I have built and flown (with varying degrees of success) three different parachute deployment mechanisms that do not use pyrotechnic devices or regulated materials (such as "E-Matches").
The first device I constructed was based on an article titled "Eject-O-Matic" [1]. This uses a stick that goes from a bulkhead, up to the nosecond. It is held by a 'hook' that catches on the lower bulkhead, and elastic is used to propel the stick out of the rocket. The hook is held by a piece of brass on a bulkhead and is relased using an RC Airplane servo. Here is the hook mechanism:
I control the servo using an Atmel TinyAVR microcontroler. My intent is to some day integrate the servo control with a homebrew altimeter. However, for the first test flight, I reused an old board from a staging timer.
The switch on the left side is used to manually control the servo. Ground tests show this is capable of pushing the nosecone off and the laundy out. I have flown this once, and the timer did not activate the ejection.
I found an article by Doug Steinfeld [2] describing a spring-based ejection mechanism. I built one of those, too:
I controll it with another TinyAVR:
I have flown this twice. The first flight had strings inside jam up and prevent the plunger from deploying. On the second flight, the homebrew altimeter did not trigger until long after apogee -- in fact, I think it triggered when the engine fired the ejection charge. I'm not sure if this was an altimeter issue or a mechanical issue.
After those two problems, I started experimenting with using nichrome wire to cut various types of lines. My first try was a 1/8" braided nylon line that I use for shock cords on model rockets. It did but, but took a second or two. Dacron fishing line also cut, but took a second. Cotton thread cut instantly!
This set of pictures shows the BT-55 dual-deploy gadget. In the first picture is the altimeter, the plunger with spring, and the tube where the plunger is held. The second picture shows a similar setup for a plunger with elastic (like ejectomatic).
This is a recent verion of my homebrew altimeter with a board that has a MOSFET for switching high-current power.
Not pictured here is an up-scale (2.6") version using a pair of springs on the plunger. I ground tested it many times, and did one (seemingly) successful flight test. But, the first time I flew it with a hybrid motor and no other ejection, the parachute never deployed. I'm unable to tell if this was due to the device not cutting the string, or because of some other factor.