Tag Archives: battery

Respect Your Batteries

Seriously.  These seemingly innocuous little things can kill you if you mistreat them.  So be kind to your batteries, alright?  Don’t overcharge them, or short them out, or put them into hot ovens, or stab them, or shoot them with rifles, etc, etc.

Just an ordinary SLAB...

You see, I’ve had more than one close call with a battery in my day.  Like the time I was working on a regenerative motor drive circuit and suddenly had a gate driver IC turn into a dead short across 24V of lead-acid batteries.  Have you ever seen a 14-DIP glow red hot?  I have.  The aftermath looks something like this:

A Very-Failed Gate Driver IC

And then there was the time I was trimming some wires on the end of a string of twenty LiPo cells.  Somehow, I managed to temporarily short the two ends of the string with the diagonal cutters I was using.  End result?  The enormous surge of current actually  vaporized a chunk of metal from the tip of the cutters.  Oops.

Of course, misuse isn’t the only cause of battery misbehavior.  Overuse can cause problems as well.  Just yesterday, my UPS shut down and started beeping incessantly.  Now, in its defense, I’ve been using it almost non-stop for the past six years and have never replaced its battery.  In the last year, it’s probably warned me two or three times that it needed a new battery, but until yesterday, I’ve ignored it.  For one thing, I operate it at only about 10% of its full-load rating.  For another, it’s no longer supporting any equipment that needs to remain online during a power failure.  So I figured, eh, I don’t mind if it only lasts a few minutes during an outage – I’ll grab a new battery for it with my next Digi-Key order.

Well, when the unit finally shut down for good yesterday, I figured I’d better go ahead and pull out the battery.  Once I had the cover removed, however, I discovered three causes for concern.  First, the battery was unusually hot (about 130F according to my IR thermometer).  Second, it had swolen so much that I could no longer slide it out of my UPS.  Third, it had split in four places on its bottom, as you can see here:

Battery failure...

Battery failure (closeup) ...Yea, that looks pretty bad, right?  Electrolyte had actually started leaking out of one of the splits and had begun a bit of rusting on the battery housing cover.  You can see a close-up shot of this to the right.  Perhaps even more unnerving, however, is that when rotated, you could hear bits and pieces of something rattling around inside the battery.  That can’t be good.  Batteries aren’t supposed to have things rattling around inside of them…

So presented with this new problem, I grabbed my safety glasses and moved everything into the kitchen.  Why?  Well, I figured that if the thing started smoking or something while I was attempting to remove it from the UPS, I’d just toss the whole works into the oven (which was of course turned off), close the door, and grab the nearby fire extinguisher.  Fortunately, I didn’t need any of that.  After a bit of sweat and a lot of not-so-gentle prying, the battery popped lose and started to slowly cool.

Here’s something I don’t understand though: why did this sealed lead-acid battery split open at the bottom?  Isn’t that what the circular cuts on the top of the battery are designed for?  To vent any pressure accumulating inside the cells?  And yet, if you scroll back up to the top of the page and take a look at the first image shown, it sure doesn’t look like any of those vents have even cracked.  I don’t think I’ll be buying my replacement battery from YUASA though, I can tell you that…

And now, to wrap things up, more examples of why you should be nice to your batteries (especially LiPo cells… these things are psychotic):


By the way, Lithium battery chemistry has greatly improved over the years.  The risks of such spectacular fires are now quite low.  But please, don’t push your luck.  Respect.

Iontophoresis: Pharmacology, Meet Electrical Engineering

Iontophoresis Patch Dispensing DexamethasoneI don’t mean to brag, but I have a really great mom.  Sadly though, last week she twisted her ankle quite badly.  According to her physical therapist, the injury caused damage to one of the nerves in her leg.  So, in addition to the more traditional PT remedies (cold and hot packs, stretching exercises, etc), he prescribed a disposable iontophoresis patch, pictured here, for the administration of the anti-inflammatory drug dexamethasone.  And the day after she was done with it, my mother, knowing of my fascination with electronics, mailed it to me for disassembly.

So just what is iontophoresis?  (And why isn’t it spelled ionophoresis?  That letter “t” really feels out of place to me…)  Well, it is a method for drug delivery which utilizes direct electrical current to “push” charged ions through a patient’s skin – no needle required.  It works based on the simple principle that like charges repel and opposite charges attract.  So, if we have a drug which can be ionized – either positively or negatively charged – we can apply a like charge to the delivery electrode, and an opposing charge to the skin itself.  This difference in electrical potential (aka voltage) will cause the charged drug ions to flow into the skin.  Cool, right?

