I saved a hard drive with some very important stuff on it exactly the same way. I had ham fisted the power connector in backwards and the sacred smoke had escaped. Lucky for me, I had an o-scope and was able to trace to a capacitor that had shorted in the 5 volt path. I wired up a new one and all was well.
You know that awful sinking feeling you feel when you've just done something monumentally stupid and embarrassing that you're going to have to fess up to? It feels twice as good +5 when you rescue yourself from it with your geek super-powers.
With hard drives, the easiest thing to do is use the circuit board from a different drive of the same make and model. Of course, if you don't have another circuit board, then you're in deeper trouble.
That's outdated advice, unfortunately. You can't do this with most hard drives anymore. To use one of my most-loathed manufacturers as an example (and the worst offender when it comes to hard drive recoveries), Western Digital stores head calibration data in two places: in the system area on the platters, and in an eprom on the circuit board. Both sets of data are matched at the factory and have to stay matched.
You can replace the circuit board on a WD drive with a match from the exact same model, from the same factory, manufactured the same day (guessing by the serial number anyway), and the drive will still fail to give you your data back.
When the system area becomes unreadable or doesn't match the contents of the eprom on the controller board, WD drives will show up in BIOS with a generic model name and no other information about the drive. There's no easy way to coax them into cooperating.
If you're lucky, you have one of the WD models with an accessible eprom on the controller board, and you might be able to either desolder and swap it with the replacement board, or, if you have the appropriate equipment, use a special clothespin-like tool to latch on to the chip's pins, read its data out, and write that to the donor board. If you're unlucky, the eprom is integrated into another chip, and nothing short of some pretty expensive equipment will be able to handle it.
However, at this point, almost everyone that doesn't work at a professional data recovery lab is waaaay over their heads, and only making the recovery job harder (or impossible) for the lab.
Additionally, a lot of WD drives now have the head stack spindle mounted by a screw to the top cover; once you pull the top cover, you've lost the head calibration for the drive. Getting any data off the platters at all at that point requires a moderately expensive homebrew tool with several calipers that screws into the head stack spindle; you manually adjust calibration until data starts coming in reliably.
We no longer do physical data recoveries in our shop; about the end of last year these piece-of-crap drives started coming in, and none of the old tricks worked on them. We looked into the equipment (and subsequent training) required to do these sorts of recoveries, with an eye towards providing very-low-cost recoveries (since we have plenty of other revenue streams), and we simply can't afford the initial $15,000 outlay for it.
We've made it our mission this year to get effective data backups working for every single one of our clients.
TL;DR: swapping controller boards and other tricks no longer work. You really really really need to have backups.
I did this once when a 120GB Western Digital hard drive failed on me. Spent several months searching ebay for "used" listings with actual images of the drive for sale, matched up the WD P/N, swapped the pcb boards, and was able to recover a year's worth of lost photos. Then I re-swaped the pcb boards and resold the "used" drive back on ebay. Good times:)
I know the feeling well. When installing an SSD into my pre-unibody, I screwed a big screw where only small screws go. Straight through the machine, and into the keyboard. I was lucky, and everything was okay, but it was a tense few minutes as I considered what a screw hole might do to a logic board.
Hmm... I guess I'll be the one who posts the Fauxgeek's Lament on this thread.
Reading stuff like this always makes me feel like a second-rate geek! I've had my fair share of soldering iron burns (for my senior project in college, I cracked open an RC car's remote and connected it to a BASIC stamp, which I controlled via RS-232 and a network socket), but the parts of this article which the author considers "easy" stuff are still a mystery to me. For example, "According to the application note I could use a simple signal diode, so I just pulled a 1N4148 diode out of my component drawer." A simple what diode? A 1N4-what diode? My component drawer? I don't even have a... help! S.O.S.! Man overboard!
I'm sure that I could come to feel more comfortable with stuff like this if I went out of my way to learn more about it, but the applicable extent of my electronics hacking day-to-day has been unplugging the rumble packs from my Xbox360 controller. Most of the things I use on a day-to-day basis are not devices that I feel comfortable screwing around with!
