Archive-name: rockets-faq/part8
Posting-Frequency: monthly
Last-modified: 12 November 1994
*** PART 7: Guidance and Control Systems
Any additions or corrections should be sent to that address]
Updates
-------------
November '94: Added "Recovery Aids" section.
August '94: Added "John Sicker" to "Sources" section.
June '94: Added "Designex" and "muRata" to "Sources" section.
April '94: Split off from "Payloads" section and renumbered Part 6.
Added "Pendulum" discussion and sources of components.
-------------
Introduction - The only universally accepted method of model/HPR guidance is
the static fin. While these take on an enormous variation in shape, size
and location, they all do the same thing: move the CP back behind the CG
so that the rocket is statically stable. Since the beginning of the hobby,
folks have been experimenting with more sophisticated ways of guiding their
rockets for lots of different reasons. None of the systems have been so
successful that the technique has been incorporated into a "regular" rocket,
except to further the experiments.
Compared to other parts of this FAQ, the following will seem much more
theoretical than practical. This is necessary because a flying rocket is
an extremely dynamic system and controlling it with an active guidance system
is a very non-trivial task. Before you jump in and start grafting gyros and
servos to something that can approach the speed of sound you have to
understand the basics behind how such things work...or, more importantly,
don't work. After the theoretical grounding, specific examples and references
will be given so you can review the work of others.
After the Guidance Systems is a section on Control Systems. While they sound
similar, Control is much simpler in a rocketry context. This section will
deal with the sequence controllers, timers and other increasingly sophisti-
cated means of making sure that the rockets (especially the big HPR stuff)
perform as desired.
Finishing up this part of the FAQ is a new section on yet another niche
for electronics in the hobby: Recovery Aids.
7.1 What kinds of Guidance Systems have been tried? Does anything work
besides fins?
7.1.1 Guidance, Active vs Passive - All hobby rockets, with the exception of
the experimental ones described later, are passively stable. They achieve
this by the simple expedient of placing the CP behind the CG, which is almost
always done with fins. Passive fins are placed at the back of a rocket
because that's where they are needed to provide negative feedback. As much
as this sounds bad to the touchie-feelie crowd, negative feedback is simply
an engineering term which means that any disturbing forces are fed back into
the system in the opposite direction. Thus any force which causes, say, a
positive pitch will make the fins generate a negative pitch force to help
put it back right. Positive feedback, conversely, causes continued motion
in the *same* direction, exemplified by the tight spirals of an unstable
rocket where the fins push it further in the direction of the error.
I say that passive fins 'help' put it back right because, being a passive
system, it has no way of knowing which way 'right' is. If you launch
vertically with a side wind, then passive fins will steer the rocket to a
direction which is a combination of the side wind and the apparent wind
caused by the rocket's own motion (this is known as vector addition and will
come up again further down). We all know this as 'weathercocking.'
Active fins are a whole 'nother story (we'll continue using fins in this
discussion even though there are many different ways to affect a rocket's
flight path as we will see later in the 'Gyro' section; fins are just the
most common). Active fins are usually pivoted on their root edge like the
rudder on an airplane. The biggest debate on active fins is which end of
the rocket to place them. Both front and rear have their trade-offs.
6.1.1.1 Rear Mounted Fins - Rear mounted fins seem more "right" because that's
where we're expecting to see them. They also have a legitimate benefit in
that should the guidance system fail, the fins will add to the static
stability, just like always. One consequence of rear mounted fins to keep
in mind is that the control inputs must be reversed. Just as an airplane
rudder must push the tail left in order for the plane to turn right, rear
mounted fins must turn in the direction of the error to correct the rocket's
path. This leads to them being very effective since you get to add the fin's
angle-of-attack (alpha) to the rocket's.
For example - If the rocket develops a 5 degree alpha with respect to the
vertical, then the fins, since they're pivoted the same direction, will
add another 5 degrees thus doubling the amount of corrective lift generated
(assuming of course that the angles are small enough that the fin doesn't
stall). The down side of this is that the fins might prove *too* effective
and could cause the rocket to overshoot its desired path and head the other
way. This could lead to an unstable oscillation if not damped out. One
way to lessen their impact is to make the fins small, which has the usual
benefits of less weight, less drag and (if this is a scale model of a
finless missile) less visual impact.
