That's a Big Transistor

re: That's a Big Transistor

Jonny D — Sat, 22 Jan 2005 02:19:00 GMT

well there is on other huge factor that has been ignored.

basically there will be inductance in the object traveling through the coil. so whatever you are trying to send up to space will heat up.
Sure you could try to use no-magnetic materials, but then you can't accelerate it. So. you are bound to melt anything that you try to send up. and further more, the magnetic fields will rip it to pieces...

this really might be the most entertaining way to quite literally burn money.

re: That's a Big Transistor

G. Man — Sat, 22 Jan 2005 02:35:00 GMT

Awesome! Man what the heck did I read before blogs?

re: That's a Big Transistor

Robert Hahn — Sat, 22 Jan 2005 14:36:00 GMT

Fascinating writeup, Eric. You got me wondering though - if 1000kg is going to be a problem, what if we scale the payload *down*?

A lot of progress is being made in the robotics field. I would suggest the following: You want an object weighing 1000kg in orbit. Design the object into 500 2kg objects with enough smarts (and a bit of power) to track and locate their 'buddy parts' and self-assemble.

Naturally, I would expect that a design like that may result in an object that would weigh more than 1000kg, because there will be overhead, but how much really? Artificial intelligence code optimized for self-assembly won't be needed after - so the RAM can be flashed with the mission specs to make that object useful. 500 small powersources might be a nice way to design this object because it would be more fault tolerant. Or maybe the power source for each object can be detached and collected in one place so that orbital changes can be made as the mass of the object increases. Or, have the object self-assemble on the moon, then blast off from there into a stable earth orbit.

it's easy for me to say this stuff - I'm not doing the math ;)

good article though. really fun!

re: That's a Big Transistor

Ian Griffiths — Sat, 22 Jan 2005 19:42:00 GMT

"Unfortunately, electrical resistance is equal to current divided by voltage"

Err... That's not how Ohm saw it. Ohm's law states that V = IR. If you rearrange that to express R in terms of voltage and current you get:

R = V/I

So that would be voltage divided by current, not vice versa. So given what you go on to say:

"That's why we have high-voltage power lines to deliver power"

given that resistance is apparently voltage divided by current, wouldn't a high voltage system be worse because there is more resistance?

Actually, no - stepping the voltage up or down doesn't in fact change the resistance of electrical cables at all. The resistance remains constant. (Well, it changes according to temperature, but it's not directly dependent on the voltage or current.)

So in the equation that relates current, voltage, and resistance, resistance is effectively a constant when dealing with power cables. All you do by upping the voltage is to reduce the current - the resistance remains the same.

So why do we use high voltage power cables? The answer lies in *power*. The goal is to minimize the loss of power. Here's the equation that lets us calculate the power disippated:

Power = VI

The the amount of power dissipated by an elecric power cabls is equal to current multiplied by the voltage drop across that cable. (Note that this is *not* the voltage of the whole power supply. It's just the amount by which the voltage drops across the cable - this is the power loss in the cable we are remembering.)

So how do we work out what the voltage drop is actually going to be across a given cable? Well for that we use Ohm's law, which remember is:

V = IR

So the voltage drop will be equal to the cable's resistance multiplied by the flowing current.

By substituting in this equation for V into the equation for power loss, we can work out the power loss as follows:

Power = V * I
= (I*R) * I
= I*I*R

So the power loss in an electric cable is equal to the square of the current multiplied by its resistance.

Notice that the voltage doesn't figure anywhere in this equation. Only the current. Of course voltage and current are related, in that in order to send a given amount of power down a cable, Voltage*Current must be equal to the power. But you are at liberty to trade one off against the other - if you need a kilowatt, you could have 1V at 1000 amps, or 1000V at 1 amp.

And bearing in mind that the equation for power loss in a cable only features current, and not voltage, one of these looks more attractive than the other. With 1000V at 1 amp, you've got a power loss in the cable of I*I*R, thats 1*1*R, or R. I.e. the power loss will be whatever the resistance is. But with a volate of 100V and 10A, the power loss is now 10*10*R, i.e. 100 times greater!

The resistance remains constant throughout these calculations. It's just not right to say that resistance drops as you increase voltage. It doesn't. The power loss through heat dissipation in the cable drops, but that's because this loss is a function of current and not voltage.

