to Niagara Falls and help me put in that regulator motor contraption. Good job on this video.

[@robertthomas8240]

My girlfriend just replaced her battery and then had P/W problems. So before digging in I happened to watch your video. Boy am I glad I did! I learned so much by watching all 3 videos. I will throw my "shade tree mechanic" test light in the trash! Thanks!!!!! Great Job!

[@poching0521]

Thanks for sharing. I guess all new cars have power windows step motor control circuit now...Share my basic understanding below. It looks as if 12V you fried and bypassed is the power supply from the switch to the sensor/encoder circuit. U1 and U2 probably are voltage regulators/controllers from 12V to 5V for the sensors/encoders. Zener Diodes z2 and z3 are probably the over current protection diodes. I guess the actual sensors/encoders probably are a b c d still berried under plastic cover. 5 V quadrature signals you observed are feedback signals back to switch control. When the smart switch is on and sends 12 V power to the motor, it monitors the 5V quadrature feedback signals to control motor rotation steps. If the switch did not see the the 5V quadrature feedback signals, it will stop 12V supply to the motor.

[@dianekeim1611]

I hear you on your exclamation, how about circuit overload protection on motor on the circuit board at the motor also, without it may burn up board. Usually you use a relay (12v) pulled in by 5v switch, the clicking he was hearing may have been a relay. And using a test light is trouble exspecially if you don’t have a print on the circuit board, putting 12v with current on the output pins of a 5v chip will damage it(voltage will back feed through the transistors and toast them)If you can see the numbers on the chips google them, you can then see the circuit schematic. I am guessing here on the 12v voltage regulator

[@dianekeim1611]

I am guessing here on the 12v regulator, may be using 5volts from switch to bring in 12v regulator( making it look like a relay) . Thus bring in the motor and using regulator as over current protection too

[@kanjorsx4969]

It’s very common I work at a used car dealership and we’ve had at least 3-4 of these come in with window issues

[@briantii]

BMW: We are happy to announce we have created the best, most complicated automotive window management system ever devised by man. Subaru: Hold my beer.

[@oldschool0072]

Lol

[@bigdaddymak1439]

This Subaru system goes wonky when it doesn't have good battery voltage! Would have liked to seen if you would of put a new battery in to start maybe only the window switch would of been bad! I think on modern cars don't diagnose anything until you have a known good battery! Especially one that's spanked

[@d.d4184]

I always feel if the battery is over 2 years old replace it. Charging seems to be a waste of time. I rarely change batteries.

[@bigdaddymak1439]

@@d.d4184 I own a 2016 outback the batteries from the factory are junk! But in these modern cars where everything is electronic you have to keep a good battery!! Can't wait to replace a bad one!

[@darwindwelle5267]

You know Ivan just watching these 3 videos reminded me of why I hate wiring problems. My hats off to you, your patience far surpasses mine because I would have been throwing things in the first video. They really are making cars to where we can't work on them anymore aren't they??

[@DIYDaveOK]

Love the post-mortem and teardown/failure analysis. Tiny current spike from who knows where blew that Zener diode. Makes me ask, "what's wrong with a hand crank" LOL 😄. Thanks as always, Ivan!

[@ufartface]

nothing except you cant reach them

[@leefra1]

Good job on a insanely complex system, Ivan. As a retired electrician I was always taught when soldering, never rely on solder for a mechanical connection. Make the mechanical connection first ( as in twisting together), then solder for the electrical connection. Not meant as criticism, only to teach. You have taught me much, thanks. Lee

[@rollandsicard1628]

Ivan, you are persistent. I have one of these cars with this same power window system. I am so glad you ran into this situation. I now know what the system is made up with. I was an Aviation Electronics Technician troubleshooter and can very much appreciate your work. Just trying to understand this power window system design. Looks like 5 volt zener diode and perhaps a power mosfet to actuate the regulator motor. Finally got the original Subaru Service Manual for the car. What a MESS of a manual. So glad you are on KZbin. Thanks again.

[@PineHollowAutoDiagnostics]

Thanks Rolland! Yeah I don't even read Subaru factory service info anymore...it is useless haha

[@shawnlockard2370]

Truly a pleasure to watch you work. Thanks again.