Other problems with rear mounted fins is that the back end of a rocket is
already a pretty crowded place. There's that massive heat generator (the
motor) taking up most of the airframe, and the last thing you need is more
mass from fin actuators, bellcranks, bearings, etc. at the "wrong" end
detracting from the rocket's static stability.
6.1.1.2 Front Mounted Fins - For front mounted fins, one is tempted to say
"take everything in the previous section and reverse it" :-) It's not quite
that simple as they have their own subtleties. The visual response from
seasoned rocket designers is usually a shudder since front fins are what
everyone is always trying to avoid. Indeed, should the guidance system fail
they will pull the CP forward enough to make the rocket statically unstable.
For this reason, forward mounted fins are almost never used by themselves,
but rather in conjunction with static rear-mounted fins to provide "trim."
Front mounted fins are not very effective; the reason being that when
they're up front, the fins *subtract* their angle of attack from the
rocket's. Think of it this way: While the rocket is ascending vertically,
the fins are also vertical, of course. If a gust of wind induces, say, a
positive pitch then the control system will command the fins to a negative
pitch position. This puts the fins nearly vertical again with almost *no*
angle of attack WRT the relative wind to generate corrective lift. It's not
until the rocket starts developing some horizontal velocity (since the motor
is no longer pointing completely vertically -- vector addition again) that
the fins start generating some lift to put the rocket on course.
While this makes front mounted fins less desirable as a primary guidance
solution, it actually enhances their use as guidance trim. By adding active
forward mounted trim fins to a basically stable rocket (with passive rear
fins) you can "fly it all over the sky." Since the fins are less effective,
your steering inputs don't have to be so subtle and the risk of overshoot is
greatly diminished. As speeds increase you really don't want a control system
that's too "touchy." This is, in fact, why supersonic missiles like the
Sidewinder and AMRAAM use exactly this control scheme.
Finally, front mounted fins can have all the benefits that rear mounted fins
lack: system comptactness (since the control system, fins and actuators are
all at the same end), static balance (all the system mass is at the front),
avoiding the heat and space restrictions around the motor, ability to be
recovered separately in their own payload compartment, etc.
Well, with that "ground school" out of the way, let's get on to some
specific examples:
7.1.2 Historic - Ram Air. The old _Model Rocketry Magazine_ carried a two part
article by Forrest Mims in the February and March 1970 issues on Ram Air
guidance. In this system, air entering through a hole cut through the
central axis of the nose cone was redirected out side ports by a rotating
duct. The idea was that under normal flight, the air would "puff" out the
four side ports (along the pitch and yaw axes) in rapid succession causing
no more than a wobble in the flight path. When you wanted to change course,
the duct would be quickly stopped in front of the appropriate side port to
let air effect the change then it would go back to spinning. The problem
(which should have been obvious, IMHO) is that a flying rocket is a classic
free body and that the torque necessary to start and stop the spinning duct
would cause the rocket itself to spin the opposite direction. This means
that the port you were lining the duct up with was no longer aimed in the
"right" direction. The system used a sun-seeking sensor which looked out the
central ram air inlet.
An equally creative (and unworkable) guidance method by the same author was
detailed in the Nov '70 MRM. This one started out the same with a central
ram air inlet and a side port (but only one). Inside the side port was a
pair of electrical contacts connected to an enormous capacitor. The air
entering the top spun a small propeller connected to a screw. The***
closed the contacts discharging the capacitor and causing an acoustic wave
to exit out the side port. The idea was that this acoustic wave could be
used to influence the rocket's path. Of course, this setup was just used to
see if any course change was detectable at all (it wasn't). How to recharge
the capacitor and time/aim the discharges was left as a project for future
generations :-) Gotta give this guy an "A" for lateral thinking!