Reminds me of when I tried to fix my fried amplifier last Christmas. The amp was around 20 years old and had mysteriously stopped working in the night. Being cheap I enlisted an electrical engineering friend and together we stripped the amp down. However, the problem turned out to be 2 diodes failing hot. The short fried the micro-controller and associated eeprom thus making the fix more trouble than it was worth. Still, process of diagnosing the issue was a great learning experience for me. I almost exclusively work in software so it is always a pleasure to learn about the electrical engineering side of things.
I almost exclusively work in software so it is always a pleasure to learn about the electrical engineering side of things.
So true.
I was poking around inside a small headphone amplifier to see what was causing a crackling noise (I assumed a loose wire some place) when I tugged a bit too hard and popped off some otherwise healthy wires. I was crushed.
Then I realized I have a soldering iron. I did a crap job (I too am way more CS than EE) but I reattached the wires and got it working again.
Later I learned that the crackling was coming from the cable being plugged into the amp, which was basically fine.
A had a similar problem with a 20 year old amplifier that came for free when I got a pair of speakers at a yard sale. I ended up reusing half of the power supply and case to power a modern chip based amplifier that I built (PCB toner transfer and all):
There was another time when I fixed a big old CRT TV by just replacing a shorted capacitor (€1 part) on the high voltage transformer circuit.
The brand representative at the store where my parents took the TV said that it would be too costly to repair and to throw it into the trash, so they kept the TV until I went there for holidays break.
Last year a friend had her laptop showing just half of the screen image. I opened it, saw that it was a flat wire connection problem and installed ubuntu, replacing XP in the process. She would need to buy a new laptop, so instead she used the money "saved" and bought herself a nice digital camera.
And this goes on an on...My point is that some things are easier to fix than others and if you enjoy electronics and tinkering, it's a fun way of spending time, learn new things and to keep (working) stuff away from the dumpster.
Good story. Reminds of when I fixed a PlayStation 2 that wouldn't turn on by replacing a blown diode in its power supply. I wonder how many complex electronics are out there Sitting in landfills just a single diode away from working perfectly.
Probably most of them, even if you are experienced with electronics, with today's electronics fixing them is more of a hobby than economically viable. Something has to be worth a very large amount before it makes more sense for the regular person to seek out a professional to fix something like this.
Haha, I did the EXACT same thing on my (original 2006) macbook, 14 months owning it. My power supply died, so I clipped the cable, hooked it up to a 1A 18V variable power supply, but it wasn't charging fast enough, so I hooked it up to another that was apparently broke and sent 120V to my laptop (people at work don't throw broken things away apparently)
Anyways, I had experience soldering SMDs before so I wasn't too worried about the repair, but there were too many power mosfets and diodes and and a ton of devices at the same potential. Testing voltages and continuity with my multimeter was mostly futile (and I used datasheets the entire time). I'm glad you found the problem.
When I was young I remember consumer electronics, like tape recorders, coming with their circuit diagram as part of their documentation.
I bought a Teac 4-track open-reel tape recorder many years ago. Still have it, as well as the docs. I swear you could build one from scratch with the info provided.
The option to open stuff up (phones, radios, TVs, tape players, etc.) and make them do things it was not intended for was magical. Maybe Arduinos allow for something like that feeling now; seems infeasible for most general consumer electronics nowadays.
My best "open stuff up and make them do things it was not intended for" war story is from 1988. I had just purchased a Mephisto 68000 dedicated chess computer for my own recreational enjoyment. The embedded dev lab where I worked won a new contract and the hardware dudes decided a 68000 was the way to go. Not content to wait weeks, maybe months for first hardware, I decided to get some early experience by hacking at my chess computer. Basically I opened it up, reverse engineered the keyboard scanning hardware (which was implemented with standard jellybean CMOS logic), and commandeered a couple of inputs and outputs for my own purposes. I hooked these up to the printer port of my PC.\
Then I worked out a simple serial bit twiddling scheme to exchange bytes a bit at a time and coded it up in C as a kind of software serial port. On the PC side I ended up with something that looked just like a simple terminal that happened to talk out the printer port. On the Mephisto side, I replaced the 64K byte EPROM with a 128K byte EPROM. I changed the cold boot vector to point at "my" 64K, which first checked to see if the printer port was hooked up. If it was hooked up, then the CPU stayed in "my" area and ran a monitor I had coded up for the occasion. If it wasn't hooked up it vectored back to the original boot code and the chess computer worked as well as it ever did. I layered a loader on top of everything and had the incredible satisfaction of running standard C sieve of Erasthonese (sp?) on my new 68K computer, (complete with 16K RAM).
