Treadmill Design

Please, start defining what do you call "good".

In engineering, nothing is "good", but, "more suitable to determined task or operational condition".

Start by defining operational conditions, loads, and what do you expect it to do.

The best way to approach a new design is to look at, and learn from, existing designs. Buy a treadmill and take it apart.

To sell or use or manufacturer? The context is important and the question not answerable without it. Treadmills are not simple devices to design to work reliably. Many more problems to solve than one would think on first glance.

First, go to school and learn all the engineering arts so that you will have the skills to do the massive amount of work involved that no one on here can possibly do for you unless you implement my second solution. Second solution: Hire several engineers to do the massive amount of work involved in such a project. I would be happy to design the electronics for you. I would estimate that that amount of work would incur design charges of fifteen to twenty thousand dollars plus parts and other services to produce a prototype. Do you need me to contract the mechanical engineer? That will probably cost about the same. Did you want just one or will you also be funding a marketing effort of at least a million dollars to realistically compete with existing high quality manufacturers of treadmills. This project will probably cost a couple of million dollars to get off the ground. When should we start?

You definitely need to specify the end use.

Without knowing any details I will recommend using one of these as a model. www.woodway.com

Many college bio-mechanics labs build their own as a good treadmill can be expensive. From my observations, they were all well overbuilt. One had a 20HP motor that could drive the belt over 40MPH with enough torque to rip a limb off. It reminded me of a giant scale belt sander. Use caution and common sense whatever you do.

Similar to designing a good (or "better") mouse trap . . .

I agree with rcapper.

What's wrong with ones on the market?

It looks like you've started off on the right foot as the first step seems to be to free load a design off of some engineering forum. You have that one covered...