replacing/bypassing dc jack on a laptop

[bpesta22]

Greetings

Anyone ever repair or replace the DC jack on a laptop? It seems like a straightforward procedure (from posts on the net).

I bought a new jack and was going to soldier wires to the mother board (and just let the new jack hang outside the case). The problem is, with the old broken jack off, I can't tell which connectors on the board go to the positive line and which go to the negative.

Can anyone help? I will post a picture of the MB soon, gotta dig up my camera.

[bpesta22]

I doubt this will be helpful, but here goes...

http://c2.ac-images.myspacecdn.com/images02/19/l_36d7053588c74a8394a63d8ff3059909.jpg

[bruto]

If you still have the broken jack, can you trace which contact on the board goes to which contact in the jack? If so, you can use a DC multimeter to get the reading off the plug.

I did a similar replacement on a laptop a few years ago and it worked all right, but in the one I did, I couldn't find a center-pin style power jack with an outer casing, so I used an in-line headphone type jack. If you do this, you should be careful not to plug the power supply jack in when it's plugged in at the mains end, because the plug will momentarily short and make a nasty little spark as you plug it in.

[ThatSoundAgain]

Measure which pin connects to the large exposed copper islands around the screw holes. That would be my guess at the minus.

Usually, it's best to determine polarity before removing the part. Are you sure there's only two leads in the plug? I ask because some power bricks are intelligent (I guess have part of the PMU system on them), and need more than two.

If you're worried, and the repair is worth a few dollars to you, service manuals can often be found for sale online as downloadable pdf's.

ETA: IANAEI, but I've done this type of repair before on other stuff. Also, if you decide to have the lead hanging out, be sure to use solid wire and to secure it to the case with some slack inside, so the next accident doesn't break the board. What I usually do is make the hole for the wire myself, with as tigh a fit as possible. Then thread the wire through (a bit longer than needed), solder both ends. Then I pull a bit of slack into the case and fasten two small locking strips around the wire on either side of the casing. Though, if you can manage, a flush-mounted plug is much nicer and more durable.

[Dan O.]

It's actually quite obvious. If you have a volt/ohm meter, you should find that the 4 points around the perimeter are all connected together and these would be the outer shell of the connector which is probably ground. It looks like there is ground trace going up along the back edge of the circuit board to the screw hole. The volt/ohm meter would confirm that these are connected. The center point would then be connected to the center pin of the power connector.

[Dan O.]

Measure which pin connects to the large exposed copper islands around the screw holes. That would be my guess at the minus.

Though it is typical for most power supplies to use a negative ground, it is not universal. Your power brick will usually indicate whether the center contact is positive or negative.

[bpesta22]

Though it is typical for most power supplies to use a negative ground, it is not universal. Your power brick will usually indicate whether the center contact is positive or negative.

Thanks for the help so far all.

So, I would need a negative wire that touches all 4 outer points, and the positive to the center?

Sorry if this comes across as dense-- never done this before (I figure either it works or I still have a broken laptop).

[ThatSoundAgain]

bpesta, be sure to listen to Dan O. Hopefully you still have the broken off part so you can check that the lone center pin on the PCB is indeed the center pin in the DC plug.

Thanks for the correction, Dan. I'll keep that in mind when looking at scrounged electronics. Is the negative ground an engineering convention, or are there specific reasons for choosing one or the other?

[Dan O.]

Thanks for the help so far all.

So, I would need a negative wire that touches all 4 outer points, and the positive to the center?

Sorry if this comes across as dense-- never done this before (I figure either it works or I still have a broken laptop).


If you provide the make and model of the laptop, we could look up the specs and give you a better answer.


I'm assuming that this one uses a coaxial power connector (plug with a band on the outside and some form of contact on the inside) given the round hole in the case. The outside connection is the ground and the inside connection could be either positive or negative depending on the whim of the manufacturer. Your power cube will probably have a little icon that shows which is which. [something like: (-)---C---(+) or visa-versa]

[Dan O.]

