Brian Rauchfuss wrote:
> > The Navaid uses a macro-size gyro, mounted in a normal gimbal.
> >
> > The problem with the cheap MEMS gyros I have looked at is poor stability
> > of their zero output. They work well for damping rapid changes in
> > attitude (e.g., model helicopters) or for holding a heading reference
> > briefly (e.g., during a GPS dropout), but they would drift excessively
> > if used for something like a traditional gyroscopic heading indicator.
> >
> > Dave DeFord
>
> But they would be ideal for a turn-coordinator or a wing-leveler. In that
> case all you need to know is that you are not turning.
>
> Brian
If I might add observations, there's an issue beyond the hardware.
What correction signal should be fed to the servo depends upon how
much and what rate the deviation occurred (why turn coordinators with
their tilted gyro are useful as input). But in turbulence you can't
have the thing making the ride worse with excessive corrections, 1-2
seconds behind what's happening. IOW, in an RC model, nobody aboard
to complain about nausea. Historically they use analog circuitry with
design examples very hard to find, though now a microprocessor design
has the advantage of debugging it via only software. When I peeked
inside the Navaid autopilot, I saw a fair amount of circuitry, and if
the TC didn't partially solve the problem automatically, gonna need
lots more. Quite a bit of tricky "wheel reinvention" needed either
way.
Also, it was originally suggested that one could use aileron trim for
the servo, but Navaid et. al. provide for control stick type input.
The problem with that is you want it neutral when you disengage the
autopilot...unless the tab's effectiveness isn't enough to make that a
problem. But then it may not work well except in smooth air, in which
case it has little benefit.
Regards,
Fred F., A063
|