-------------------------------------------------------------------------------------------------------------------------------------------- 

Comment 1:

Just to add to the comments so far, aside from the rare desktop molding machine, here's how I believe most of us prototype plastics:

1. Start with a detailed knowledge of the process (to design the part correctly)

2. Once the concept is verified with low fidelity prototypes, create the CAD model

3. Create 3D print / SLA prototypes & test

4. Create injection molded samples using prototype tooling (outsourced to Protomold, Model Solution, etc)

5. Release to production tooling, then test / verify / tweak during manufacturing ramp

Generally speaking in my experience there's no in house molding involved in the design cycle.

--------------------------------------------------------------------------------------------------------------------------------------------

Comment 2:

Never screwed around with the tiny guys but spent a lot of time with the Morgan press. (14 ton clamping force, if I remember correctly.) This can be placed on a desk but is still heavy. imagine a 1 x 1 x 4ft tall footprint. It was very capable of making good parts but they retail for $20k, sell used for maybe $8k if you're lucky. Few things to know in general:

Take some parts that you want to mold and do a pressure and/or clamp force calculation based on volume and wall thickness. This will be an estimate but help you see if the little guy can even fill that parts you want to make. You can call the pressure fixed based on the max machine spec to solve one unknown and try this, for example:  http://www.improve-your-injection-molding.com/clamp-tonnage-calculator.html

Provided that you have a machine that can give you enough pressure and clamp force, don't forget about shot size. This will be the max volume of plastic it can shoot, including your sprue and runners. Some of these things don't do much. You also need to check max and min temperature ranges to make sure you can use the resin you're interested in. 

Lastly, now that you have a machine that works, you need to design the mold. This can be a pain in the ass, trial and error process. Every part is different but at least sometimes the things work. There are a ton of little tips and tricks for small molds but on the whole, start with a mold base that has room for pins and die springs, not just a screw driver pry slot. (Subtract those die springs from your clamp force, too.) You'll thank me when it opens itself ;) Having your own CNC helps as you'll want to be working with aluminum tools to simulate. Also, thanks to the internet, you can pirate moldflow software and tweak your runners in much less time than it takes to make a mold that doesn't work and make it again.