Sunday, November 25, 2012

Speaking of Fixing Junk

I spent today catching up on some things I wanted to do on the holiday weekend.  It seems I never get everything done I want to do on any holiday, but I mixed this "work" with some goofing off and sleeping late, so I can live with not getting everything done. 

One of the things I wanted to do was take a look at a pair of cheapo iPhone chargers that have died on me.  Has everyone seen Dilbert's take?
As always, Dilbert is pretty much a documentary.  Both of these chargers plug into the car's 12V system (which is nominally 13.6V) and convert it to 5V DC at 1A - the USB charging "standard".  Both of them died while in my car, keeping my iPhone 4s topped up on the 3 hour trip over the river and through the woods to grandmother's house.  (I was using streaming radio, so the phone by itself may not last that long on its batteries).  Both died about halfway.  Not even two hours of operation.  Both gave off a tell-tale smell that all technicians know - burnt electronics.

Since we're talking about fixing things, I thought I'd share a look inside them.  They are typical of the unrepairable items you're surrounded by.  They are similar in design and use a single integrated power supply controller circuit.  Here's the iCrap charger with an obvious indication of being blown:
and here's a close up of the blown component:
Yes, that's a seriously overheated spot on the side of the IC. Pretty foregone conclusion that the part is toast. 

And that's the problem.  If the system is down, that's unrepairable.  If the infrastructure is still in place, I might be able to buy one of those parts.  Yeah, I'm guessing it would still be cheaper to buy a new charger.  If this design was made with discrete transistors, or older technology that's more widely available than these one-chip wonders, it would be easier to fix.  But then it would be more expensive to produce, and would be harder to fit in the nice little lighter plug package it's in.  They don't use these parts to be perverse, they do it to hit a price point people like. 

In this magnified view, you can read the part number for the IC: it's RT8250.  I've never heard of a company that uses RT in its numbers, but searching the web for that part number brings me to a free archive of datasheets and I have the datasheet downloaded.  Turns out RT is "Richtek Technology Corporation" - whoever that is.  Likewise, the other one was an RT part, and RT34063A.  The second one uses fewer parts but is otherwise very similar. 

Unrepairable?  How much of a scrounger are you?  While I couldn't make this work with the existing design, I could put in a different regulator and take advantage of the form factor that includes the USB plug and car lighter plug.  At this point, I'm probably just going to toss them , but it is worth thinking about what to do if there are no cheap/widely available replacements and how I'd generate something to do the same function. 


RegT said...

You also need more data than many of us possess, or would know where to find: is it enough to simply supply 5VDC at 1A, steady, or is the circuit built to prevent overcharging the battery? Does the amperage or voltage taper off as the battery reaches a full charge?

I'm definitely not electronics-savvy, and there may be other things that little circuit does that I'm not even aware of, but it would suck for me to build something to recharge an iPhone or an expensive laptop (my $2K MacBook Pro comes to mind, as I type on it), only to kill it by missing something obvious to an electron-wizard.

Graybeard said...

A reasonable question... these parts don't appear to be anything more sophisticated than a simple voltage regulator. That's all the chips say they are. They are switching regulators, because those can be more efficient (less heat in the charger) but still pretty simple.

They're not smart chargers, and maybe that's the problem.

I didn't point out that these are not Apple's chargers. They're the generic chargers you can find around.

RegT said...

BTW, that wasn't meant as criticism. I like the idea of work-arounds and DIY repairs/construction. It just struck me that as ignorant as I am about electronics, I'd have to be careful.

drjim said...

A 7805 three terminal regulator on a heatsink, and a couple of small ceramic disc caps.

What else can I help you with today?

Eric Wilner said...

Hmmm, I wonder how much market there'd be for a *reliable* car-to-5V adapter.
TI is making some spiffy tiny high-frequency switching regulators these days, e.g., TPS62163DSG: add two ceramic caps and a little bitty inductor, and you've got an efficient 12V (17V max) to 5V (1A max) converter. The package is conducive to good heatsinking, too, so keeping the junction temperature below the rated +125C should be dead easy.
Designing the board and getting it built are straightforward; now if someone has a source for the plastic housings and associated fiddly metal bits....

wheelgun said...