[@alward9901]

Hi Ivan I did the same as you stripped down the gear & motor after getting a new actuator cleaned & greased the old gear cleaned the motor & ajusted the brushes and the old one worked fine on direct wire and battery test , so I just kept it Honda actuator is $ 380 Canadian I also put extra grease on the cables of the new one as you can tell I worked backwards should not assume you need a new one defore looking at the old one so I have a spare .

[@n3roc]

Ivan is the best! Even when he blows a diode! Keep up the great work. Lookin forward to the next project.

[@billyyoder8171]

Thank you Ivan. Good job. I believe the wire you call orange is actually pink. Orange is on terminal 10. Pink is on terminal 5. Pink Terminal 5 is next to the large Green Terminal 14. Have a blessed and safe week to you and your family.

[@PineHollowAutoDiagnostics]

Remember the incorrect wiring diagram?

[@cinquecento1985]

What a saga. Im so sorry for the incident with the test light. Very frustrating, but you did it. But in one thing I disagree: If you make hacks on a car, like jumping wires, you should make it obvious and dont hide it, you should know that the next guy is having a bad time finding that one out. I once had a annoying problem with a microswitch in my driver door, after changing the 300 dollar doorlock 3 times I decided to do a jumper wire, because the car did not lock the doors. I left a note in the door and in the fusebox for the next guy. Because as I found out later my jumper wire had another consequence: If the car doesnt know that the door was opened, after 90 seconds it locks all doors, and in this way you can lock your key in the car. No problem if you know it. I follow you channel for a week now, and I learned a lot, thank you. Good job on doing the videos, your work and the explanations.

[@johna2327]

Digital chips often work on 5V. One should never try to apply more voltage to any of their pins. Many digital components have little or no protection against the inputs or outputs going above their positive supply voltage or below their negative supply voltage. Most digital components use CMOS technology in them. They are voltage sensitive. It is not always the current that will destroy them, but the voltage. Connecting a power probe or test light between the input or output of a digital component to the car's positive battery can destroy them with very little current. It only takes a micro amp or less to fry a 5V device with 13.8V car voltage. If you wish to test if an input of a digital device works by applying a positive signal, it must be done with a voltage source which is the same as the digital device. In this case 5V. Never try to force an output of a digital device to ground or to positive at any voltage, unless you are certain that the device supports it. Such as open collector outputs. They can be pulled to either the positive or negative voltage, but rarely both. Normally open collector can be pulled to ground, but not to positive. If you are working with a 5V digital circuit, don't use power probes or test lights. Use a scope or meter. If you want to pull an input to ground for a test, use a 1K or 10K resister and don't connect the other side of the resistor to positive 13.8V. Connect it to DIGITAL ground only. The 5V reference from the ECM/PCM is generated by a voltage regulator in the ECM/PCM. They usually have built in overload and short circuit protection. They can normally handle 1A to 2A. It is designed to be a regulated power source for digital devices. Digital components have outputs often designed to run other digital components that use only a few milli- amps or even micro-amps in some cases. They can't run heavy or even small current devices like light bulbs without drivers or buffers to handle the current needed. They are designed to operate based on digital voltage levels, not current. Most of the ECM/PCM outputs have drivers designed to handle moderate current to run things like injectors. Some of the inputs to the ECM/PCMs are designed to be connected to 5V sensors. They may not be able to handle your test light connected to positive car voltage either. Using a 5V power supply running through a 1K resistor and into the input of the computer would be better than taking the chance of damaging the computer with a power probe or test light.

[@hnd2893]

I'm just seeing this. I don't know if you got a fix yet. Here's a few tips from my experience. All I work on is Subaru's. 20+ years. I had a Genuine Subaru of that vintage window motor/regulator assembly that was mis-boxed, the difference is the number of pin connectors in the motor connector, auto up/down vs non. If the wrong motor and or switch combo is installed, you will only be able to operate the window up in increments as yours is doing in the veideo. The front passenger window switch has to be relearned first for auto up/down. Never use a test light, see service manual for window testing, use a scope to check for command signals. The upper window rubber run channel gasket gets sticky and will trigger the anti-entrapment function and make window bounce right back down, there is an updated window seal available. Lube channel with slicone spray, if it fixes issue, replace channel rubber seal. There's a TSB for parasitic battery draws, remove FOB from car. There are recommended electrical accessory switch settings to be left in the on position that are controlled through the body control unit before testing. There is a time frame for the test in which the ELCM may activate during the test. There are software updates for charging system logic. The original equipment batteries are underrated, install a 550CCA battery. I typically see 30-40mA's as a norm.

