[Top] [All Lists]

Re: [TowerTalk] Flexible coupler assembly

To:, "'Chuck Lewis'" <>,<>
Subject: Re: [TowerTalk] Flexible coupler assembly
From: Jim Lux <>
Date: Thu, 25 Jan 2007 16:12:31 -0800
List-post: <>
At 03:30 PM 1/25/2007, Gary Schafer wrote:

> > -----Or
> > The donut helps because it spreads the impact over a longer time. Without
> > the donut, the force is applied over a short time after the backlash in
> > the
> > gear train is taken up, and the gears absorb the total energy almost
> > instantaneously. But the donut allows the same total energy to be applied
>Hi Chuck,
>Yes I understand how the force is spread over time with the absorber. My
>question is more concerned with how it will affect the forces when
>oscillations are present.
>We know that energy can not be created or destroyed it can only be stored.

Practical couplings like this can be made to be quite lossy, esp. at 
high frequencies, which is where the high accelerations come 
from.  There's a whole science/engineering of this which might be 
worth thrashing through if one were developing a product, as opposed 
to scrounging in a junkyard.

>So most of the energy that is absorbed by the donut will be given up in the
>opposite direction to cause the antenna to rotate backwards from its initial
>direction from when a wind gust moves it.

And some significant amount of the energy is also dissipated in the 
air drag of the antenna.

>I know, energy can not be destroyed or created but it can be stored and
>released at different times and if that release time coincides with a wind
>gust then peak loads may be increased. I am not saying the donut is not a
>good thing just some things to think about.

Exactly... you want to make sure that the natural frequency of the 
antenna:mast:donut doesn't happen to match, for instance, the natural 
frequency of the elements or the period of the gusts (or more likely, 
the turbulent eddies from a gust).

And this kind of thing is real hard to predict.  Some adjustability 
is nice, for instance, in the rotator case, just adding some 
frictional drag on the mast (above the coupling) might solve a lot of 
problems.  If you added friction of say, 20-30% of the inertial load, 
you won't change the slewing time all that much (esp for short moves) 
but that's a heck of a lot of damping.  Some sort of spring loaded 
band around the mast would be one idea.

The other approach, if you have a moderately sophisticated motor 
drive system is to put the damping in electronically.  Even the 
little $100 motor drives have this sort of intelligence these days, 
although doing the tuning can be a bit exciting if you aren't 
careful.  Everyone who's done this has probably had an experience 
where they pushed one of the gain terms a bit to far and it breaks 
into uncontrollable oscillation or a hard to the stops limit 
move.  There's even automated programs for doing the tuning (as well 
as self tuning controllers, but they're pretty expensive).

>Another thought on the subject: It may be more convenient to mount the shock
>absorber under the rotor between the tower and rotor.

That's a possibility, but one that works better for things like high 
frequency vibration (like, say, a fan motor).  I always prefer to 
bolt the motor down securely and deal with mechanical issues 
downstream.  But hey, it works for cars and airplanes..where the 
motor is connected more or less rigidly to the drive shaft, and the 
motor mounts are flexible.  In the airplane case, the mass of the 
load is much, much less than the mass of the motor, though.

  Bringing up an interesting approach.. bolt the antenna rigidly to 
the motor, which sticks up, and bolt the output flange of the gear 
box to the top of the tower.  Nothing says the motor has to be 
unmoving.. you're already running a flexible coax to the antenna, 
running a flexible cable to the motor is really no 
different.  Doesn't change the forces, but does mean that you have 
more flexibility in motor and drivetrain size, since it is "outside" the tower.

Jim, W6RMK. 


TowerTalk mailing list

<Prev in Thread] Current Thread [Next in Thread>