If you go to 10 minutes in to video 2 the design seems pretty bulletproof, you do need to get it up to a sufficient speed though

I have built this same model …twice, one with a plain inner ring & the other with the holes in… & have had the same exact problem, I thought it was top heavy too so switched them both around 180*… no improvement… there seems to be an out of balance somewhere but can I as heck find it.. I set it off with a 3ft cord so it gets up to some speed but then the ali ring starts to spin with it, which is not s'posed to happen, when the spin eases off it gets top heavy again & then go shooting off the bench… all the materials are the same, & I made sure that all dimensions were exact using DRO's, even ensured the shaft had sharp points & oiled the brass cups with a drop of light oil…I even checked Springsprings video 3 times & I haven't done any thing different to his procedure, I thought it was just me that was having probs… glad to see not, so Bruce where do we go from here … ……?

George.

if you scale the dimensions to 0.74 of original the mass of all the bits reduces to 0.74^3=0.41, but since the the moment of inertia is mass*radius of gyration^2 it reduces to 0.74^5=0.22. It's a very long time since I did the maths for gyroscopes, but reducing the mass of the non rotating bits is an obvious start, and as Peter Krogh says above getting the mass of the flywheel in the rim will probably help as thinning the web will reduce the mass more than the moment of inertia.

A good start would be to spin up the flywheel and axle without the frame. There is some magic ratio between length of axle and inertia of flywheel which must be met or it will topple over. If it won't work without the frame it has no chance with it. Perhaps some mathematically minded reader will take this on as a Christmas brain teaser!

Taking Duncan s formula as read.
Mass by 0.74*0.74*.74= 0.41…

MI of rotor.. M *r*r

So thats 0.41*0.74*0.74..= 0.21

so assuming the mi is acting against the torque ( on this case to fall over)…
We have torque T proportional to mass* half height…ie T proportional to 0.74^4
And MI proportional mRR. 0.74^5
This means the restoring “torque”.. seems to reduce by 0.74 relative to the tendancy to fall over…
Thus a design that works at 1:1 might be marginal or not work at 1:0.74

Ants ( bees, and other arthropods)..don’t work scaled up to the size of sheep.

A bicycle wheel makes a good gyro even though it has a low mass rim, it's just a large radius. Perhaps you could try removing material from the inside of the rim so you have a lower overall mass but keep the most effective mass, that is the mass that is as far from the centre as possible. also thin down the centre boss as much as possible, this is mass that is supplying very little gyroscopic effect. Finally drill holes in the outer ring to reduce its weight. It does not need mass or great strength.

Martin

Thanks to the discussion, the relationship popped back to mind. IIRC it's something like W.r.r/2 for the inertia. Hard to math on the keyboard! But the inertia is an r squared function as everyone has stated so for a given total gyro weight the larger the rim radius the better the stability.

I love gyros; it's been a long time since I made one. Somehow I feel another coming on!

Pete

Here it is : **LINK**

..unless some other phenomenon is at play?

At one job the only way i could get a readable ECG trace was with the patient positioned N-S. Any other direction and the baseline oscillations on the trace made it unreadable. The proximity of ley lines, my boss's antimagnetic personality or the electrified railway at the bottom of the garden were all considered causes..

Where i live now there's a 600ft ridge just across the field. One day i shall find the energy to climb it and recover all those small parts missing from my shed. I'm convinced of the mass attraction that causes them to vanish.

Fast spinning objects may well be communicating with an alternative reality.. why in some places you can put a pair of socks in the spin drier and only ever recover one sock. It's no stretch to consider a perfectly spinning gyroscope can also exhibit abberant behaviour <s>