[@fieldsofomagh]

I am a big fan of tear downs and surgical procedures.

[@Chris-Fennimore]

Great fix Ivan! So when my wife's 2018 Impreza starts having window switch issues, I'll be sure to call my favorite Russian KZbin Diagnostician. Very cool 👍😎👍

[@ednewell1847]

I have let the smoke out of an a/c amp on a 05 outback, testing the ground feed for the mode door actuator with an incandescent test lamp (250mA). I had to own it and replace it with a used unit out of pocket. Learned a lesson that day

[@PineHollowAutoDiagnostics]

Interesting...

[@shuriKen469]

that's some sweet bonus footage right there! at least the power windows didn't crap out in the middle of winter.

[@kamikazekeebler]

2011 forester here. Driver side and both back passengers windows don't work. Great series of videos!

[@jeffjankiewicz5100]

Neat soldering iron, Good job Ivan. That job would have driven me nuts. My test light never shorted anything.

[@jcmobilediagnostics8611]

Ivan the surgeon! That’s a neat little solder iron.

[@miken8150]

Post a link to that soldering iron!

[@fapapas4534]

Wow... I'm having same issue but only the driver's side window. New driver side switch not the problem. Thanks for excellent video.!

[@pauldavis5075]

The incandescent test light has zero ohms and super high current for a few milliseconds as it is turned on. It's pretty cool on a oscilloscope. Ohm's law and a resistor in series for 250ma would fix it. I work on radio equipment and my radio has 10 indicator bulbs that are 10 years old run 12 hours a day but there is a current limiting resistor in series with each one. The $250 transmitting tube has a 6 volt 20 amp filament that has a soft start system that uses a relay to step up current in two stages and the transmitting tubes seem to last forever. Thanks for all the videos.

[@TreyCook21]

I have an '03 Legacy with the big EJ25... sure, power windows work beautifully. She can't hold her oil in though!

[@mikechiodetti4482]

Zener diodes by the band on ZD2. There's up and down positions on the switches and when the window reaches full top or the bottom, the sudden rise in current should be enough for the controlling module to turn off the power to the switch or motor. Or the resettable circuit breaker opens. BUT NO! We have to have an encoder (counter) to number the counts from top to bottom and not expect the occupants to have to release the switch itself! My 2020 RAV4 has the same one touch up and down for driver and passenger windows. Don't know about the rear windows. Can't wait for that problem to surface! Oh well, vehicle is fixed. That's all that encode, I mean counts!

[@luigigonzalez5062]

Finally part 3 is here :) Tell owner of the car to keep an eye on the fuel pump as well

[@Paolo0902]

You have to start using LED test light for this kind of electronic components. You did not take into account in-rush current for incandescent test light (it can draw for the first time 1.5 A). Check with picoscope and amp clamp.

[@rusty6666]

Glad to see you aced that one Ivan and it all came good in the end once again nice fix . Thanks

[@thomasbyzewski2687]

This lady is very lucky to have you!

[@baxrok2.]

Well, this makes me even more resolved to buy older vehicles. I don't need a single thing that newer cars have to offer. Great series Ivan. Impressive as always!

[@dc5723]

Haven't damaged anything yet since I have 2 LED test lights. My only incandescent test light has a burnt bulb and never got around to replacing it.

[@AP9311]

Wow, Ivan, great fix. Awesome car story. Pretty interesting find!! Just that tiny thing causing it to go wonky!!! Great video series!!!

[@HUBBABUBBADOOPYDOOP]

Thanks for the dissection. The filament of the test bulb is huge compared to a Zener that small. I can imagine it was a millisecond of Hell inside that circuit.

[@ratbag359]

they should of gone with a higher limiting resistor.

[@TheRetiredtech]

But he burned out something in the switch. The motor had already quit before he probed that side

[@throttlebottle5906]

@@TheRetiredtech I missed that at first also. I'd like to see a switch tear down next.

[@vettepicking]

A lot of window motors in new cars with limit stops need to be run "with the up switch" outside of the door before installation to reset the limits . This is done on cars that have auto up switching to prevent crushing dogs and kids.

