one of our people spilled diet coke on his laptop while it was on...instead of turning it off immediately, he kept using it because it was still working...he turned it off later and ever since it won't boot (but it powers on)...a cursory review didn't get it working quickly, so we just bought another one...i, of course, get to keep the old equipment

it's not an old laptop...a pavilion dv2000 series with a 14.1" screen, 2ghz c2d, 4gb memory, and 400gb 7200rpm hard drive...well, i have the same series myself (though with a faster proc), so at the very least i have parts (including a new screen) in case anything goes wrong

but what i'd RATHER do is get it up and running for use as a file server (my current one is a p4m laptop with no screen and 1gb of DDR, so it's running a bit long in the tooth)...i don't need performance, as it's for my own personal use (mostly i just pull files i need when i need them, but i leave it on 24/7 so minimal power use is nice)

anyway, i posted looking for a new board (socket P, aka micro-FCPGA) since i have all of the components, but if i can just fix this one, i'd rather do that (otherwise i'll just part it out and keep the screen)...as it stands, it powers on with no beep codes, and it never posts...it'll keep running indefinitely, as far as i can tell, but i don't think the hard drive is spinning up (though i know it works)...the CPU fan does adjust speed, rather than running full




is it shot? if so, meh

4/20/2009 8:47:32 AM

Shot, you'd need some fairly sophisticated tools to even have a chance of reworking it.

4/20/2009 8:49:51 AM

pics of the other side plz

4/20/2009 9:05:58 AM

I have an SMD workstation that would easily pull all the surface mount chips off--but beyond that I have no idea. If you need to use it just let me know.

4/20/2009 10:55:18 AM

clean everything up, look at the traces and see if they're shot or not. if they aren't, you have a better shot (since it probably didn't corrode the lower layers as well)

you need sexpencil's rework station. luckily it looks like it's confined to one small area so it shouldn't be TOO horrible.

i'd definitely say it's at least worth a shot, though.

[Edited on April 20, 2009 at 11:04 AM. Reason : also, be glad it wasn't regular coke ]

4/20/2009 11:03:50 AM

Quote :

"clean everything up, look at the traces and see if they're shot or not. "

Seriously. I've cleaned up similar looking nastiness before with just some isopropyl alcohol and a pencil eraser, and the board underneath was just fine. You'd be amazed what dirty traces from an acidic soruce will fuck up.

4/20/2009 2:01:01 PM

Quote :

"clean everything up, look at the traces and see if they're shot or not. "

Seriously. I've cleaned up similar looking nastiness before with just some isopropyl alcohol and a pencil eraser, and the board underneath was just fine. You'd be amazed what dirty traces from an acidic soruce will fuck up.

4/20/2009 2:01:01 PM

Quote :

"clean everything up, look at the traces and see if they're shot or not. "

Seriously. I've cleaned up similar looking nastiness before with just some isopropyl alcohol and a pencil eraser, and the board underneath was just fine. You'd be amazed what dirty traces from an acidic soruce will fuck up.

4/20/2009 5:13:22 PM

looks like a maxim dc/dc controller fried.

whats the top say?

http://datasheets.maxim-ic.com/en/ds/MAX8743.pdf


I take it there is a fried inductor / pair of mosfets on the back of the board?

[Edited on April 20, 2009 at 8:33 PM. Reason : .]

4/20/2009 8:29:33 PM

Donate it to me I need something to tune my car with and its romulator locked....i dont want to lock my work laptop to the car.

4/21/2009 7:59:43 PM

top says MAX8743 +EEI 0819

i cleaned with some 90% alcohol, a q-tip, and an eraser...i'm putting it back together now, to see if that did anything...i'd really like to salvage this thing if possible...what are the odds it's salvageable if the chip is replaced? would any of you work on it, and how much would you charge?

4/21/2009 8:57:42 PM

no dice, still not posting (though still getting power, cpu fan spinning, all LEDs working)...the first picture is the cleaned up area, and the second is the bottom of the board directly beneath the damaged area)

4/21/2009 9:29:25 PM

You see the shorted traces on the dc/dc, right? Have you tried to take a multimeter and see which voltages you have and which you don't? As that is apparently a cpu regulator, it certainly makes since that you are getting some power (including the CPU fan which is fed from a different regulator) but no posting.

If you have a fairly fine tipped soldering iron, get some solder and clean those traces up and any of the others that look like shit and try again.

4/21/2009 9:52:18 PM

I find it difficult to believe that the MAX8743 DC-DC chip is still okay after being shorted like that. It almost certainly needs to be replaced. Its a good chance the passive components are okay, they're usually a bit more forgiving.

