Reply To: Fusion Oil

[mchargue]

Aeronaut wrote: Thanx for the research, Pat. I’m impressed that it wasn’t behind a paywall. Being a State fan, this could help me rethink my attitude about U of M 😉

I have two questions concerning the laser power requirements: How much power, size, and unit cost per switch, (very rough guesstimate, of course) and
Might it be more cost-effective to use a single large laser?

Cost… Um… Bigger than a bread-box?

Not helpful, I know, but I haven’t even done a back-of-the-napkin estimate. As well, there’s the added complication that I don’t work in manufacturing, so it’s hard to know what materials’ cost would be. My thinking is that switches now represent the biggest issue to date – both for experimentation, and for robust operation of a fielded system – and that the problem must be solved.

Whatever solution comes from this is going to be more expensive than plunking down $50 for a set of NGK spark-plugs, but the solution should resolve the switch-to-switch jitter issues, and guarantee that all switches fire.

LASER selection & power.

Based on what I’ve read, the best choice for a switch gas is a noble gas with a low atomic weight (i.e. He) in order to make the switch rise-time fast. I would think that this will also make the LASER more effective, as there will be a single, dominant, frequency for gas excitation.

That LASER power – power/unit area. (W/cm^2) – that would needed to pre-ionize the switch gas should be available from the paper(s). Multiply that by the number of switches that you want to fire, and you that will give you a LASER power value to shoot for.

I would run all switches off of the same LASER, as that should help ease switch-to-switch jitter. Doing this, however, removes the ability to dynamically ‘tune’ the firing pattern of the switches through any other means than by lengthening the LASER light-path distance relative to other switches.

Although…

One neat possibility, one that would decrease complexity by removing the LASER path, would be the use of individual LASER diodes (solid-state LASERs) on each switch. Operated in pulse mode, you can get several watts out of ’em, but frequency selection may be an issue.
http://www.repairfaq.org/sam/laserdio.htm#diocss7b

This would, though, put the ability to tune the timing of the switches (relative to one another) back in the realm of electronics. (or maybe lengths of fiber-optic lines) Care would have to be taken, though, to try to minimize the differences between LASER diodes. (rise-time, mostly)

Here are a few high-power LASER diodes.
http://www.sony.net/Products/SC-HP/datasheet/90216/data/a6804813.pdf (20W)
http://www.sony.net/Products/SC-HP/datasheet/90216/data/a6804814.pdf (40W)
http://www.sony.net/Products/SC-HP/datasheet/90216/data/a6810228.pdf (60W)

http://www.lumics.com/Multimode-Diode-Laser.361.0.html

Anyway, the point is that there seem to be high-power, solid-state lasers available (you can get a 2W version on ebay!) that should be able to drive pre-ionization in a gas switch using pulse-mode operation. If I were engineering this I would:

Select working gases.
Determine excitation frequencies for each.
Determine pre-ionization power requirements.
Talk to a solid-state LASER diode manufacturer, and design the system with them. (use expert help)
Make & test one.
Clone the working version.

Tougher than using spark-plugs, true, but it should carry you through experiments – hopefully to a working system.

Pat