[@xanderlander8989]

ZD3 is probably a zener diode. Thanks for the diag. We have a '16 Forester, same window system.

[@briankotze57]

On the front of the shaft that black ring has magnets. So if the motor turns it works like a normal hall effect ckp. So the magnets pulls the circuit down. So it doesn't have any current going through those smd's that burned. Only job is to react to the magnetic field as motor turns. And that is the square wave you are seeing and that is why it needs to be initialized so that the window can learn the open and close positions. All modern windows and adjustable seats that have a memory works that way. On seats you get motors that have 2 sensors build into them

[@androidemulator6952]

..also learned that when diagnosing Subarus, at least keep a trickle charger on the battery,,, notice how more and more weird faults crept in as the battery volts dropped lower. (Wot, still no "low-voltage" alarm on modern cars ???. ) Keep up the great work Ivan !! :)

[@hatfez]

Have "Big Clive" check out the circuit!

[@douglasjones3799]

You are a very determined man Ivan. Good series!

[@PineHollowAutoDiagnostics]

Ha this car went from amusing to frustrating at one point :)

[@douglasjones3799]

@@PineHollowAutoDiagnostics I saw! Very perplexing case. But it made for a good video!

[@ggstephens50]

Ivan would be interesting to take those door switch’s apart and trace the 12v feed , could have a really thin trace that couldn’t have handled the 200mA load of the test light.

[@Automotive_Solutions]

I love pulling stuff apart to see how it works / what broke.

[@walibran]

ZD 1,2,3 are zener diodes used to set up either voltage references or constant current sources for U1 and U2 which I believe are Hall effect sensors. The black ring that has 6 or so circular segments are probably magnets that when the motor spins pick up the gear spin and are delayed by a few degrees between each sensor giving you that delayed dual square wave to indicate direction of motor spin. That is read by the electronics in the switch to control the motor. The test light probably induced a surge current that fried ZD3 and disabled one of the Hall effect sensors. Without BOTH square wave signals the circuit in the switch could not provide continuous motor operation thus shutting down the motor after a short run. Test light bulbs only draw 200ma while lit and the filament is hot. When cold (initial startup) it can draw about an amp for a few milliseconds. That is enough to fry the zener. Those zener diodes are only rated for about 100 Millie arts and are NOT FORGIVING when too much current is applied. In the future, especially with newer cars, use a 500 ohm to 1k resistor to jump voltages in sensitive circuits. Carbon film or carbon resistors have no surge current and should be safer. Love your channel but I advise caution with newer car circuits. They are all low power control with pulse width modulated FET outputs these days. Power ratings CANNOT stray from specifications even a little bit or it will blow, usually with expensive consequences. Thanks for reading. Electronics tech from jersey.

[@PineHollowAutoDiagnostics]

Awesome info! EXCEPT the test light burned out the 12V source in the SWITCH, not the motor. The motor sensing circuit was already failing way before I saw the car as evidenced by the funky "Auto-Up" operation ;)

[@walibran]

Sorry for that. I can’t argue with positive results but that whole repair made me cringe. I hope it doesn’t come back to bite you in the future. As I am dabbling in automotive repair in my spare time ( mostly to save money) I hav found that if you fix a car for someone, even if you don’t charge for your time, you become “married” to the whole care for life. I also recently repaired my Mitsubishi ECM by purchasing a used ecm on eBay and used only the flyspec sized FET from it to fix my original ECM so I wouldn’t have to spend $140 to have the ebay ECM reprogrammed. Car is running fine 6 months later. Keep teaching the good electronics knowledge. I am learning a lot even at 58 years old!!

[@PineHollowAutoDiagnostics]

@@walibran That's fantastic! Thanks so much for the comments :)

[@chuck10088]

I suspect ZD3 is a 5.6V 300mW zener, (using ZD2 as a reference) R3 looks burnt, but, it is a 10 ohm resistor. A 100 ohm would be marked 101. Not sure what U1 and U2 are, as I couldn't track down the actual part numbers. The car manufacturers certainly don't make your jobs any easier. Enjoy your videos, keep up the good work.