You'll need a hot-air station to do it though (or some very creative soldering iron/heat gun tricks) Luckily like most chips you can always get a couple free samples from the manufacturer! (http://www.maxim-ic.com/quick_view2.cfm?qv_pk=4356&t=or) You want the EEI+ model.

4/22/2009 12:27:37 AM

maxim controllers in that package don't have heat slugs. you could remove that with two good soldering irons that have half inch flat tips. if you're asking for help here im assuming you dont have that kind of tool at home.

As far as debugging goes this is what I would do.

turn it on and probe the inductor , square labeled 2R2 , with a volt meter. probe the side facing the maxim chip. report pack what voltage you see. it will probably be easiest for you to measure directly across the cap beside it labeled 220. you could measure the power good pin but just measuring the output is going to be easier and if the controller implements it as open drain (i'm sure they do) and the designer didn't pull it up anywhere you will not measure anything anyways. Plus its hard to probe those damn legs.

The worse cast scenario is the power supply failed in such a way that the high side mosfet stayed on and shorted the bucks VIN to the output (through a small LC filter at least!!!!). The fact that you need a charge pump powered 5V cap placed on the switch node to even pull that off makes it unlikely (VGS > Vin + Vth). The next thing to fear would be that the power supply going to zero caused some power sequence latch up.

the good news is that is a horrible dc-dc controller so you wont be chasing down feedback components to replace.


good luck!!



[Edited on April 22, 2009 at 7:47 AM. Reason : .]

4/22/2009 7:46:10 AM

^^ haha, that's awesome...i filled it out and they're sending me two

^ & ^^^ i have NO idea as to what the majority of what you wrote means...i have a multimeter at home (which i've never used), so i might as well give it a shot and i'll report back the numbers i get...if they seem wildly off, then i'd assume i'm doing something wrong or testing the wrong area

as for the rest, i spent all of a semester in computer engineering before i decided to switch colleges, so i have absolutely zero experience working with components past basic computer building/repair (i mean, i consider myself more capable than your average user, but not regarding the technical physical aspects of the components themselves)

i'll do the test and report back...i would assume that this isn't really all that hard if you know what you're doing, and shouldn't take TOO long to fix...if that's the case, and anyone is willing, i'd pay to have it done...just let me know what you'd charge

4/22/2009 8:27:43 AM

Quinn, do you have switchmode design knowledge? I've dabbled with them enough that I generally know how they do what they do, but I don't know them well enough to make statements like this off the top of my head

Quote :

"The worse cast scenario is the power supply failed in such a way that the high side mosfet stayed on and shorted the bucks VIN to the output (through a small LC filter at least!!!!)."

4/22/2009 8:41:45 AM

The bottom fet would have to turn on occasionally or I don't see how the boost capacitor could maintain a high enough voltage on the top fets gate. With damage like this anything is possible....

To answer your question though I design dc/dc converters for a living. The grand majority are sync. bucks just like this one.

I'll be honest though I haven't used a controller that REQUIRES >=ESR on the output cap in a long time. I've actually never done one myself just changed/fixed old stuff. It's hard (read impossible) to have a low ripple voltage when your putting the current sawtooth out of the inductor into a large ESR value. I also graduated college when large value low voltage ceramic capacitors were affordable. No 'taint'-alums here!




[Edited on April 22, 2009 at 2:41 PM. Reason : so can i have it?]

4/22/2009 2:35:08 PM

Do you mind me asking how you went about getting that type of job? When I came out in Dec 02, I really wanted to land some sort of an embedded systems type of job. Now that I am unemployed I would like to do the same thing, however, just as it was then, I have yet to find an entry level embedded systems job, they all want at least 5 years and some of them that know tons of engineers have dumped people out on the street are asking for Sr Embedded engineers with 10 years of experience and the damn pay range is 60-70k.

4/22/2009 2:49:17 PM

Co-op transition which seems to be the only way to get past the 5 year or masters requirement.

If you want to start off in analog , power supplies are probably a good stepping stone. Almost every element is used in a digital fashion so its a good "walk before you run". Anything half on outside of pass transistor in an LDO is usually a very very bad thing. Kind of conducts, kind of gets hot as hell , kind of smokes, kind of reflows...... You will also learn a bit about control systems but for the most part the feedback loops are simple. If i make that little in 10 years I will have already chosen another career path. There is nothing about engineering that excites me enough to put up with that. I will say I am currently thankful to have a job and I do enjoy it. Good luck though man!



so can i have this laptop?

[Edited on April 22, 2009 at 6:32 PM. Reason : .]

4/22/2009 6:31:50 PM

no, i want it to work, especially since it seems like a pretty simple fix!

4/22/2009 6:54:32 PM