Ah good times. Pretty hard to do stuff like this today when everything gets packaged into mega chips, not to mention the tiny scale of everything these days.
Luckily not everything is unhackable. Sometimes you're lucky and find a company who is pro-hack-their-stuff who release everything (echematics, firmware source code). Heres something myself and my brother recently completed:
We plan on releasing some proper video demonstrations of it in action and to write a few blog posts about how we did it and our experiences when we find some time.
My electronics instructor still uses his amplifier from the 70's because he can rebuild it with common parts...nowadays if an amplifier section breaks chances are that it is a chip which may or may not be available (many manufacturers keep no backstock of one off chips, therefore there is no way to get replacements).
One of the things I really took with me from his instruction is a disgust at the "Death of the TV repairman". People would rather throw something in the garbage instead of having it fixed...not that it is always their fault. It has become cheaper to purchase an item than to have it fixed, that combined with the fact that most manufacturers typically do not release schematics until the item is out of production (if they do at all) and an increased effort by manufacturers to not allow anyone other than their techs to work on things (i.e. special Iphone screws, warranty voiding tape, part numbers removed from chips).
Remember, just because it is small and complex does not mean it cannot be fixed. If it was built it can be rebuilt (or unbuilt, see ifixit).
I'm not so sure it's like that. Consumer electronics have gotten more and more integrated, and the bits that are assembled rather than integrated are assembled with a higher and higher density. The point is, the time spent finding faulty components and fixing and / or replacing them (or jerry-rigging something like the article here), multiplied by the hourly cost of a person with sufficient training to do that job, at this point usually exceeds the cost of replacing the complete component.
He could have found the same SMT diode as he was replacing had he tried...he rigged it because it was the same part (electrically) just a Through-hole part.
Just because a device is "dense" or heavily populated does not mean it cannot be repaired if done properly, if the schematic is provided (and accurate per revision) and you can still source the part it can be repaired. Even IC's can be removed it just takes different tools (i.e. solder pot, hot air pen, hot plate, etc).
I assure you there is no difference between troubleshooting old tech and newer stuff its just different. An IC is just that an "integrated circuit", meaning they just took what used to be an entire daughter board and made it on a silicon chip. I would actually argue that it could be easier in some cases to find an issue on newer stuff since if a chip is failed you just replace it (if you can get it, which is the real problem).
It's just like he said in the article use the datasheets if you don't have a schematic typically they are just using some circuit that someone designed 20 years ago and put in the datasheet as an example.
I understand the labor cost involved in fixing an item but I would argue the long term cost of replacement/improper disposal (even recycling...have you seen the little kids with the campfires of circuit boards in india/africa that are trying to reclaim the valuable metals...that is the cost I am talking about).
As an electronics tech I can tell you that I have worked on many complex circuits in the field and in production and I have never had a problem with the complexity...but I have had many issues with finding "one off" IC's and schematics.
Things are not always about money...sometimes its best to fix things just to fix them (and hopefully learn something at the same time).
Also, consumer electronics are built to fail. If they were not the manufacturers would require some IPC class, nobody wants quality though...they want cheap.
It's not just consumer electronics that suffer from integration -- it's pretty much all appliances.
We have a fantastic old (probably 10+ years) clothes washing machine. The lid switch went dead about 2 months ago. We looked at (possible) replacements at a retail store and realized that there was absolutely _no way_ to repair many of the newer models -- a process you can count on given the MTBF of many electronics components. A $10 replacement lid switch from Ebay and a 15-minute solder job and the old machine was back to working beautifully.
It's sad... despite having a professional electronics lab in my home (including hot-air rework, stereo microscope, and a great oscope), I doubt I could repair many (most?) modern electronics failures.
Nice, we had a 2 year old range stop working in our condo...it wouldn't heat. So since we were upgrading appliances one at a time that was obviously next (it sucked anyways). So I traced the problem back to the control board...$500+ online, so I ripped into it, found the burned solder connection (admittedly very simple fix, but still a very simple circuit otherwise), resoldered the connection and now we have a working stove to sell/trade/donate. Total cost: 2 hrs and less than a penny worth of solder.
