Damn, finally broken

[elwing]

Damn, my pandora is finally broken, I'm not sure how it happened but I suspect my childs are partially responsible. the display backlight does not power anymore, the rest of the unit seems fine trough (if you except the power led that stays one no matter what but this posed no problem over the past few years).

It seems like a common case of the infamous coil getting loose. now I'm wondering what I should do, should I open it an take a look first or should I directly send my unit back to ED's? I'm unsure that I can do the soldering myself, I got to see the PCB directly to get a good catch on the pitch of the component, but opening the case and taking a look/picture should be no problem...

br,

 

[EvilDragon]

Well, if it's the coil, we can fix it for sure

 

[rygD]

Early reports suggest there will be massive flooding...oh, "damn" not "dam".

That sucks.  If there are other things that might not be right that you want looked at I would it to ED, but pop it open and take a look first.  There is no point in sending it away if you don't know what is wrong, and there is a chance you could fix it yourself.  I don't let anyone touch my Pandora anymore, just in case.  

I am still really confused about this coil issue.  Is there such a great chance of it falling off that I should open my Pandora and add hot glue as a preventive measure?

 

[elwing]

well, I suspect it might have been dropped or worst... you know kids... it survived for some years now...

that other issue of having the "power led always on no matter what" is not that bad, it occurred during the first few week after I got my pandora... no idea what the reason is, but as this led is more or less directly connected to the power management chip, I'm wondering If i might have a problem with it (ESD problems maybe?) repairing that seems quite hard. I'll take a look once it's open if I do not simply have a short there. in any case, that alone do not justify sending the pandora back... now if I find my coil too hard to solder back or if it's not the coil, I guess I'll probably ask ED to take a look at that too...

 

[levi]

FWIW, I had a lot of trouble desoldering my broken LCD connector, such that in the end I got it off partially through melting the solder and partially by breaking the component apart and ripping the bits off the board, so most of the traces to the part are now damaged, so it's pretty much only usable as a TV-only unit for the forseeable future.  I think the flux I bought turned out to be crud.

Now, resoldering anything is easier than unsoldering, and a two-lead coil is simpler than a ~40 lead LCD connector, but take my story as a warning - it may be better getting professional help from ED.

 

[elwing]

ok, I openned it, and the coil is rather loose it seems one pad is not soldered anymore... seems like an easy enough repair... but I've only got a 1mm tip with some 0.8 solder... seems rather chalenging...

There's a nice solder mask and enough clearence for an iron of that size... what I'm afraid is not being able to warm enough the coil and the PCB tab to get a nice solder...

edit: any advices there? I supose that attacking the problem with a tiny blob of solder at the end of the tip to make sure it contacts both pad and eventually add a bit of solder if needed once it flows is the way to go... When I learned electronics, before going to computer the majority of soldering was definetely still through hole, I've limited experience with SMD.

 

[levi]

Can you be sure the coil hasn't taken the pad with it?  If not, I'd tend to agree that it's achievable with something to take the solder on both sides.  You using leaded solder?  Shouldn't matter too much in that there shouldn't be too much lead free stuff already on that spot to alloy with it weirdly, but it will mean it melts better, and possibly mean there's no chance the other lead will melt off.

 

[elwing]

Without removing completely the coil, it seems hard to be sure, but it looks like the cut is between the coil pad and the pcb pad, and not directly on the component/PCB, I'll take a closer look... as for solder I have both, and I hate leadless one with a vangeance, that's stuff is only good to damage the tip of the iron even trough I'm using one regulated for leadless temperature... you recommand using leadless, right?

edit: picture time



sorry for all the color aberations, getting a picture without macro lenses and trough a table magnifying glass is hard... it's more visible on the right image but I can move the coil a little bit vertically, and the space that "grows" is between what is visibly a flat lead out of the coil and a metal pad on the PCB...

edit2: concerning that other problem... there's nothing to see there... according to the schematic, that led is connected with just one resistor to the TWL4030(which is in fact a TPS65950 on my PCB) ... which seems partially defect in my case... I might have missed a bridge somewhere, the resistors are so tiny and I do not know where is R110...

 

[rygD]

You say the power LED stays on.  Do you mean the green one on the right, or the red charging indicator LED on the left?  I can't remember if you can control the charging LED via software, but you can change the settings for the green power LED.  I have mine set at medium brightness when fully on, and very dim in power save.  Did you try that?  I am a bit curious about this, so keep us updated on both problems.

 

[elw3]

 I am a bit curious about this, so keep us updated on both problems.

Me too.

And yes the Battery led also is controllable. 