Is the negative ground an engineering convention, or are there specific reasons for choosing one or the other?

US autos are negative ground so that's what most engineers grew up with. Telephone systems use 48v positive ground and probably had a very good reason likely dealing with corrosion of exposed equipment (or maybe it was just to be different).

[BigAl]

A VOM used on the wall wart will tell you the polarity unless I am missing something.

[bpesta22]

Well, I'm a definite rookie, which is why I posted this.

I have a volt meter but there's no power to the board, so I'm not sure how that will help. I know there's a way to measure whether things are connected-- is that what I want to do?

Also, suppose I do figure it out and the get the soldering done. I suppose there's no way to know it works til I get the laptop back together (or could I test it with the power on once the motherboard is back in and secure)?

what would happen if I tried to plug it in-- torn apart-- after the soldiering job?

We could be witnessing me destroying my laptop, but I'm ok with that.

[Gagglegnash]

Hi

What's the make and model of the poor baby?

Have you googled for technical and repair stuff on Teh IntrTubz?

One way to find out the polarity of the expected stuff is to put the old broken off connector onto the business end of the wallwart cable and VOMulate the sundry power-related bits that come out of the connector. Of course, this only works if what's left of the connector isn't s dead short. The deadshortedness of the connector can be determined using the O part of the VOM as a continuity tester while it's still disconnected from any wallwartesque entities.

[ThatSoundAgain]

Basically, you measure if things are connected electrically by measuring resistance. This is done using an ohmmeter, which is very commonly combined with other instruments. If you have a voltmeter, chances are it's really a multimeter that can measure resistance as well. Look for the word "resistance", "ohms", or the greek omega.

You can test your ohmmeter by touching the electrodes together. If it's working and powered, you should see it read a value very close to zero. This means zero resistance, or directly electrically connected. Infinite resistance means no connection at all.

So one simple test to do before you plug it in would be to measure the two leads / pins inside your newly mounted socket. They should not be directly connected.

If at all possible, get someone knowledgeable to help you. The mechanical aspects of this takes practice, and it's hard to convey in writing how to know if you've soldered a good joint or not.

In any case, try to minimize the heat and handling you put the rest of the board through. Some of the chips and surface-mounted components are likely pretty sensitive.

[Bob Blaylock]

Basically, you measure if things are connected electrically by measuring resistance. This is done using an ohmmeter, which is very commonly combined with other instruments. If you have a voltmeter, chances are it's really a multimeter that can measure resistance as well. Look for the word "resistance", "ohms", or the greek omega.


Of course, if you're a Borg or a Vogon guard, then this won't work, because in those cases, “Resistance is futile!” or “Resistance is useless!”, respectively.

[bruto]

Well, I'm a definite rookie, which is why I posted this.

I have a volt meter but there's no power to the board, so I'm not sure how that will help. I know there's a way to measure whether things are connected-- is that what I want to do?

Also, suppose I do figure it out and the get the soldering done. I suppose there's no way to know it works til I get the laptop back together (or could I test it with the power on once the motherboard is back in and secure)?

what would happen if I tried to plug it in-- torn apart-- after the soldiering job?

We could be witnessing me destroying my laptop, but I'm ok with that.

Use the voltmeter to determine which pin on the power supply plug is positive, and which negative. Using that, you should be able to determine which pin in the jack you have removed was which, and using that, trace which pin went to which traces on the board. Even if the jack is broken, it's almost certain that the outer conductor on the power supply plug is going to go to the motherboard ground, so if that is negative, it's negative ground, etc.

p.s I think that's basically what gagglegnash said, too, but there's probably no harm in saying it two different ways.