BTW, I am jealous of the stereo microscope and hot air rework station...those are on my list for my shop (Just got a Rigol oscope to hack a few months back)...Sometimes, I miss working in production and having access to all that..although I have better handtools now.
top tip: use an ice cube tray to hold all the screws you remove. I remove the first set, put them together in the first cube hole and continue. then just reverse the order of cubes when you are putting it back together.
I've been bitten by that scheme after I'd carefully disassembled my laptop, putting all the screws on a neatly labelled sheet of paper, only to later attempt to blow out some dust, and have them scatter across the desk and floor.
Double-sided tape, or actual compartments, sound like a much better idea.
Another strategy is to screw them back in, lightly, into the same holes they came out of (after removing the thing they were holding down of course). Not practical with a lot of gear though.
And use wide masking tape over each cube as you are done so that when you (or the cat or the kids) knock it the screws don't disappear into invisible cracks in the floor.
Actually I tend to have strips of tape attached to my desk, with the sticky-parts up. The ends are turned over and stuck to the desk.
I stick the screw heads to the tape. They stay pretty well, are easy to get-to, and it's easy to arrange them as you want. I usually put them in a compacted form of the layout they were on the board.
I just put the screws back into the screw holes that they came out of, after removing the component that it was holding in place. And take lots of pictures along the way.
Another technique that I found handy when replacing a Macbook keyboard assembly: I used a piece of foam roughly the same size as the notebook, and stuck the screws in their representative places after they were removed.
That model in particular had many screws that shared the same pitch, but with different lengths. The thought of piercing something with a 10mm screw in a hole meant for a 6mm screw was scary. I thought this technique worked well.
We use a simple sheet of pre-printed paper with diagrams for common laptop parts and areas, and then cover it with sticky-side-up packing tape. Each screw gets stuck head-down to the tape, according to the precise location and laptop part it came from.
I use my Flip for this. Photos good, video better. Also a great idea for when working on cars. I'm thinking of getting a gopro sd hero so that I can track my attention as well. For only $100 why shouldn't I document all of my repairs?
A trick I sometimes use on cars is that when you take each part off you put all the bolts in a ziploc and mark it, then put the bags in series in a cardboard box or whatever...then you just take each bag out of the box in reverse during assembly since typically the same parts go back in the same way...helps avoid the extra screw (or the dreaded short one screw/nut/bolt which is usually more puzzling for me).
I always want to do photos but I am usually too dirty and/or lazy.
I had a Samsung Laptop that I broke while trying to downgrade the Bios image without properly downgrading some other controller image first. The only problem was that the soldered EEPROM had the wrong image and thus the laptop decided not to start. Samsung wanted to charge 700+ Euros (CPU was soldered to the motherboard).
I was really confused that the world's biggest electronics concern wasn't able to replace an EEPROM.
Luckily I came across bios-fix.de and paid 65 Euros including shipping to get the old eeprom soldered out, flashed and soldered back in.
I wish bigger companies would actually try rather than going down the service path of least resistence :(
Always remember the Maker rule..."If it is broken it is fair game..". Open it up, do your best, if you don't fix it but you learn something then its time well spent.
If you are like me you also salvage as many parts from what cannot be (or is not worth) fixing there is always something else to use the parts on.
Really handy stuff and a reminder that "this product contains no user-serviceable components" is rarely true. I also liked the structure of the post. Isolating the lessons and documenting his work while still discussing the higher-level implications certainly made it more useful than just a walk-through of how he fixed the motherboard.
Having such a position in Greece right now can lead to bad reputation I think (of course the state of the economy is not his fault) :) Diomidis Spinellis is an alpha geek and a great scientist. He has written great books such as Code Reading, Code Quality and he is also a FreeBSD committer.
Replacing through-hole parts on motherboards is exceptionally difficult. There's so much copper in a modern multilayer board that it's very difficult to get enough heat in without cooking something delicate. Even with preheat, a reflow gun and a powerful iron it can be very hard going.
Replacing broken PSU connectors is one of the most common laptop repairs and should be an easy job, but has a dismally low success rate.
