I have made a lot of unique business card holders over the years, but I always wanted to make one that could automatically dispense cards.

It sounds pretty straight forward, but I know that to make a reliable and simple machine that can do this is not an easy task and there are several risks that turn it into more of a research project than a sure-fire design project.

The Project14 Geeky Gadget is a good opportunity to see what can be accomplished, since there are no product sponsors counting on a successful project.

So this project is dramatically different from my usual projects, where I do enough up-front homework to ensure the resulting design will work on the first try.

I expect it will be a meandering narrative of tests and discussion of what works or doesn't work - it might provide insight into the crazy tangents my mind is capable of, but I will probably sensor some of that.

There is a master plan which involves quite a lot of work, but I don't want to build up expectations and pressure to deliver, so the plot will thicken as progress is made.

It is not set up to stick rigidly to the Project14 schedule - that is just an excuse to work on it - if it works by the deadline - great, if it doesn't work by then - still great.

I use low cost components which equals long delivery, so missteps will slow it down significantly (already happening).

This initial part of the blog will cover some of the issues with motorized ejection of cards.

The debate over whether to eject the top card or the bottom card is still not settled.

Ejecting the top card means the stack needs to be spring loaded or motorized which adds to complexity.

Ejecting the bottom card allows gravity feed, but the force will change as the stack shrinks.

In both cases there is a problem of how to eject only one card. Any restraint or pushing lip that acts on the edge of a card needs to be thinner than a card, but cards can easily be bent by much more than one card thickness.

If a friction wheel is used for propulsion, ensuring a constant amount of friction has many problems. Both gravity and springs provide a different force depending on stack size.

There needs to be force on the card to generate enough friction to move it, and the friction with the wheel needs to be greater than the friction between cards.

The wheel cannot be flush with the surface or it won't contact the card, but how to prevent cards from warping around the wheel and still only feed one at a time?

If there is a funnel for the ejecting card to ensure it doesn't get caught on a lip, how to ensure multiple cards don't get jammed in the funnel?

These are the kinds of thoughts and questions I have running around in my head - each one leads to a whole series of tangential possibilities.

I decided to start with a simple gravity feed hopper and single drive motor to see if it would work.

My design calls for a small motor (required by the mysterious master plan), so I ordered a tiny flat motor when Geeky Gadget was announced - it looks like one I've seen in old slot cars. But there were no specs.

When it arrived a week ago, I got to measure it and print off a card-handling hopper.

The following video shows the best results I could get with this prototype.

The simple system worked after a fashion, but I want something much better.

I learned that I wanted slower motors with more torque and the feed system needs a lot of thought to improve its performance, so some really small gearmotors are on order but again - no specs.

I think I will try a 2 stage ejector - the first motor just moves the bottom card into the "breach" where I can control the friction and ejection speed because the second ejection motor only has to deal with one card.

The hopper "nozzle" will be redesigned to only have critical tolerance at one point where card selection will be controlled. Everywhere else the system will use smooth guides. I may be able to prove this concept with another bigger gearmotor I already have.

To be continued......