As to plugging it in while still disassembled, that will depend a good deal on what you had to take apart to get access, but you should be able to plug in the power supply, and if there's a battery in the machine, it should light the charging light even if other parts of the computer are unplugged/disassembled, etc. I would try that if possible before snapping everything back together. If you get the charge light, it should mean that you are getting current, and that it's the right polarity.

[Gagglegnash]

Hi

Of course, if you're a Borg or a Vogon guard, then this won't work, because in those cases, “Resistance is futile!” or “Resistance is useless!”, respectively.


Futile and useless, but still measurable!

[Gagglegnash]

Hi

Oh! Hey!! AND!!!

Remember that one of the best ways to protect the rest of your computer from excess heat, quite counterintuitively, is to get a VERY HOT soldering iron!

The hotter you can get the tip, the faster the place you actually want to solder will get up to soldering temperature. The faster it gets up to temperature, the less time there is for the heat to migrate to somewhere you do NOT want it to go!

Also, in the long run ('long run' being a measure of the length of time you have to expose each element of your work site to the heat of the soldering device) you'll pump less overall BTUs (or Watts, I forget) into the board by drastically shortening the length of exposure.

Alligator clips and scrunchy aluminum foil 'U's that you can pinch down around your work site make Ok heat sinks, too, and a wet sponge nearby for wiping excess solder off the soldering tip will do a lot to change your attitude about soldering.

[bpesta22]

Thanks all. I plan on doing the operation tomorrow night. I will let you know how many burns and / or electric shocks I get.

fwiw, here is exactly my laptop mb:

http://www.powerjackrepair.org/gatewayta6.html

and here is the fix that I am going to attempt:

http://www.laptoprepair101.com/laptop/2006/05/27/failed-laptop-power-jack-workaround/

[bruto]

That should work fine. The one I fixed before was a Gateway similar in appearance to yours.

I do wonder why they can't do a better job on these things, considering how heavily stressed that connector is. How hard would it be to anchor the power jack to something other than the motherboard?

[Dan O.]

This is a laptop power connector done right: Magsafe.

[ThatSoundAgain]

So, any news? I'm running out of finger nails, here.

Dan O., you're right. Now if only they made a magsafe for USB, firewire, ethernet, DVI, and audio connectors, I'd be golden.

[bpesta22]

Eh well, it took forever to figure out using the resistance thing which prongs mighta been positive and negative (the one in the middle beeped when the other lead touched any of the 4 outer prongs. I finally found what I thought was it and soldiered things in. I didn't put it all back together, but did test whether the battery would charge-- it wouldn't.

So, then I figured screw it and I spent 10 bucks on ebay for the actual replacement plug, which when it comes should figure out the positive/negative thing itself.

I did not burn myself, and two of my soldiers actually looked decent, though one would require a do-over. Also, my iron turned black with soot (I think) for reasons I don't understand.

[ThatSoundAgain]

Hmm.

IIRC, the explanation the alloy of the soldering tip will react pretty quickly with air when heated. That's why the first thing you do is to tin the tip, that is, cover it with a thin layer of solder. This will both protect the tip and conduct heat more efficiently to what you're working on. After you solder each joint, and after tinning, wipe the tip on a damp sponge (NOT nylon).

Here (http://www.makezine.com/pub/nl/33) are some tutorials in video and written form that should help. I like this one (http://www.instructables.com/id/How-to-solder/).

I recommend reading / watching some of those, and if you happen by an electronics store before your part arrives, a few dollars will buy you a prototyping board and some cheap resistors. These are perfect for getting the knack of soldering quickly. It's not difficult, and within the hour you'll be able to reliably solder a passable joint before you go on to fix the computer.

You'll also need to get the old leads off your board. De-soldering. Some people like a spring-loaded vacuum tool for this, others, like me, use solder wick. This will expose your board to more heat, but less mechanical stress than the sucking tool. Be careful, and if you at all can, try it out with throwaway parts.

Good luck!

[ktesibios]

Solder wick is a very bad choice for desoldering components from multi-layer PC boards with plated-through holes, which the board in your laptop almost certainly is.

