# Ultracapacitor-powered drill from NASA

I was reading my copy of NASA Tech Briefs and came across an article about a ultracapacitor-powered drill. Some of the claims were a little suspect. They said it could be charged in seconds instead of hours and that the charging would be way cheaper and more efficient than a battery charger. It should have been part of the article but there were no real performance figures on the charger. Similarly there were no real performance figures on the drill itself. I downloaded the pdf.

Sorry, no new information there either. All I wanted to know was how many screws the drill with the ultracapacitor can drive versus how many the original DeWalt can drive. That seems like a pretty important figure-of-merit to me. Now the claims about charging the drill in seconds with a very inexpensive charger did not ring true to me. A drill may put out a couple amps for maybe an hour. That is 2 amp hours. To charge it in 10 seconds you would need to supply 720 amperes. Now this is a capacitor, not a battery, so the concept of amp-hours is a little sketchy. The voltage constantly drops as the drill discharges, and that means you would need a boost converter or some other expensive device to keep the motor operating off a constant voltage. And how cheap is a charger that can supply 720 amperes? What will really happen is that IR losses somewhere will have to limit the charging current so now the charger is very inefficient. So you can put in a constant-current switcher to keep the current manageable and you are back at the same cost as a regular battery charger. It takes a long time to charge as well.

The drill has 6 one-hundred farad capacitors in series. This is so the 2.5-volt ultra capacitors can reach 15 volts. So the equivalent capacitance is 16.7 farads. Now the equation “farads = coulombs/volts” implies we will have 15×16.7 = 250 coulombs. Now a coulomb per second is equal to an ampere. So I was wrong above because I assumed the ultra caps would be equal in capacity to the battery. So how much current is needed to charge this particular gizmo in ten seconds? Well, 25 amperes. OK, how much power in the capacitors? Well, if you drew those constant 2 amperes out of the caps they would discharge linearly, which is nice for the math. Since an ampere is a coulomb per second you could draw the 2 amperes for 250/2 = 125 seconds, or about 2 minutes. Since the voltage is dropping linearly from 15 to 0 in that time, the average voltage is 7.5 and so the average power is 15 watts. Since a battery does not discharge to zero volts as it is used, it delivers far more power.

So lets think about this some more. A new DeWalt drill uses an 18-volt pack rated at 2.4 Amp-hours. That’s 144 amp-minutes and 8640 amp-seconds. If ni-cads had a flat discharge this would be 155,520 watt-seconds. So if we discharge it in the same 125 seconds that we used for the ultra caps we could get 1244 watts, almost 2 horsepower. But that is a little silly since we know it cannot discharge it that fast. So jump up a decade in time. The ultracapacitors now take 1250 seconds to discharge but draw 0.2 amperes and we get 7.5×0.2 = 3.25 watts average over the 20 minutes. Now if we discharged that battery over 1250 seconds we could still get 124 watts. Now that is optimistic since the discharge of the battery is not ruler-flat. Say it went to zero like the ultracapacitor. That still would be 124/2 = 62 watts, versus 3.5 watts for the ultracapacitors. So it is a pretty safe bet that the ultracap holds way less than one-tenth the energy of the ni-cad battery.

Are we sure I got this right? Lets look at the energy. Energy is (0.5)xCxV^2 so that means the ultracaps hold 1875 joules of energy when charged to 15 volts. A joule is a watt-second. So a perfect DeWalt battery with a flat discharge would be 8640 amp-seconds times 18 volts or 155,520 joules. But we know ni-cad has a far from flat discharge. So even if it droops to zero that is 77,760 joules, or 41 times the ultracapacitor. Hmm, this does look bad—almost unbelievable, so lets do another quick check. This website has a nice chart of energy density of battery chemistries. The chart says that ni-cad batteries have a energy density of 140,000 joules/kg. So go look up that battery I linked to above and how about that—it weighs 2.5 pounds, just over a kilogram. So the energy in the DeWalt battery is probably closer to 155k than 72k. The ultracapacitor is about 75 times worse than the battery.

This thing does not work, in any meaningful sense of the word. The charger is not cheaper, and the tool needs a boost converter, otherwise you can only discharge the caps from 15 to 10 volts and you get 3 times worse than the already miserable performance. But who cares because the thing won’t work for more than a few minutes. And this is work done by NASA rocket scientists.

So why would NASA do something so silly? That is explained at the bottom of the article. I quote: “Inquires concerning rights for the use of this invention should be addressed to NASA Glenn Research Center…” I always considered NASA to be a patent mill turning out obvious patents to tie up the licensing, but in this case the darn thing does not even work. We know the game though. It is a classic Lemelson move — write a broad sweeping patent to tie things up so in case any better ultracapacitors get invented, Black and Decker will have to go to NASA to license the obvious use of the technology. So this means your tax dollars are paying for a bunch of people to sit around all day and think up ways to tie up intellectual property so that when you buy products in the future there will be a hidden tax—the licensing fees to NASA, that you have to pay on top of all the other taxes you get to pay. I discussed despicable patent behavior in a previous blog. Also let me point out a patent that was granted to a guy that was just trying to prove how broken the system is. Paying NASA to work on stuff that will cost us more in the future is pretty stupid; you don’t have to be a rocket scientist to see that.

Fairchild acquisition of Xsens

The third anniversary of the passing of Bob Pease

DC to 110 GHz: Analog Devices and Hittite

It’s not too late to register for EELive in San Jose beginning March 31

Touchstone Semiconductor falls prey to hard times?

Student engineering labs revived with hands-on electronic kits

Accidental engineering: 10 mistakes turned into innovation

That 60W-equivalent LED: What you don’t know, and what no one will tell you…

6 famous people you may not know are engineers

DC distribution in your house and 42-V cars

10 tips for a successful engineering resume

The 5 greatest engineers of all time

Higgs Pt. 9: What makes King Carl XVI Gustaf think it’s the Higgs Boson?

10 things you may not know about Tesla

Analog Fundamentals: Instrumentation for impedance measurement

The ASUS EA-N66: A firmware upgrade doesn't translate into a fix

The connected car and congested traffic

Rule of Thumb #4: Skin depth of copper

10 things you may not know about Tesla

HP is incorporated, August 18, 1947

Debug ARM Cortex without a header on your PCB