7.1.3 Sun/Horizon (target) seeking - The Ram-Air system described above was
a sun seeker, albeit a very crude one. The sun is a marvelous target for
experimental guidance systems. It's massively bright, so it's easy to
detect (in fact, it's hard to miss :-). It's always "up" in the sky [if
you're not launching too close to sunrise or sunset] so you don't have to
worry about violating the safety code WRT flight path angle. Finally, it's
out at infinity, so you don't have to worry about actually reaching it (i.e.
a constant target).
Perhaps the most in-depth sun guidance system yet built was done by the
Zunofark team of George Gassaway, Matt Steele, et. al. and covered in mind
numbing detail in the May/June and July/August 1992 issues of HPRM. The
articles are a reprint of the research paper that took first place in the
Senior division at NARAM 30 (this paper is also available from NARTS).
It covers all phases of the project from basic theory and construction
details through experiment design and execution. The research vehicle used
the passive rear fins/forward trim fins control scheme. Control input was
by a quadrature style sun sensor looking through either a transparent or
translucent nosecone. Fin actuation by RC airplane type servos. The
articles include appendicies.
7.1.4 Pendulum - An article of similar depth to the "Sun Seeker" above, but all
theory, was presented in the July/August 1993 HPRM. Author David Ketchledge
starts with the basics by revisiting the famous Barrowman Equations for
determining the CP of a rocket. He then continues on to explore the dynamics
of rocket flight. My only problem with the article is that the equations are
presented in FORTRAN or BASIC style arithmetic statements rather than using
standard mathematical notation which makes it difficult to follow. At the
end of the article, Mr. Ketchledge proposes a guidance system design based
on a free*** pendulum to use as the vertical reference. He provides
multiple computer simulations to show the effectiveness of this sensor
compared to several others (including the Zunofark sensor described above).
Penudla are an interesting concept, and are often chosen by those designing
their first active guidance system. Mostly this is due to being simpler
and more intuitive than gyroscopes in their operation. However, even a
simple analysis of operation in the "real" world (as opposed to a computer
simulation) shows their deficiency as a primary guidance tool. The following
7.1.4.1 Pendulums don't react to gravity - Pendulums react to any moment
(torque, couple) that appears between the weight and the pivot. When a
pendulum is stationary this moment would be created by the force of gravity
on the weight and the reaction force from the ground applied through the
pivot. Now, what happens in a rocket (or other free body)? When the system
is in free fall, the mass is pulled down by the force of gravity so it is
accelerating downwards at 1 gee. Also, the pivot (and attached rocket) is
pulled/accelerating downwards at 1 gee. Net result? Nothing. The force due
to gravity is equal to the force required to accelerate the masses (inertia)
at 1 gee so there is no net moment acting between the pendulum and the pivot.
The pendulum would react to external forces (thrust, drag...) but even with
these present, it will still not react to gravity. That is, pendulums react
*only* to forces applied to the pivot that are not applied to the weight. A
pendulum inside a free body will exactly align it self with the forces that
are applied to that body but not to the weight.
Not convinced?
Think about what happens when someone lets go of a pencil in a spacecraft
that is orbiting the earth - it doesn't appear to fall. Observing a space
craft from "outside" it is very clear that it is falling quite rapidly. But
from inside the space craft, that acceleration is not observable - things
appear to be "weightless" - not affected by the forces of gravity. Likewise,
a pendulum would not tend to point "down".
A pendulum in a rocket that is thrusting horizontally will not point at
some angle between the horizontal thrust and "down", it will point exactly
horizontally (and the rocket will fall towards the ground at 1 gee). It
should be pointed out that a pendulum in a rocket that is moving truly
horizontally (due to aerodynamic lift keeping it from falling) will point
at some angle towards the ground. But, this is not because it is reacting
to gravity, but rather because it is reacting to the force (lift) that is
acting on the pivot but not the weight.
Bottom line: A pendulum inside a rocket in flight will give absolutely no
indication of which direction is up.