In multi-layer boards, connections between layers are made by plating the inner surface of the through-holes with copper. When the components are soldered in place, the holes become filled with solder all the way through the board. Wick, which soaks up melted solder by capillary attraction, like a sponge, will remove the solder from the solder-side pad but usually fails to clear the hole. If the hole isn't completely cleared of solder, when you remove the component it's very likely that it will take the through-hole plating, or a component-side pad, with it. When this happens you have the tedious job of restoring the broken paths with wire jumpers.

The only good method of desoldering from a multi-layer board is to use a tool which sucks the solder out of the hole. There are power-operated vacuum desoldering tools which do an excellent job, but they're too expensive to be attractive to the casual DIYer.

I can recommend one of these (http://www.radioshack.com/product/index.jsp?productId=2062731). The iron has a tip with a hole in it. You squeeze the bulb, place the tip over the protruding lead you want to desolder, allow a few seconds for the solder to melt all the way through and release the bulb, sucking the solder into the iron. Then you squeeze the bulb again to spit the molten solder back out, preferably into a safe container. It takes a bit of practice to become handy with the tool, but once you've mastered it it will do a very good job.

I do this kind of work for a living and I've been using this kind of tool regularly for quite a few years.

An obvious alternative which I also recommend, is to canvass your friends to find one who is an electronics tech and solicit their help.

[bpesta22]

KT

Thanks, I was just going to post asking for tips on de-soldering!

Got the part and if I can get my smegged up soldier out, I should be ok-- assuming I didn't ruin the board with my tinkering.

The saga continues.

Problem is that pump you linked to costs more than the dc jack I just bought. Eh well.

[ThatSoundAgain]

ktesibios, that desoldering tool looks great! I'd never seen one of those before, but luckily a bit of searching reveals one obscure source in these parts.

I've used wick exclusively after I'd seen a spring-loaded vacuum tool tear the copper right off the board once too many. I take it the bulb type is more gentle and easier to control than the spring? It certainly looks that way.

[ktesibios]

Another tip about using a vacuum desoldering tool:

When you've cleared the hole as well as you can, grab the protruding end of the component lead with a stout pair of tweezers and wiggle it. You may hear a faint "crack" as the lead breaks loose from residual solder still clinging to the inside of the hole. Once all the leads of a component wiggle freely you can try removing the component.

Grabbing the component with your needlenose and twisting or wiggling it a bit to break loose any "klingons" at the top side of the board before you try pulling the component out is also a good idea. I do this when desoldering DIP ICs.

When you remove the component inspect its leads to see if there are any bits of through-hole plating or top pad stuck to it. It is infinitely easier to repair damaged traces when you know they've been damaged before you install the replacement component than if you have to figure out why the -----thing doesn't work right after you've put in new parts to conceal the damage from you.

Also, keep a piece of solid wire handy (the leads of a 1/2W resistor work well) for poking out the hole in the tip of the desoldering iron so that it can breathe freely. And, as with all soldering tools, keep the tip clean and well-tinned, both to preserve it and promote good heat transfer from tool to workpiece.

[bpesta22]

Looks like I failed. Got the dc jack on all snug, and the voltmeter indicated it was working. Alas, though, it won't power up or charge the battery.

Eh well

:(

[bruto]

Looks like I failed. Got the dc jack on all snug, and the voltmeter indicated it was working. Alas, though, it won't power up or charge the battery.

Eh well

:(

YOu may have to look carefully at the circuit board, and see if the old jack tore a trace loose or cracked the board. It's not uncommon if it was stressed enough to break. A cracked circuit board can break the conductor in a way that is not always easy to spot without a magnifier.

If the case is still open, try with the voltmeter following the circuit board traces past where the jack is attached, and see if there's an interruption. If there is, it's sometimes possible to bridge the broken traces with a little piece of wire.