If the infrastructure has failed to the point where you can't get a replacement car-charger for your iPhone, I'm not sure that your most pressing need is going to be a car-charger for your iPhone.

This is different from "the government shut off the internet" and has moved over to "TEOTWAWKI." Even if it is going to be short-duration or localized. Think the Great East Japan Earthquake and Tsunami (with the following nuclear emergency) from a few years back.

Water. Food. Clothing. Shelter. Heat. Security. Medical supplies. Communications technology is pretty far down on that list. And I am a HAM radio operator.

And while I often buy the cheap generic gadget, there is a reason that some things cost more than others.... maybe you should spring for a decent charger. (Maybe the iPhone is a bigger current draw than the typical cell phone of a few years back.) I burned out a laptops power cord (120 AC to 19 DC) a few years back, and when I went to replace it found that HP had significantly increased the amperage rating on the replacement. They had used the cord from the older PC, but the new one - bigger screen, multi-core, more options - chewed thru more amps. My bet is the iPhone is similar.

And it doesn't look like the damn thing is fused. It should be fused. What if you were in the restroom at the gas station. Could have caused a fire.

Graybeard said...

You left off "That'll be $500 for my time, and $5000 for the design reviews". Or something like that.

Graybeard said...

This is different from "the government shut off the internet" and has moved over to "TEOTWAWKI." Even if it is going to be short-duration or localized. Think the Great East Japan Earthquake and Tsunami (with the following nuclear emergency) from a few years back. I'm thinking less like TEOTWAWKI and more like the collapse of Argentina, as Ferfal describes.

Most of the infrastructure is there, but it's spotty and less reliable. The roads have more potholes, power is on less than 24/7, phones work less of the time. Basically, everything just kinda everything sucks. Life is tougher. If you're lucky, you keep your job, but probably work fewer hours and your costs go up. Fewer things get transported across the country because the cost to transport is either rising too fast or unpredictable. So supplies of parts and things get worse, and the premium is on keeping what you have running.

BTW, this one is fused. The other one wasn't. I mean, besides the car's fuse. Not like things don't blow themselves up to save the fuses, though, right?

Graybeard said...

I was thinking of that. Unless you want to hollow out old ones and use their "fiddly bits", you're kind of stuck. Until 3D printing gets cheaper.

drjim said...

Naaaah.....I only charge friends for the cost of the parts I use!

Anonymous said...

Echoing drjim's suggestion: A good old LM350 3-terminal adjustable reg on a small heat sink is a good choice. It is even available in a di-electrically isolated package for added safety.

A lot of modern portable gear is fussy about the "USB" charging voltage range, and may not charge if that voltage is outside the 4.75V - 5.25V range. That includes brief moments of voltage ducking or surges.

Put a series protection diode in series with the 12V input, and bypass the reg's input and output. Trim the LM350 to about 5.2 V to compensate for charging cable IR drop and let 'er loose. Not only is it damn near indestructable, but there is no aggravating RFI and hash from a crappy noisy switch-mode reg to wipe out the reception on the car ratio's AM band...or your mobile HF rig.

Graybeard said...

The only problem I have with that approach is that the power dissipation in a linear regulator is the voltage drop times the current, in this case about 8W. I don't think that works well in a closed plastic package. The heat sink needs to be connected to something, or have air blowing on it.

You need a good way of getting the heat out of there. Switching regulators can be noisy, but every radio I've worked on in the last 30 years has had one inside. They can be tamed. If I can get 80% efficiency, pretty reasonable in a switcher, that means much less dissipation in the circuit. If I'm running 5W, to the battery, I'll dissipate only 1 1/4W. Sounds better.

Graybeard said...

everything just kinda everything sucks obviously should have been "everything just kinda sucks"

Anonymous said...

If you want to make it REALLY simple and still use a switcher, Linear Tech and TI make very nice little micro-modules that contain the inductors, switch transistors, etc. all in one nice, easy to heatsink package. Something like this would work really well.

The problem with that is that it's much more expensive than a discrete solution, but it's a trade of part cost vs. design time and convienence.