7.1.4.2 Does this mean that I can't use a pendulum to create a guidance system?
Yes/no/maybe... It depends on what you mean by "a guidance system." From
some of the criticisms of pendulums (pivot friction, small deflections
relative to thrust, etc) some folks mean "something that makes a rocket fly
straight up". If so, a pendulum clearly can't work. But, that must also mean:
Fins Don't Work.
Well, it's true. Passive fins as found on 99% of all hobby rockets don't
work. You can't use fins to be sure that a rocket will fly straight up.
Why not? Well, as described previously under "Active vs. Passive," fins
don't know which way up is. Plus they don't react to gravity. What they
react to is a difference between the orientation of the rocket and the
apparent wind. For example, when a rocket comes off the end of the launch
rod a cross wind will tend to force the rocket away from it's vertical path
and make it point towards the wind. Also, they only react in proportion
to the magnitude of the angle between the rocket and the apparent wind
(angle of attack). So, for example, if thrust is applied off center the
fins will not completely correct for this and the rocket will tend to curve
off in one direction.
Now, I probably didn't tell you anything you didn't already know about fins,
right? So, why did I waste your time with the last paragraph? I wanted to be
very clear what it is that we really need from a "guidance system." We don't
need something that senses up. We don't need something that is particularly
accurate. All we really need is something that tends to keep the rocket
pointing in the general direction of the launch as long as the magnitude or
duration of external disturbances or internal imbalances are not too large.
So, if we don't need a system that senses gravity and will not insure that
a rocket goes "up" (e.g. fins)... perhaps we shouldn't be too quick to say
we can't use a pendulum just because it can't sense which way is down or
because there will be friction in the pivot.
7.1.4.3 So how can a Pendulum be used?
Now, given that a pendulum does not know which way is down, can we still
get any useful information from one?
Let us consider a pendulum hung at the center of gravity of a rocket. What
happens when it is hit by a crosswind (say, from the left)? As the rocket
is accelerated to the right, the pendulum (as viewed from inside the rocket)
would tend to move to the left. If it were attached to movable fins (or
whatever) it could be made to cause the rocket to turn to the left, i.e.
weathercock. When the acceleration stops or is offset by the rockets
thrust, the pendulum would return to the center. What happens if we have
an off center motor that tends to make the rocket yaw counter clockwise
(to the left)? The forces that make the rocket turn about it's center of
gravity are not directly applied to the pendulum so it will lag behind
(as seen from outside) and from inside the pendulum will be seen to be
deflecting to the left, same as above. Oops, the above system will tend to
make the rocket turn more to the left...positive feedback...Loop...Crash!
Bad idea. Reversing the connection between the pendulum and fins would
make it tend to react properly to yaw but make the rocket turn downwind
in a gust. Not good.
Plan B. Suppose we put a pendulum ahead of the center of gravity. In the
above cross wind it will try to steer the rocket to the left and as the
rocket starts to turn the nose will be accelerated to the left and center
the pendulum. It seems likely that the net reaction will be minimal. Now,
if the rocket starts to yaw counter clockwise (accelerating the pivot to
the left) this time the pendulum will swing to the right (as seen from
inside the rocket) and tend to steer the rocket to the right (clockwise).
Hmmm, just what we want. We should note that the distance between the
center of gravity and the pendulum will determine the relative magnitude
of the reaction to lateral accelerations and yaw accelerations.
Ok, so far it looks like plan b might be possible. What happens when the
motor burns out? Oops, now the primary force on the rocket is drag and not
thrust. A weight on the end of a stick style pendulum will want to flip
over and make the rocket fly backwards. Bad idea.
Plan C. How about a weight that is free to move horizontally (with respect
to the body tube)? Well, it would react just like the pendulum in plan B
except it would not need to "flip over" at burnout. Doesn't sound too bad,
does it?
Bottom line:
Will it work? I don't know, maybe it would. I certainly wouldn't argue that
it couldn't work, particularly when you consider that model rocket guidance
systems don't have to work very well or for very long. But is it really worth
all that effort to build a system that works about the same as passive fins?
7.1.5 Can I use Gyroscopes to stablize my rocket without fins?
This one's going to need a little more ground school.