(Of course the fact that cable heats up as it dissipates heat and that changes the temperature complicates matters a little. R isn't really a constant. But even taking this into account, high voltage still ends up looking attractive.)

re: That's a Big Transistor

Eric Lippert — Sun, 23 Jan 2005 23:47:00 GMT

You're absolutely right -- I totally screwed up that explanation. I'm not sure what I was thinking when I wrote that. I'll make the offending paragraph more clear. Thanks!

re: That's a Big Transistor

Enigma2e — Mon, 24 Jan 2005 06:45:00 GMT

Back in high school, my senior year, i did this as a science fair project. I actually went through several designs before i settled on one. I still havent tested the latest incarnation for one reason alone. The switching transistors i use are 7 bucks a-piece and rated at 70 Amps each. However, last time i 'tested' it with only 3 of the 6 stages operational. I ended up with just a metal backing of the transistor + a heatsink. the ceramic and chip had literally disintegrated themselves. And the only thing i was using, as i couldnt find any caps, was a 12v deep cycle marine battery. I have pictures of my experiements here http://archives.rivin.net/files/images/coilgun/ hosted on my {slow} DSL server. Username is "coilgun", Password is "coilgun", if prompted.

re: That's a Big Transistor

Scott — Mon, 24 Jan 2005 14:51:00 GMT

There's also a problem that has nothing to do with electronics. Inertia. You're taking an object at rest (an object with alot of mass) and accelerating it at an incredible rate. You're going to need a large platform capable of holding its own against the (for lack of a better term) "back energy" in order to achieve the maximum forward energy.

The platform size is already going to be required to be large and strong enough to support the weight of 10k meters of coils, the electronics and cabling, potentially a lift system for maintenence personnel.

It could be an interesting sight though. Imagine the cabling running from the ground to the upper levels of the coil gun, like you would see on a cellular tower. Now imagine firing the gun, the inertia actually compressing the structure toward the ground and those cables going slack. Then going taut again when the projectile is released (hopefully!)

re: That's a Big Transistor

Eric Lippert — Mon, 24 Jan 2005 16:37:00 GMT

Since it would have to be kilometres long, you'd want to build it up the side of a mountain, going from horizontal to vertical gradually. Of course, then there are all the problems associated with making an object with lots of inertia curve!

I think one of Heinlein's novels featured a coil gun up the side of Pike's Peak for launching stuff into space.

re: That's a Big Transistor

Peter Torr — Mon, 24 Jan 2005 17:16:00 GMT

Eddy's in the current again! Somebody get him out!!

Plus strapping stuff to objects and blowing it up in a sem-controlled fashion is much cooler than having silly electrons do all the work... imagine watching "Armageddon" or "Apollo 13" without all the cool take-off effects.

On second thoughts, you don't have to imagine anything at all -- just watch "The Core" and realise how crap the world would be without rocket take-off sequences in blockbuster Hollywood movies about deep-sea oil drillers saving the world from ulitamte destruction.

What would Jerry Bruckheimer do?!?

re: That's a Big Transistor

Patrick — Mon, 24 Jan 2005 19:39:00 GMT

"Until there are radical new advances in material science and power management we're going to be stuck with strapping big tanks of liquid oxygen onto the sides of projectiles if we want to get them into space."

The conclusion does not follow from the argument. You've demonstrated that coil guns would require radical material advances (radical, but hardly magical).

You haven't demonstrated that all alternatives are infeasible. There may yet be a non-liquid oxygen solution with current materials and power management technology. Perhaps solid fuel rockets, or propulsion by ground-based lasers. Who knows.

re: That's a Big Transistor

Zac — Mon, 24 Jan 2005 21:19:00 GMT

Eric, your Mr. Fusion arithmetic is bogus. If you listen closely, the flux capacitor clearly required 1.21 "jigga"-watts. Clearly Doc had some sort of communication back from the future related to the power of Jay-Z, but we don't really know what those units scale to.

Honestly, such a lack of rigour. Tsk tsk.

re: That's a Big Transistor

Lance Fisher — Mon, 24 Jan 2005 22:15:00 GMT

Like Patrick said coil guns aren't the only way to convert electricity into kinetic energy in order to move an object to space. You could use a giant elevator, which also have their problems, but at least you can convert the energy slowly. Check this out: http://en.wikipedia.org/wiki/Space_elevator

re: That's a Big Transistor

Eric Lippert — Mon, 24 Jan 2005 22:22:00 GMT

As the article notes, a space elevator has to be made out of fictionite. Again, this is a technology which is predicated on radical advances in material science.

re: That's a Big Transistor

Kalon Jelen — Tue, 25 Jan 2005 04:55:00 GMT

Nice reference of Ghostbusters in the title, BTW. Great article too.