[@Graham_Wideman]

Did you manage to read an SMD code from ZD2? I think I saw maybe 562, which leads to PZM5.1NB2, a 5.1V zener, but it's in a SOT346 3-pin package, so not the diode on the Subaru board. FWIW, my schematic: grahamwideman.wiki.zoho.com/PHAD-Subaru-window-regulator.html

[@pvillehick67]

These parts should be covered as a safety item.

[@larrylytle6482]

2015 Forester 86000 miles. Driver's side window stopped working on the auto down mode. Then would only go up 1" at a time. Returned to normal operation after a few days. Probably need the updated switch. Just bought a 2007 Ford Focus with 22,000 miles and manual windows. Perfect solution.

[@philipaustin8710]

2014 Forester did the 1inch increment after a winter freeze of the window. Returned to normal a few days later after warmer weather.

[@toppledgod]

ZD = Zener Diode, the question is what voltage, but you could probably thow a few guesses at it...

[@kenalex0353]

Was the low current circuit (which was damaged) there just to allow the window to operate after the ignition was off for a period of time or until the driver opened the door to exit..

[@Cowboy_Ash]

I believe you are correct

[@throttlebottle5906]

it's all part of motion/speed/load sensing setup for when something gets closed in the window, it senses the stall/speed change, then the "door-switch" circuitry reverses it to free whatever was stuck in it(fingers, kids, pets, other junk)

[@SmittySmithsonite]

Wow, what a nightmare - I'd have been losing my mind trying to diag this one! Nice work, Ivan! Love that soldering iron, too. Nice FAST heat-up time! Automakers just love to throw electronics where they don't belong! My 34 year old Grand Marquis with 269k miles (and counting) still has all 4 ORIGINAL OE motors and switches in it, with ZERO electronics, and they still all work great! I had one problem with one of the rear window motors many years back. The plastic drive broke inside the motor. Turns out, .45 ACP lead bullets fit perfectly in the 3 equally spaced parts of that triangular shaped drive, so I stuck those in there, and it's been working since 2011! 😂

[@PineHollowAutoDiagnostics]

Amazing fix haha

[@rickrogers2649]

A case where firing the (little) parts cannon worked : )

[@roxanneabbott8424]

Awesome Ivan you nailed it as always!! Loved this case study!

[@PineHollowAutoDiagnostics]

Thanks Roxanne!

[@Jason7591969]

Great video! I'm about to change out my master switch. In my case, none of the windows will go down. Not even from the actual door. None of the lights under the switches are on either. Do you think it would be the master switch?

[@arthurfricchione8119]

Ivan first I'll say 👍 your interest in trying to find out what failed. To bad you didn't have an updated switch to disassemble to see what was new and improved LOL sure it cost a lot for just another cheap update. Anyway another interesting video Ivan. Ivan I'm taking a In Cylinder Waveform class given by Brandon Steckler. 8 hour class broken down to 4 two hour sessions. Really knows his stuff. You and wife stay safe and well Artie ❤️

[@tedgourley7783]

My 2017 Outback driver side switch will not operate the passenger side window; I bought a new battery yesterday. The moral is, sell the 2017 Outback before (rather than) repacing the battery. Could the battery installer have temporarily connected the new battery (with cables??) before removing the old battery, thus avoiding the momentary total power outage?

[@rampage28200]

Been a mechanic over 40 years , it’s just one of those jobs should’ve walked away from. Those cars are from Mars

[@PineHollowAutoDiagnostics]

The only two jobs I have walked away from was engine work on newer BMWs xD

[@2packs4sure]

@@PineHollowAutoDiagnostics Sounds like there's a story there and I'd like to hear it,, I loves me a good BMW horror story !!

[@fmbfla]

"Men Are from Mars, Women Are from Venus (1992) Elders should know of this writing your comment suggests that we men made this thing

[@tomvleeuwen]

I'm not a mechanic in the old-fashioned way, I only do diagnostics and electronic repairs on cars, and those are the kind of jobs I love. Especially when the dealer walked away, when it starts to involve board level repairs or immobilizers. I like to know every little detail on the type of cars that I work on, and I discover new details every time I do a job. Getting my hands dirty on a bearing replacement? No way, where's the fun in that?

[@cletusgaming6108]

A friend of mine that loved to go hiking had a subbie and didn’t have brake lights on it when I was following him in a unfamiliar area I noticed it so I informed him of said problem and ended up being a connector that became unpinned

[@bobbydelamar606]

So what about those test lights with the circuit board and LEDs? Are they safer? I've never fried anything either but here at my job they told us not to use test lights.