Gyros fascinate us because they violate our common sense perceptions of
mass and force. Everyone who has played with a toy gyroscope marvels at how
it "resists" the twisting and turning of your hand. Eventually, most rocket
hobbiests come up with the idea that if you put a gyro on board a rocket
then it would "resist" all of the external disturbing forces and cause the
rocket to fly straight without any fins.
Sorry, it doesn't work that way.
Gyroscopes work on the principle of rotational inertia. Just like with linear
inertia (where a mass moving along a line will continue along that line
unless disturbed by an outside force) a mass set spinning on an axis will
continue to spin around the same axis unless forced to change. If you do
force it to change, however, the results are not what you'd expect.
The reason the passive gyro won't work is due to the physics of rotation. The
basic gyro "law" is as follows. Gyroscopes have three axes: the spin axis,
the input axis and the output axis; all at 90 deg to each other. Twisting
the gyro about the input axis will cause a torque about the output axis.
Putting this in rocketry coordinates, if the gyro rotor is spinning on the
long (roll) axis of the rocket, then anything that causes a rotation about
the yaw axis will torque the rocket in pitch and visa-versa. This means that
if you launch your finless rocket with a gyro spinning vertically, then a
gust of wind from the North will cause it to veer East or West (depending on
which way the rotor is spinning).
"Well," some folks argue, "then all I have to do is put another gyro with its
rotor spinning along the yaw axis and maybe a third spinning on the pitch
axis. That should 'resist' torque from any direction." Sorry again. Just as
weathercocking is an example of linear vector addition, the angular momentum
of spinning gyro rotors add up in the same manner. Three identical rotors
placed orthogonally like that will cancel each other out (vectorially adding
up to zero) and act as if they weren't even there (except that you'll be
lifting a *lot* of useless mass :-).
The correct way to use gyroscopes in a guidance system is as a REFERENCE
PLATFORM. What that mean? Well, remember how a spinning rotor will continue
to spin on its initial axis unless disturbed by an outside force? Rather
than lashing the gyro to your payload section and forcing it to twist with
your rocket, it should be mounted in a gimbal (a two axis bearing originally
invented to keep ships' lanterns vertical in heavy seas). In this way, no
matter how much your rocket pitches and yaws, the rotor will continue to spin
about the same axis it started with on the pad. With this constant "up"
reference, you can build control systems to keep the rocket heading in that
direction. There are several ways to do this:
7.1.5.1 Fins, Mechanical - This is probably the closest thing to the ideal
passive gyro that everyone thinks of since it's all-mechanical. With a
fairly massive gyro rotor spinning in a gimbal, bellcranks can be run from
the gimbal axes down to the fin pivots. The only tricky parts are remembering
to cross the belcrank rods to get the reversed action required (See "Rear
Mounted Fins" above). Also, some means of providing recovery that doesn't
blast the gyro and linkages with ejection gas is needed.
I've heard of such an all-mechanical design by word-of-mouth, but was unable
to find any references in either AmSpam/Sprocket or HPRM (but my collection
only goes back to late '91 for the former and mid '92 for the latter) so
I decided to design my own. It uses RC airplane components for all the
movable pieces (cheap and reliable) with a homebrew gyro and gimbal. It
will be about the size and shape of a LOC Onyx, since that will give me a
baseline comparison.
7.1.5.2 Fins, Electronic - The more sophisticated approach to gyro control is
to use a "real" electronic control system combined with the reference
platform. While I haven't found any reference to such a system actually
being built, an excellent source to draw from would be the Zunofark design
described above. If one were to replace the 4-direction Sun sensor with
rotational position sensors on the gimbal axes (along with the appropriate
signal conditioning), you would have a very workable setup. A setup, one
might add, that could be programmed to head in any direction, not just
towards an external signal source (can you say "Inertial Guidance"? I knew
you could :-)
Historical note: This is exactly how the German V2 guidance system worked,
the only differences were in the details: It had fixed rear fins for basic
stability, trim tabs on the trailing edges of the fins, plus (since it had
to travel outside the atmosphere) graphite vanes to vector the exhaust. While
it didn't have modern digital electronics, it used very similar analog
predecessors. Also, the gyro platform had a three axis gimbal so that roll
was controlled as well as pitch and yaw.
