I want to design a boost converter that can supply 50V@2A assuming i can get at least 70% efficiency from my 24V, 6.5A supply.

I've done a little bit of research on them, and I tried building a simple converter but my mosfet was getting very hot, very quickly with a 1M load coming from my multimeter.

this is the website I used
http://www.ladyada.net/library/diyboostcalc.html

Is there a good website that can help me design the converter? (wikipedia wasn't much of a help)

Also, can I make it adjustable from maybe 1.25-50V? Sorry for my ignorance, I recently discovered how much better these dc-dc converters are over linear regulators.

 

Try going to the National Semiconductor site www.national.com or Texas Instruments www.ti.com and use one of the power supply design tools they have. They will specify the regulators to use and all the other components you need. Not sure about the variable one though.

 

If your MOSFET is getting hot with a light load then, all other things being equal, it's a sign that the switching losses are very high. Or it's stuck on permanently. There's a trade off between static losses due to high Rds(on) and switching losses due to parasitic capacitances. If the latter are too high, reduce the frequency (and increase the Ls and Cs) or choose a smaller MOSFET with a higher Rds(on).

National's WebBench is pretty good, saved me half a day on my last boost design, but it did give me a bum steer on the frequency-setting resistor, which should have been the easiest item in the design to calculate. So watch out for bugs in proprietary design tools, they're not perfect.

I like Linear Technology's switchers. They don't have a turnkey design tool, but their SwitcherCAD SPICE tool is optimised for SMPS designs, and is many times faster than normal SPICE tools for such things. Naturally LT have suitable models for their devices.

 

Another thing I thought of is your "load". 1M is not really a load for a PSU. Also a lot of SMPS NEED a load to run. I don't know if this is causing your heat problem or not.