[@PineHollowAutoDiagnostics]

You have to know the circuit

[@bobbydelamar606]

@@PineHollowAutoDiagnostics 👍🏻 If in doubt, DMM.

[@justincase3880]

In the hobby servo world they call it an “H” bridge .. runs the motor in either rotation on demand .. what the 12v regulator/bias voltage supply does is a mystery .. old school H bridge system was mechanical..used 4 micro switches at the levers, that opened/closed “diagonally” (with the motor in the bridge of the H) when the levers were pushed down, or pulled up...

[@josephwash109]

Yes, I'd agree the ZD components are Zener diodes. They're probably used for voltage regulation purposes (that's where the 5V square waves originate). If it literally burned up, it may have just failed because it was a defective component (the heat cycling eventually killed it). The resistors are probably used with the Zener diodes to maintain the 5V regulated circuits. I really find it hard to believe that the test light fried anything. I read comments about people talking about inrush current that caused the failure. However, if that 12V circuit is a low current signal circuit, it should have a terminating resistor that will limit its current, even if it's shorted to ground (like you mentioned). The inrush current of the light bulb should be of no consequence; it's not like it pulls a ridiculous amount of current at first like literally shorting a wire to ground.

[@PineHollowAutoDiagnostics]

Exactly my thoughts, Joseph. The test light didn't even flicker on that 12V source from the switch. Never seen a 12V feed that wasn't protected by either a fuse or resistor :)

[@bluegrallis]

My 2014 Subaru Outback recently lost the auto up and down on the drivers side window. It also only goes up in one inch increments.

[@High_Tech_Mountain]

Great case study Ivan. I haven’t damaged any circuit w/a test light. Seems like those switches are subpar quality. The switches failing after test light use. Is why you spliced the wires together right? Great case study and repair. Looking forward to diagnosing these more complicated window switches and systems! Always a pleasure!

[@PineHollowAutoDiagnostics]

Thanks so much for the comment Angel! Can't believe that I had to splice those wires to get the stupid windows to work, but hey they work perfect now ;)

[@joolsgrudgings5776]

I would advise that you tell the customer to lubricate the window runners with a PTFE based lubricant periodically to stop future failures. I also think that the flat battery certainly played its part in this too because as the the voltage decreases the current increases putting extra strain on those already fragile components - as always a great video ......

[@aurthorthing7403]

I had similar issues with Hondas. I'm still guessing you need a logic pulser to deal with those things.

[@bobweiss8682]

The "ZD3" probably means a Zener Diode, used to drop the 12V supply down to 5V for the Hall Effect sensors (U1 and U2)

[@ratbag359]

ZD1 is power supply ZD3 is protection

[@ckm-mkc]

ZD3 is identical to ZD2 from looking at the circuits. Probably directional sensors. Shameful to have a part fail because of a $0.001 component. The manufacturer might as well have done it on purpose.

[@TheOnespeedbiker]

Did a search on blown zd3; Re polarity - the diode should have a band nearer one end. This band is the cathode. Usually labelled k in schematics. The un-banded end is the anode. Usually labelled a in schematics. On a standard rectifier diode: The positive + supply enters the diode at the anode and exits at the anode. However, the component's ID is ZD3, which tells me that this is a Zener diode. A zener diode is connected opposite polarity to a standard rectifier diode, as it only conducts when reverse voltage from k to a exceeds a pre-determined level. This is probably part of a voltage regulation or voltage clamping circuit. ie: when voltage on 'k' line which is positive goes above 'X' volts the zener will conduct, shunting voltage to ground or another part of circuit. So on ZD3, band is +. A new diode is advisable, as it serves as over-voltage protection. The caveat is what voltage rating of zener diode you will need. A schematic of the sub will be needed to know this....

[@kennethbode2017]

sometimes you just have to re-engineer stuff. your diligence paid off

[@TheCrazy243]

Hey Ivan! Great video! I think you should get “Lisle 24550 Computer Safe Circuit Tester” . Most test light draw about 225-230 mA. This one only draw 15-17 mA. It’s inexpensive test light and I think it should be safe for modern vehicles. Again great video. Learned a lot from you.