The gyro platform was set up to keep the rocket absolutely vertical WRT its
launch site. To hit a target, the launch pad was aligned with the rocket's
pitch axis aimed towards the target. After liftoff, an actuator pushed
on the pitch axis of the gyro forcing it off center. The guidance system
interpreted this as an error and "corrected" it by pitching the rocket the
other direction (towards the target). As the actuator continued to push,
the rocket continued to pitch over until it ran out of fuel; at witch time
it was (theoretically) directly over the target and heading straight down.
Range was controlled simply by controlling the rate of the pitch actuator
and how long into the flight the pitch program started.
7.1.5.3 Gimbaled Motor - The only example of hobby rocket guidance done the
way the "big boys" do it, was covered (somewhat sketchily) in the May/June
1993 issue of HPRM. Richard Speck designed a gimbaled motor system consisting
of a two axis gyro reference platform combined with a two axis motor gimbal.
This was an all analog system which even included phase comparitor circuitry
to prevent over correction. The first version of the test vehicle hedged its
bets and included fins, but the second one had none; a true finless missile!
The design was used as a basis for an eight foot "high fidelity" Saturn V
model with five engines in the first stage; the central one being fixed and
the outer four each being on a one axis gimbal along the pitch and yaw axes,
just like the real thing! A "progress report" photo appeared in the Sept/
Oct 1993 issue of HPRM.
7.1.5.4 2nd Gyro torquing - Finally, there is a technique for controlling
rocket attitude without fins, gimballed motors or any outher external
affectations. In fact, it's very close to the presumed ideal of a gyro
that "resists" external forces all by itself. The technique was originally
used to stabilize ocean liners along their roll axis but is now used in
some spacecraft to do attitude control without the use of gas jets.
The first thing you need is a reference platform to tell you which way is
"up". This can be a small mechanical gyro with encoders on the gimbal axes
like we've been discussing, or even a non-gyro system like horizon sensors
(this is the way satellites do it) or a "target" sensor like sun guidance.
The second part is the control gyro. It must be fairly massive and positioned
somewhere around the rocket's CG. Actuators are placed on the gimbal axes so
that when the reference platform detects an error, say on the yaw axis, the
actuator twists the gyro's pitch axis which forces it to precess in yaw. If
you've got the signs hooked up right, this will counter the disturbing yaw
and put you back on track.
While the theory is good and has been proven out on real spacecraft (such as
the "Magellan" Venus orbiter) it all seems quite involved for a hobby rocket.
I'm sure someone out there will try it for just that reason :^)
7.1.6. All this talk about reference platforms seems so complicated. Why
not get some of the gyros the RC-Helicopter folks use? These are small,
relatively cheap, and are designed to hook into servos.
This comes up almost every time the conversation turns to gyro control. The
main reason this type of gyro won't work is that it is a *rate* gyro. This
means that it doesn't give you the absolute "up" reference like a position
gyro, but rather just how fast you're turning about an axis.
This is not to say that a rate gyro can't be used in a guidance system. It's
done all the time in professional rockets, but the sophistication necessary
to integrate the rate signal to get a position is beyond the capabilities
of most hobbiests.
talking with Rob Rau of High Technology Flight about this. The reasons why
R/C helo gyros will not do for rockets are as follows: the accelerations
are so drastic that RCG (r/c gyros) cannot react fast enough. Also, the
sampling rate is also too slow, and are only good for about 5 degrees of
deflection from whatever normal reference point is set up. We all know that
rockets get far beyond that 5 degree limit. I suggest that if you want a
system that will work with rockets, contact Bob Rau at HTF. [See the address
section for his address and phone - JH] Call him up and get one of his
catalogs...you will be impressed!"
is true that heli gyros are rate gyros and will not hold an absolute heading. HOWEVER, they are still useful for damping undesirable deflections (which is
why we use them in tail rotor circuit). I have seen many launches where a
rocket clears the launch rod, then turns 20-30 degrees in random direction.