Higher voltage = less power loss

Martin — Tue, 25 Jan 2005 09:20:00 GMT

Has anyone worked out how many megawatts of power would be saved if the USA converted to the same standard as the rest of the world and used 230V as its voltage for domestic electrical supplies? 3kW electric heaters would only use 13A. Ohmic losses would be halved.
Or you could halve the amount of copper used in the wiring...

re: That's a Big Transistor

Steve — Tue, 25 Jan 2005 14:16:00 GMT

I knew a 1.21 gigawatts reference would be in this article. Eric, you have a lot more time on your hands my friend.

re: High voltate = less power loss

Carmen Crincoli — Tue, 25 Jan 2005 17:50:00 GMT

Martin,

That's one of those pie-in-the-sky savings though. The energy, (both human and electrical) that it would take to get the entire country to switch over would be immense, costly, and most likely tragic for the economy.

Microsoft isn't the only entity not interested in breaking backward compatibility just for the sake of new design or efficiency. :)

re: That's a Big Transistor

Norman Diamond — Wed, 26 Jan 2005 03:19:00 GMT

1/25/2005 1:20 AM Martin

> if the USA converted to the same standard as
> the rest of the world

and had a mix of 100, 120, 220, and 240, matching the standard of the rest of the world?

In a house in a medium-sized Indonesian city, when a refrigerator was moved from one room to another, they had to buy a transformer.

> used 230V as its voltage for domestic
> electrical supplies?

Yes, that part would be beneficial, except for the minor issue that Carmen Crinoli pointed out. Maybe the whole world should move to a metric 1kV.

re: That's a Big Transistor

Simon — Wed, 26 Jan 2005 15:46:00 GMT

Excellent discussion. Got me completely in the mood to help my son with his physics homework.

I'm glad someone remembered Eddy!

I'm sure it won't be long before one of the eggs at MIT or O&C discover a new form of energy that can aid propulsion.

It just might take a few years to get past all the Health and Safety standards and application forms of our respective nations, before you can buy a domestic version of it in WalMart or Woolworths.

re: That's a Big Transistor

Phylyp — Fri, 28 Jan 2005 12:12:00 GMT

Eric,
You missed your vocation - teaching. If you had been my physics teacher, I'd have been inspired.

Reading through this, e = 1/2 * mv^2 seemed familiar. Then it hit me: I'd learned this so *many* times in drab proofs, this was a thrilling account of its actual usage!

re: That's a Big Transistor

Eric Lippert — Fri, 28 Jan 2005 16:23:00 GMT

Thanks, that's a very nice thing to say.

I've often considered whether I'd enjoy teaching, and the conclusion that I've reached is that trying to teach people who didn't want to be there would be intensely frustrating for me. I don't think I could be a high school teacher, for instance.

I suppose I could teach college courses though. Perhaps after I retire from Microsoft I'll spend my time teaching classes and writing books! :)

re: That's a Big Transistor

Bramster — Thu, 03 Feb 2005 20:10:00 GMT

Great discussion! And there are other considerations.

Why not a mixture of technologies. It seems to me that it takes at least a couple of seconds for the shuttle to clear the launch tower. . . (I'll try to time it the next launch). How much fuel is used to lift all that fuel past the end of the tower?

How about an elevator with the mother of all counterweights for the first part of the journey up the inside of mountain, a coil gun for the rest of the time inside the mountain, and then conventional chemical rockets for the rest of the trip to orbit?

Maybe even a super laser to pump energy into an ablative stage before the Goddards kick in.

Recall that Armstrong & Aldrin's Lunar Excursion Module had enough fuel to both slow from lunar orbit to a dead stop, and then return to the same orbit.

Now, if Bill Gates really wanted to leave a legacy. . .

re: That's a Big Transistor

Rolf Viehmann — Wed, 30 Nov 2005 22:39:28 GMT

There's one other thing to consider: The sattelite you may want to launch has to include quite much fancy electronics to do it's work, but if you have coils which induce a high current into the electronics (and you would need enormous shielding to prevent this), how do you manage to NOT fry the electronics? How could such a great shielding be made?

re: That's a Big Transistor

Tanveer Badar — Fri, 25 May 2007 13:42:26 GMT

And just not forget how particle colliders work these days. I am not talking about child stuff of betatron and synchronotron.

Consider Stanford Linear Collider (SLC) and CERN's proton-antiproton collider. They work exactly like having multiple electromagnets pulling the particles and imparting energy to them during the process.

The only difference is that the payload is around 10^31 kg this time and travels an equivalent distance to a couple of earth's circumference in case of CERN.