Empi Iontophoresis Action Patch: Underside

In practice, drug delivery using iontophoresis is quite simple.  First, a solution of the drug to be dispensed is applied to the appropriate electrode (either positive or negative, depending on the charge of the ionized drug).  Then, both the delivery electrode and a second, oppositely-charged electrode (necessary to complete the electrical circuit) are applied to the patient’s skin.  Finally, a small electrical current, usually no more than a few milliamps, is applied between the two electrodes for a set amount of time.  Dosages are specified in mA-min: the number of milliamps applied, multiplied by the treatment duration in minutes.  The battery-powered patch shown above is rated for 80 mA-min when used for 3 hours.  Thus, it delivers an average current of 80/(3*60) = 0.444mA.

Alright, let’s get started with the teardown.  First off, I removed the labeled cover from the top electrode, which was held on by only a fairly mild adhesive:

Iontophoresis Patch Electronics

As you might have guessed, I found a whole slew of 1.5V alkaline coin-cell batteries, linked together in series to produce 10.5VDC.  Towards the bottom-middle of the battery compartment, you’ll notice a small bracket-shaped device.  This is actually a spring-loaded switch which is held open by a removable tab (no longer present).  This switch is necessary to prevent the batteries from discharging until the patch is applied and the tab removed.  The underside of the battery compartment doubles as the positive electrode, and is attached to the conductive gelatinous pad shown in a previous image.

Before going any further, I was curious to see if I could bring the discharged device back to life once more.  So, I punched a couple of small holes in the battery compartment and inserted my own wires.  Those wires were connected back to a DC supply set to 10.5V:

Repowering the Iontophoresis Patch

Indeed, the LED (what Empi’s marketing department calls the “Smart Light”) illuminated once more!  Now, I’d already suspected this to be a fairly simple system – basically a set of batteries in series with a resistor – but I decided to take some current measurements between the two electrodes to confirm this.  To do so, I grabbed a bit of aluminum foil, some wires, a section of plastic wrap, and assembled it like so:

Measurement Connections

Not the prettiest thing, granted, but it did the job.  The adhesive backing on the patch held the plastic wrap tightly in place, pressing the aluminum foil snugly against the two electrodes.  I first inserted a 50k resistor in series with the electrodes, and measured a current of 143uA.  I then lowered the resistance to 25k, and recorded a current of 251uA.  Finally, I shorted out the two electrodes and saw a current of 1240uA.  Clearly, there was nothing here performing current control.  The amount of current this device delivers depends solely on the resistance of the patent’s skin.  But, that’s probably fine.  I mean, does skin conductance really vary much from patient to patient?

With that test complete, I removed my wires and proceeded to extract the opposite end of the circuitry from the drug delivery electrode (it was again held in by mild adhesive):

Iontophoresis Patch: Complete Electronics

Nothing terribly exciting here, honestly – a flexible circuit board with five components, seven coin cell batteries, and a switch.  The whole thing was quite easily traced:

CircuitNow I’m guessing, but I’ll bet that extra diode D1 is there to cause the LED to shut off sooner than it would with just a series resistance.  The forward voltage across D1, about 0.7V, requires that much more voltage from the batteries for the LED to be illuminated.  The purpose of R3, I’m guessing, is to prevent too much current from being delivered to the patient.  And the function of R1 can only be to discharge the batteries more rapidly, which I’ll wager is done to guard against an excessively long dosing time.

So the question is, does iontophoresis actually work?  Well I’m afraid the verdict’s still out on this one.  One study indicates a measurable difference in the delivery of the drug dexamethasone, but does not demonstrate a tangible benefit to the patient.  Overall, the results are mixed, and iontophoresis is still considered experimental by insurance companies (who won’t typically pay for it).  Take a look at this great article for more info.

Oh and as for my mom, she’s doing just fine, although she isn’t quite sure whether or not to attribute that to iontophoresis.  With so many other treatments being employed in parallel, it’s hard to tell what worked and what didn’t.  But at a cost of $12, the patch she used was probably worth the try.  I certainly got a kick of out it anyway.