Bicycle iPhone and USB Charging – Battery Output

To take advantage of the power stored in my batteries, I need to regulate the voltage they produce so that it’s a stable 5 volts. That means I’ll need a voltage regulator circuit.

Now, there are schematics for these all over the web, and fancy products that combine regulators with programmable switches and blinking patterns that are designed to power lights… But all I need is a steady 5 volts, to hand off to a series of USB ports. I may also decide to add regulation for other voltages, but right now 5 volts is my target.

Let me introduce to you The Superest Regulator Ever, the Texas Instruments PTH08000W.

Yes, it is quite small. And according to the docs and the graphs in the docs, it has over 90% efficiency with the voltages I’m feeding it. There are other regulators out there … many others … but few of them are as efficient as this, as small, and as self-contained.

The simplest way to regulate voltage is to just take an input voltage and throw away everything above the target. Circuits that do this are called "linear regulators", or "voltage dividers" … and they suck. Modern power conversion has moved waaaay beyond this. Basically, what a switching regulator like the PTH08000W does is decompose the incoming current into a collection of different waveforms … and then reassemble the waveforms into current at a different voltage! It’s some serious Weird Science.

You set the desired output voltage on the PTH08000W by attaching a single resistor to it. Aside from the resistor, the only other component it requires is a capacitor to provide filtering on the input side. In my application, the capacitor is probably not even necessary, since I’m hooking straight up to a high-output battery.

So what’s it cost to get ahold of this fine piece of electronic equipment? Nothing. Go to the Texas Instruments Sample Ordering section, make an account, and drop one in your basket. You will receive a regulator in the mail about a week later, free of charge, in a little bag like this one:

What, you thought you were gonna have to design a regulator? Or go spec out a laundry list of components and build one? That is so last-century, yo. Texas Instruments has done that for you. Free of charge.

One Resistor

Once I received mine, the first thing I did was hop over to the local electronics shop to find a single 348 Ohm resistor. The closest they had in stock was a 380 Ohm, which would cause the regulator to output 4.95 volts instead of the desired 5 volts. I bought it anyway, reasoning that 4.95 volts is close enough that a USB device would recognize it.

While I was browsing around the store I went rummaging through their clearance shelves, and discovered a party-size bag of 100µF capacitors for five bucks. Yeah, this is the sort of thing you just run across when you live in the Silicon Valley. I grabbed it and set aside a handful of capacitors for this project. My household uses the rest as chips for poker games.

Solder And Test

My approach of soldering the resistor directly across pins 4 and 5 of the regulator is not the standard procedure. Usually you plant this thing in the middle of a larger circuit board and then locate the resistor somewhere else. But I’m a MAVERICK! And I like to MAVERICK IT UP!!

Once the resistor was attached I set the board loose from the vice grips so I could clip off the excess wire, and then I taped an input capacitor on and soldered it between pins 1 and 4 (like the manual suggests) for some input voltage filtering. In retrospect, totally unnecessary. And I ended up replacing this capacitor anyway, since it made the regulator too thick to fit in the case I wanted to build.

Finally I soldered three wires to it, on pins 1, 4, and 5, then hooked up my batteries and measured the output on my pathetic, cheap-ass voltmeter. The regulator produced 4.96 volts, just like it’s supposed to. FLAWLESS VICTORY, as the kids say.

It looks hilariously small next to those massive batteries.

Discharge Graphs

Now that we’ve got power flowing into the batteries and flowing out, let’s stop for a second and consider will happen over time.

As I mentioned before, the voltage that batteries put out is not constant. Over the duration of one "cycle" of the battery from full to empty, the voltage will "sag" from slightly above the standard cell voltage to some amount well below it. How far the voltage sags depends on how quickly you are draining the battery. Imagine an athlete running in a long-distance race: If she runs too fast, she’ll become out-of-breath partway through, and cover the remaining distance much slower. A battery with sagging voltage is like an out-of-breath runner … except there’s no way for a battery to catch it’s breath again. You’ll have to charge it until the voltage comes back up.

The minimum required input voltage for my regulator is "output voltage plus 1.1 volts", or 6.1 volts. Since I have two cells in series, that means the converter will stop converting power when “voltage sag” causes each cell to drop below 3.05 volts. Even if the batteries are still half full at that point, they won’t be able to drive the regulator, and I won’t be able to charge USB devices.

So how much time have I got before this happens? How much capacity can I use up before voltage sag stops me?

A fellow from the Zero Emission Vehicles Australia website has conducted some voltage sag tests with a collection of LiFePO4 batteries from various manufacturers. My specific battery is not mentioned in the tests, but the graphs on the page paint a good picture of what I can expect to see. Since the discharge rate I’m using is so very low (about 1 amp), I can expect the voltage to stay above 3.0 volts for at least 70% of the cycle time. At the 30% full mark, each battery will drop below 3.0 volts and my regulator should shut off.

I say "should" because during my testing I discovered that the regulator output 5 volts all the way down to the point where the batteries themselves could no longer push 5 volts. In other words, each of my two cells dropped all the way to 2.5 volts before the regulator quit. According to the discharge graphs, that’s easily the entire cycle life of the battery. So I am apparently able to use the full capacity of these batteries when powering USB devices. That’s good news!

Now it’s time to build something that will provide the USB connections I need.

To USB Connections …