[@alward9901]

Hi Ivan just a short word on my last comment I did not buy the actuator from Honda I got a after market for $90 so far works well for the last 10 months .

[@mrb1864]

Least it all working ... I damaged an old ford ecu testing temp sensors, had 2 and I done some tests on wrong one with the trusty test light and it was running like junk, realized likely ecu and removed for off board check and had get board repaired at my cost lol . some of best education you get is from mistakes :-)

[@paulk437]

I was wondering why you saw 2 out of phase square wave signals coming out of the motor back to the switch. The following youtube video explains this, contains other encoders but the hall effect discussion answers my question.

[@dwoodog]

This could have easily turned in to a dealer / other shop washing their hands of this window issue. Glad you stuck to it. Wish I knew you when I had my 1990 Chev Beretta.

[@alfredlyne4438]

I have seen some concerns regarding ZD1. From the reverse engineered circuit diagrams I would say that it just a voltage transient limiter with R1 limiting transient current. The Hall effect ICs can probably handle around 30V without damage. D1 provides power supply reverse polarity protection. A 200mA test lamp will kill the 5V output zener diodes if connected from battery positive to the outputs, assuming the output resistors are 10 ohm. The output resistors will get a bit hot, but should survive. As someone suggested, the circuit would be far more robust with 500 ohm output resistors. It would at least allow you to poke around with a test lamp anyway you feel like.

[@PineHollowAutoDiagnostics]

Great info! But the test light had nothing to do with this failed component...

[@alfredlyne4438]

Maybe battery positive from the driver side electronics?

[@justincase3880]

How many milliamps does the orange wire pull? Maybe use a limiting resistor .. check for parasitic draw ...

[@viewmaster6949]

Ivan thanks for the very valuable information because it would surely kick my ass in Jamaica because I don't have a scope as yet getting it soon

[@BigJohn4516]

There’s probably a surface mount fuse or diode between the ignition power and the motor feed inside the switch.

[@wwdiesel]

Wow, what a bunch of over designed power window circuit and components. Way to many fragile components just waiting to fail. Just curious, so you think my Snap On 12V LCD DC Digital Display Circuit Tester would have caused the same failure as your incandescent test light did?

[@markorlando4817]

Great informative video but, I’m wondering why you think you shouldn’t have replaced the switch you damaged?

[@boblee4508]

Is the orange wire and circuitry at risk of damage in fault circumstances now it’s fused at the higher rating of the green wire? Maybe an appropriately sized in-line fuse for the link would address the issue.

[@bettylane1984]

That was one of the biggest brainteasers! As good or better than anything out of Staten Island. Thanks!!

[@zeuscarver3112]

Im curious what the internals of the switch look like. What component would fail from a low current ground fault.

[@Mrjm5411]

To help visualize the markings on IC's and other similar parts temporarily lay a piece of frosted tape (Scotch Magic Tape) on the surface of the device and rub/pressure the tape the markings will appear. Hint #2 if when the tape is removed the markings are now transferred to the tape you have found a counterfeit part. Another thought - The window circuit as poor a design as it appears to be was designed to purpose. Do you suppose by jumpering the power directly as you did that some of the safety in the circuit may have been bypassed? i.e. Window Jam?

[@saarike]

Really frustrating case and fine solutions, but your nerves were like steel! This was like Sherlock Holmes worst case. 👍

[@johnwaby4321]

A nice tricky job . Testing your expertise. 👍👍

[@b61mack56]

I can't wait until she goes to upstate NY and has a widow problem and she stops at SMA to fix it. Eric will go nuts because everything looks like it hasn't been touch since the factory. You may want to put a tag on the wiring to give a heads up to someone else that my work on it in the future if she sells the car.

[@VWWRENCHIE]

The tag would read.."SBI" Serviced by Ivan"😎

[@stealthg35infiniti94]

Would it be nice to put a piece of Duck Tape stating " This _____ circuit has been modified." on any non factory rewiring? Techs like Eric O have pulled their hair out trying to troubleshoot modified wiring...

[@mikepratali5582]

Or Russian Modified

[@cyberwiccan3598]

I would have tried to run the jumper inside the switch and left the wiring as standard.