 

[elwing]

Hi rygD,

The green one, it's on even when the unit is powered off, as soon as there is a battery inserted, before the system had a chance to boot, also brightness settings have no effect on the led.

regarding the other problem it seems I mistaken some strange reflection on the side of the coil... both the coil pad are nicely soldered to the PCB, no issue there, it's just the coil winding that's not connected to the coil pad themself... I guess I'll send it for repair, finding a single coil like this will cost me an arm and a leg in shipping and handling, I can probably pay the shipping of the pandora to ED for the same price. and I'm not well equiped for soldering that, trying to just get a bit of solder there was realistical, desolder the old coil leads and resoldering nicely the coil is not really resonnable...

 

[rygD]

Those both suck, although I am a bit confused about the coil issue you described.  No need to explain further, the point is something is wrong and you don't have the means to repair it.  Best of luck to you, and I hope you get it back quickly.

Edit:  I also wanted to comment on surface mount vs. through hole.  I was never formally trained in electronics or how to solder, and my skills are enough for me to do what I need most of the time, but not that great.  I usually make things worse when messing with surface mounted components.  I also like through hole a bit better because it seems more durable.  My view of this is probably a bit wrong, since to me it is like using a screw or something vs glue.  More accurate would probably be to say it is like using a rivet or nut and bolt.  Just seems less likely to fall off or otherwise have the connection disturbed, though.

 

[levi]

Holes weaken PCBs.  SMD is probably more mechanically reliable than through hole especially if you tried to make anything as complex and small as a Pandora out of through hole (imagining such a thing were possible).

FWIW, my power light also comes on as soon as the battery is inserted.  I've not tried dimming mine, so I don't know at what point it would become dim, but it only goes out after I shut the machine down (without external power).

 

[rygD]

Holes weaken PCBs.  SMD is probably more mechanically reliable than through hole especially if you tried to make anything as complex and small as a Pandora out of through hole (imagining such a thing were possible).

I think in most situations if you are worried about the weakness caused by holes in your PCB your design is fundamentally wrong.  On a bare board I am pretty sure components are more likely to get ripped off than the board is to break, but it in a more sensible design where there is a case you still might have some shock or flex due to abuse.  I have never experienced a PCB breaking due to the holes, however I have seen one crack when it got bent in half (the case also broke on bothe side), and that was an extreme situation, and most, if not all, of the components were surface mount.  It was a cartridge that someone had sticking out of their console in the back seat of their car and something smashed or pushed on it.

Anyway, my point was merely my preference and admittedly probably incorrect feeling of less likelihood of components to have problems when hand assembled or modified by someone with mediocre soldering skills.  

 

[levi]

I haven't seen a fully through hole PCB for a computer since the mid 1980s.  Since then component density has shot up almost exponentially.  You might well be right to suggest it's easier for end users to hack and fix, but the fact of the matter is nothing as complex as the Pandora would exist without SMD.

I don't know why that inductor keeps popping off for people.  I can only assume stresses become so great that the case flexes and rides over the PCB edge and sweeps that component off the board - if it had been through-hole mounted it would probably stay on the board in that case, but almost certainly be destroyed instead.  Maybe it's no so dramatic and it's simply some sort of resonance that gets set up which shakes that component off the board, and it's only coincidental that it's on the edge of the board (though clearly that location is important, as IIRC there's another inductor of similar dimensions on the board that never seems to go anywhere).  I guess through hole would help in that case, but still to this date nobody understands how the board resonates, as it costs more to find out than the cost of repairing a dozen or so boards with this problem is.

I don't really know who I'm arguing with here, but sod it.

 

[rygD]

I don't really know who I'm arguing with here, but sod it.

That was how I felt when replying last time.  Someone needs to find a way to make electronics indestructable, small/dense, and easy for the unskilled to modify.  Or they could just create a replicator that we can feed our old stuff from which it will extract the raw materials and can reassemble the matter into whatever we need/want.

 

[bluedeer]

Once I took a sewing needle and used a piece of utility wire and bound it tightly to my 1mm solder tip, then I clipped the tip so it wasn't so fine a point, and it heated up enough to allow me to solder some extremely fine components but it's a ghetto workaround for not having a small enough tip. I have a 40W Soldering iron, though.

 

[elwing]

through hole component is a thing of the past and for good reason, it's huge, cost more to produce, I'm not sure for the PCB durability but it is definitely a problem for buried vias and multilayer PCB...

While I regret the simplicity of through hole components for repairs, they're gone for good reasons... what I'm not so sure is for the leadless solder change, but that's another story...

@bluedeer thanks for the tip, it's really hackish but assuming you can get solder to flow correctly on the needle that seems really good.

 

[elwing]

ok, just sent the pandora... do I need to send a mail with the tracking # or is ED already overflowing with email and don't care about tracking?

 

[EvilDragon]

Follow the tracking yourself and contact me when there are issues