Quadcopter anti-vibration tube mounts.

[Vince Hogg]

Having seen a vast improvement in the Arducopter handling when vibration was reduced I decided to experiment further. I thought it was worth trying to soft mount the tubes so cobbled together a test frame on a bit of 4x2. The 12mm tube mounts are held in place with ‘o’ rings. The video is with standard 75 shore rings 17mm ID. This was just the first test so was quite pleased but will experiment with other types/sizes.

This method will probably only work on a ‘H’ frame where the motor lift balances out.

 

 

[OlivierD.]

Cool stuff! What is the name of the App you are using on the galaxy to measure vibrations?

[Vince Hogg]

Vibration Monitoring

[Vince Hogg]

If all goes well, hope to end up with something like this.

[OlivierD.]

Very slick. I imagine H quad tuning is the same as that of an X quad? Pardon the noob question, never flown and H but the more I look at them, the more I like the idea. Thanks for the App!

[Joshua Ott]

Good looking project Vince. I've said before, multirotors are ultimately just a packaging exercise and the traditional hub and spoke frames are a bit tight, especially when one is trying to separate video Tx, RC Rx, data TRx, GPS Rx, power supplies, ESCs, batteries, etc. It's amazing any of these things work at all!

Have you considered making the motor boom quick release or rotate to facilitate ground handling and transport? I have vowed to only make folding frames in some type of H config for all my future projects. 

[Vince Hogg]

I found that my traditional open ‘X’ frame was getting splattered with mud and debris and is very difficult to clean. I thought it was just a cosmetic problem until one ESC failed. On examination it was very wet inside the plastic wrap. It started working again after being dried out.

I expect tuning will be like a normal ‘X’ except it may need slightly higher gains in pitch due to the increased inertia around that axis.

The shell will be of Kevlar/carbon around a ‘lost foam’ mould. It doesn’t produce a nice surface (without a lot of finishing) but Im not bothered. Hope to start that today. The ‘airplane’ shape fuselage is just to help with visual orientation. Some people think the 5.8G FPV interferes with GPS so they will be at different ends. I do intend to put legs on it when I attach the camera gimbal.

It probably would be easy to make the booms quick release but Im not bothering this time.

Am in a hurry to finish so I can try 2.9!

[thomas Butler]

For vibration analysis on the cheap, in the past,I've used an ADXL335 breakout from Sparkfun mounted on the quad and attached the microphone input of my PC.  Using an audo analyzer program on my PC  to capture the data through the microphone input.  This is then plotted using Excel.  Advantage is that it requires no hardware that could corrupt the sample set.

Now days though it's easy to use XBEE with my Crius all-in-one (or an APM) and capture the data in Mission Planner using the accelerometer data of the FC.  Logged data is in csv format which is imported to Excel for calculation of mean, variance, and standard deviation.

A slick addition to mission planner would be to display, in the sample plotting screen, the mean, variance, and standard deviation.  The use of statistical data is much more effective (and scientific) than watching a plot!

Additionally, Vince, I would shock mount the motors rather than the arms (jdrones sells a vibration mount for motors), but it easy to make your own using hard disk drive mounting grommets.  The quad arms can act as an amplifier of the motors vibration.  Note that the 2x4 and your are probably acting as a dampers too; reducing the resolution of your measurements.

Another trick is to inject high expanding foam (can available for sealing cracks in walls) into the beams to deaden vibration.

Yet another is to add mass to the FC board (5mm standoff it to a 4mm x FC length x FC width) aluminum plate) and then vibration mount the assembly.

Finally, the H-configuration is probably the way quadcopter design is headed.  The X and + layouts leave little room for avionics, etc.

 

[Vince Hogg]

Hi Thomas.

All good info thanks.

I did put a post somewhere (Mission Planner forum?) asking if we could highlight parts of a graph and get statistical data.

My test was not that scientific but I was just proving to myself that the o-ring mounts were worth taking further. I will use the APM to get good data when its flying. It would be nice to have a league table of 'good' and 'bad' vibration measurements. The phone program I was using gave a vector total vibration but Im not sure the APM will go that, just separate axis.

I think almost everyone misses the trick of using the battery mass to dampen vibration by just hanging the battery loosely below.

I will clamp it tightly just above the APM and will try some 5mm 'special' foam to mount the APM on but have seen it cause more problems in the past.

There have been comments that my 4008 type motors have weak shafts and that any prop ground contact will cause a bend. I think some of mine may be slightly off so will replace with hardened ground shafts.

Expect to be flying in the next day or two.

[thomas Butler]

I'm a software developer , and I've used and loaded both MPNG and Multiwii firmware (with my own mods) onto my FC boards.  There is a Mulitwii setup/analyze app for Android using bluetooth (via USB so it could use XBEE or any other wireless link ) that is excellent.  It displays waveforms for the sensors and allows setting PID, etc.  Considering all the development in APM, I don't see why someone hasn't developed a good setup/analysis app for use with ArduCopter/MPNG by now (I would , but I suspect there are a dozen people already on it and the Andoid development environment is a dog using Eclipse).  The simple, but effective MavLink protocal should make an app relatively easy to create.   I know there are a few half baked apps for Android...

 

Can you say what your source is for the carbon fiber layup materials is?  Thanks.

 

[Vince Hogg]

Im in UK so prob not much help.

http://www.easycomposites.co.uk/

I wont be buying Kevlar anymore. Its such a pain to work with.