I have a Kurt Kinetic trainer that I use for most of my interval training. I don't have the funds for a power meter but I noticed they have a power computer which convert speed to watts. I've know watts are more accurate than heart rate because the heart doesn't give immediate feed back. The pdf file they show on the site has hugh jumps in watts from one speed to another. So my question is if anyone is using this or something like it in their training?

In principle something like a kurt kinetic or cycleops - basically one of the better trainers has a fixed relationship between speed and power. Allegedly these units are reproducible enough that they use a basic equation and its good for all units. However, a better idea, and one of which I am particularly fond, would be to get some friendly racer with a power meter and ask him to calibrate it for you. I've been thinking of doing it, but havent had the spare time to pursue it.
If you put a speed sensor on his back wheel and just take a note of the speed versus the power reading, then you can make an easy look up table or curve on a chart. Wait till the trainer is warmed up as the fluid trainers seem to change a bit at first.
then all you need is a speed sensor on your back wheel and you can easily do direct power measurements.

Calibrating it with a friend's powertap is a good idea. I've seen others here remark that the KK equation is within +/- 3% or so, but not a bad idea to confirm if possible. I'm not so worried; speed here should be constant after the unit is warm, unlike the road. So higher speed means higher power.

As for the large jumps in power from one speed to another, that's the way power works. It isn't linear. If you look at the KK equation, you'll see that it has exponents. These are a little steep at higher powers, though.

The general equation for air resistance is: 0.5 * air density* drag coefficient* cross-sectional area* (velocity squared). That velocity-squared term is the one that'll change the most on the road.

Basically when you're training to get faster, you have to increase your power by a lot to go just a bit faster. To increase your speed by, say 5% (20 to 21 mph), you have to increase power by about 10%...to get to 23 mph (15% increase), it's about a third. 20017.622518.625019.627520.430021.33252235022.737523.440024.142524.745025.347525.850026.452526.955027.4 I used Excel and the KK equation to make a little power chart, to give me a rough idea of what I'm doing (Watts vs. mph). This also illustrates how steep the resistance gets at higher levels. I couldn't find the numbers that KK uses for the rider weight, etc (i know the slope is 0.1), but analytic cycling puts 500 W at about 30 mph. They do seem to diverge. Not that you'll probably be spending that much time at 500 W, but the power needed here is probably higher than on the road.

If you stay in the same gear and maintain your cadence, that's a steady effort. By knowing your gear ratio and cadence, you can calculate your speed (see the Holy Texts: http://www.sheldonbrown.com/gears/). By knowing your speed, you can use Kurt's formula to calculate your watts.

Even if you don't have a cadence meter, you can just count from time to time with a watch during your longer intervals. It's not a bad idea to develop a feel for a steady cadence.