If you are replacing the connector and you don't care about reuse I find it is best to cut off the part then just do cleanup, remove as much as you can to eliminate the heatsinking. Then its up to tip size, temperature and flux...even still I can't say I have not lifted a few pads in my day.
Common sense will get you a long way. I was replacing failed memory chips on a Vax 11/780 and doing other hardware-level things in pretty much the same way; the downside of screwing up was a multi-thousand-dollar service call to DEC, since our machines were not under maintenance contract. Well, and bringing a machine that twenty people were using down, for days (which would have been the more expensive piece of the equation).
Lesson 0 has to be 'have several years electronics repair skills under your belt' - and that's the kicker for most.
I agree though - a bit of knowledge can be a powerful thing; a 'resistor pack' for my car (to regulate the 4 speed aircon blower) is 80UKP + labour to replace when it fails but the most common fault is that the thermal fuse (approx 0.80UKP to replace) blows when the airflow is reduced by an air filter that just needs a clean-out.
I fixed my t42 motherboard by soldering a tiny wire over a blown fuse.
I was in the process of replacing the failed fan on the CPU heatsink. (Had to transplant the fan from a smaller heatsink that came from a lesser t42 with a smaller heatsink that didn't fit... Long story.) Prior to the procedure, I wanted to carefully see if there was even voltage on the FAN pins. I wasn't careful enough and thought I shorted the pins. Indeed, when I connected the working fan, it didn't work. So I figured that a fuse blew somewhere. Did some searching on the web, found a partial schematic for the board, found the 1mm x 2mm FAN_FUSE (or something) and soldered a wire on top of it. To my surprise it worked. Still using that t42 when on the go.
I once had to replace my laptop's motherboard because the gpu part of it fried due to a malfunction that cause alot of heat. I had to buy another motherboard at Asus, that was actually pretty cheap, 150 dollars, but I didnt want to pay another extra 100 dollars for, the technician to change the boards.
Well, I couldn't find the service guide, all the ones I found were diagrams, but not a guide on how to dismantle the laptoo, so I decided to figure it out myself. After 4 hours I managed to take the whole motherboard apart and extract all components, then another hour to install the new one. It was pretty challenging, but I managed to close the case with all parts in place and no extra screws lying around.
Sometime Ago, I tried to repair a Motherboard form a Dell Laptop.
The best thing that I learned these days was photograph every part of the disassemble process, and where you put each screws and cables, with this you can go back in the time and see how to reassemble much more easily
I love stuff like this because the time it took him to fix it was probably worth more than the cost of the motherboard BUT the fact that he pulled it off is what makes him so valuable (and cool ^_^ ) and he learned a lot while doing it. Simply awesome.
Yes. However, this guy is also the general secretariat of information systems of the greek government, and oversees the online tax reporting system, which has been having extended outages in the past months. I 'd rather he fixed that instead.
> Well, the diode did as it was supposed to, and prevented the reverse-polarity from damaging anything of vital importance.
I don't think so. If the diode were being used to prevent reversed polarity, it would simply have not conducted. If it were a normally-reverse-biased shunt across the power supply, then blowing it would have left the laptop unprotected but otherwise functioning (and it would have had to blow something else in the process, like a fuse). So I suspect it was intended for something else.
According to the data sheet http://datasheets.maxim-ic.com/en/ds/MAX1908-MAX8765A.pdf "D2 protects the MAX1908/MAX8724/
MAX8765/MAX8765A when the DC power source input is reversed. A signal diode for D2 is adequate because DCIN only powers the internal circuitry."
Off topic: Anyone else read this title as "What I Learned From Fixing my Mom's Laptop"? I think I have been doing too much tech support for my mom and her friends lately.
For a self-learner at a casual pace, try starting with something like an Arduino; it'll abstract away a lot of the more complicated ideas at first, and give you a ton of positive feedback along the way. It's a great bridge to hardware for a software guy.
For more formal stuff, see electronics/digital computer fundamentals/etc. courses at your local community college or university, or see if your local Maker community (if there is one where you live) can point you at more targeted local resources.
You know that awful sinking feeling you feel when you've just done something monumentally stupid and embarrassing that you're going to have to fess up to? It feels twice as good +5 when you rescue yourself from it with your geek super-powers.