[@walterk1221]

@@cyberwiccan3598 plus, if switch assembly internal space permits, a picofuse. I don't like the bypassing of the fuse protection that was obviously internal to the switch assembly. Curse on Subaru for not making it known that wire is internally fused at a low value.

[@brainndamage]

Both ZD3 and R3 look to have failed. R3 is 10 ohm, ZD3 is likely equal to ZD2 so the marking on it will say what voltage it is. They look like they're on the output signal to (try to) protect it from overvoltage and overcurrent - obviously they're not designed good enough as they shouldn't burn up even with a overvoltage. The thing is, if the overcurrent surge that burned up the zener caused it to fail open, the 12v got to the 5v sensor chip (U1) and caused it to die as well, even with no outside marks. ZD1 and R1 are obviously the input shunt regulator to drop the 12V to 5V with the small ceramic cap to filter the 5V.

[@brainndamage]

R2 and R3 should've been at least 100 ohm (worst case current 7V/100ohm=70mA, zener dissipation 350mW, assuming a 5V zener) if not even more, plus an additional resistor between the sensor chip output and the zener.. The way they designed it, the worst case current is 7V/10ohm=700mA, zener dissipation 5V*0.7A=3.5W, resistor 7V*0.7A=4.9W. That tiny zener can't be rated at more than 100mW and resistor at 0.5W, no wonder it blew both.

[@dariopalace]

Hi Ivan , I'm having almost the same problem with my 2015 Subaru impreza, but first the driver side window was not going down , then the front passenger window it was working so I put it down and stop working and now I can't make it go up , I have my battery check and charged, they told me that is ok. and then all suddenly all the windows don't work from any switch of the any door. then I disconnected the battery and connected back and now 3 windows work but the front passenger window still doesn't work , and is down, and here in Michigan is getting so cool and starting to snow , so tomorrow I will get a new battery and if you have a chance can you tell me how to reset my windows using the switches? , your video has been so helpful and very interesting, I've learned a lot from it , but I still have problem with my window, any help will be very appreciated.

[@PCBoardRepair]

26:30 U1 & U2 are hall effect IC's and that is what ZD3 protects one of these 5V lines and you test light blew it by forcing 12V onto a 5V line frying the part....do not do what IVAN did

[@cbattman]

How come it's not under dealer warranty after only 2 years? In the UK we get minimum 3 Yr warranty. Is this not the case in the US?

[@altontyler865]

Having pretty much the SAME window issues with my mom's 2013 VW Beetle! They worked fine until I replaced the battery a few months ago. Almost identical scenario. I've fired the parts cannon and replaced the door switches. Not so bad as they aren't that expensive, so I'm down to replacing the motors now. After searching online for hours and reading many forum boards about this, I'm not the only one having this problem. VW says the original motors do not have enough torque to overcome the "Pinch" safety feature. The new updated motors do, according to them. The crazy thing is I haven't been able to diagnose fully, as the owners manual has NO details on fuses on either fuse panel. No included diagram on the panel or the owners manual. I can't find an exact match online either because "They" say they changed sub models many times during the model year. These smart windows are a POS that's for sure!

[@wyattoneable]

Interesting tear down of the regulator. What a strange sequence of events to bring you to this point.

[@jjtoms3341]

Can you do the same to the driver side switch? Love to see what failed

[@PineHollowAutoDiagnostics]

Didn't see any burn marks...power transistor stuck on or commanded on by faulty logic :)

[@JD-iu3vi]

Instead of straight wiring the orange and green wires, why not connect them through a PTC resettable fuse?

[@ggstephens50]

Hey Ivan. That magnetic wheel on the motet shaft reminds me of a VRS ABS sensor

[@JOHNSUE28]

Ivan, This is a true story I swear. Last week I recieved a Call from A 2018 Subaru Outback Owner. Her concern was Her Drivers side Power Window is inop. I went to AllData and looked at the Wiring Diagram and immediatly had Questions due to the lack of explaination of the Switch Circuitry, And Diagnostics. I then found the TSB for the switch, And then watched your Repair. What is a tech to do for proper Diagnosis? I just dont feel comfortable with the lack of info. Didnt Clint Eastwood say a "Man has to know his Limitations". I'm just not clear still how to properly Diagnose this issue correctly.

[@charlesmiller5078]

Now that you have the new one installed, hold the button down and see if it burns up.