i have made torsion pendulums in the past. and the thing i discovered that works to remove the oscillations is to attach the oscillating mass not only from above to the ceiling but also from below to the ground. to measure torsion, you can use a steel wire from above and from below. the torsion of it will be the "torsion spring" you will use. changing the diameter and length of the wire you can change the torsional spring constant. the configuration would be like this: https://drive.google.com/open?id=0B6KXLlAfKuDeLUstVk8zdkE0Tjg&authuser=0

to measure the displacement, you can attach a mirror to the steel wire, just above the platform. then you point a laser at it and you can see the reflection move as the thendulim rotates. we added a linear optical sensor so we had the angular position digitalized easly. and it was a perfect damped sinewave..we used it to measure viscosity of several liquids but i think you would not want added damping.
this is another schema: https://drive.google.com/file/d/0B6KXLlAfKuDeeEpaZTVsVG9aMTQ/view?usp=sharing
this was optimized to measure viscosity, you may not need the extra mass, and you don't want the oil container unless you want to introduce extra damping. you can see where the mirror is.
this is a picture: https://drive.google.com/file/d/0B6KXLlAfKuDeY3g2MmFhUHM4dGs/view?usp=sharing
you can see the laser, the linear sensor, the pendulum, mirror, the oil container could move up and down and the wire from below was attached to a screw and those two round things below worked as two nuts so you can adjust the tension of everything pulling from the bottom steel wire.

i hope this helps you with the experimental setup.

Thats really great. I have some PSD projects running (satellite sun sensor), would work as linear sensor for sure. I´ll check our ressources if it´s plausible to build a setup like this without having to stay awake 24h :) Thanx a lot for that insights.

it may not need to be that pretty, but do it the best you can (very straight steel wire, and very tense, the nut and screw below is very important so you can add tension to the system) , you may not even need the linear sensor, you can try to start from an still position (will be hard) and watch if there is noticeable movement in the laser reflection. the further away you watch for the reflection the more amplified the movement will be

Seems like some air flow issues in that room/closet...  a couple of suggestions showing up on the NASA thread:  http://forum.nasaspaceflight.com/index.php?topic=37642.msg1389001#msg1389001

Good luck!

Thanx, we take it into acount for the next tests.

In order for the scientists around the world to help you understand and improve your results, you need to disclose the internal dimensions of your baby EmDrive first (plate diameters and cavity length). We cannot help otherwise.

[EDIT] : thanks for your answer there: https://hackaday.io/project/5596-em-drive/log/19437-cavity-parameters

I love the latest test.  This is so clearly showing the oscillations, but it also shows something unexpected although many aren't sure what to make of it.  Keep going and looking forward to more.  I think your intention was to make observations and rule out causes of random osciallations vs thrust.  Good job!

Thanx. It seemed hopeless after we saw the data, but the water damping helps a lot.

In this video:I think the oscillations are due to the platform's center of gravity being offset from the axis of rotation of the magnetic platform.  Ideally you'd want the EmDrive on one side of the platform and a dummy with the same weight on the other side to counter balance.  I see you have some equipment there, but in the video it doesn't look like it is enough to correctly balance the load.  If it is not correctly balanced you'll get a feedback loop which will give you oscillations.  (Also it looks like the EmDrive isn't positioned exactly tangentially to the platform with respect to the axis of rotation, which means it is applying some force laterally which can also cause an oscillation).