This is the sort of thing that a rate gyro could handle. Incidentally, my
heli will yaw (at full right control) at about 1000 deg/sec (3 rps) without
causing a problem for the gyro. I don't think the pitch rate in a rocket
will give a problem to these gyros! Also handles linear accelerations very
well, and costs <$100 by quite a bit."
7.1.7 Equipment Sources
If you don't want to "roll your own" WRT delicate hardware like gyroscopes,
etc., there are places that sell such things that are suitable for hobby
rocketry, or at least the HPR end of it.
The Japanese company muRata (that's the way they spell it on their
literature) makes a very compact, piezoelectric solid state rate gyro
that measures only 25 x 25 x 58mm and weighs only 45g. It is extremely
resistant to temperature, shock and noise. The down side is that it runs
$200 and that's for only one axis! If you're still interested, contact:
muRata Erie North America
2200 Lake Park Drive
Smyrna, GA 30080
(800) 831-9173
--------------
found: "In reference to the many recent discussions about the hand made
gyro published in HPR magazine and the pendulum guidance systems, here's
what appears to be an excellent source of guidance components small enough
for HPR projects:
Humphrey Inc., Dept. CA391
9212 Balboa Ave.
San Diego, CA 92123
Ph. (619) 565-6631, Fax (619) 565-6873
Humphrey manufactures a full line of guidance system components for use
in rockets, missiles, target drones, etc. Their product line includes:
"gyroscopes, vertical indicators, north seekers, rate sensors, position
transducers, accelerometers, pendulums, magnetometers, directional surveyor
systems"
As an example, they offer a 2.3 inch diameter by 3.25 inch long 2 axis
spring driven gyro. It operating time is 60 seconds, minimum, which should
give you plenty of time up to burnout. This gyro is shock proof up to
85 gees, 10 msec (all axes) and weighs 345 grams. It's output is via
potentiometer pickups.
Using this 2-axis gyro would have solved much of the difficulty encountered
by the hand made gyro team in the above referenced HPR magazine article.
Note that this will probably be *real* expensive -- all military spec. stuff.
A cheap gyro supplier that may be more suitable for HPR low-budget projects
is:
Gyration, Inc.
Saratoga, Calf.
(408) 255-3016
Gyration makes small and inexpensive (about $500.00) gimballed and single
axis gyros. I have no direct information this."
--------------
While on the subject of cheap precision parts, we have an enthusiastic report
Chicago is very possibly the most awesome store on Earth. Where else can you
get WW2 aircraft gyros, 3'x3' fresnel lens, a Tesla coil, the guidance system
to a heat seeking missile, and countless other bizarre goodies. They have a
70 page newspaper print catalog, which is revised every month. I think they
will send one free just for asking. American Science also runs two store
fronts. One is located in Chicago proper and the other is in the western
suburb of Genevia."
The address for mail order is:
American Science & Surplus
3605 Howard Street
Skokie IL 60076
(708) 982-0870
7.2 Control Systems
Every rocket has a control system. Unlike Guidance Systems, which affects
the flight path of the rocket, a Control System determines when the various
flight events take place. For a simple model or HPR Lite rocket, the control
system is built into the motor. After ignition, the only flight event to
be controlled is the recovery system activation, so the "control system"
consists of the delay and ejection charges in the motor.
Moving up one notch in sophistication, a simple system for controlling remote
staging is the well know mercury switch/flash bulb combination. At lower
stage burnout, the blob of mercury flys forward against the contacts
completing the firing circuit for the upper stage. The upper stage ignitor
is usually a flashbulb or sometimes an electric match. Power comes from
either a small "button" battery or a capacitor charged up just before launch.
But what if you have lots of events to control? Say for some very high
flying HPR rocket you don't want the parachute to eject at apogee since a
parachute opening at 20,000 feet could drift for miles. OTOH, you don't
want the rocket to build up too much speed falling back down to a reasonable
altitude. You should eject a drogue 'chute near apogee then the main 'chute
at, say, 500 feet above the ground. Tricky.
To achieve this, you need devices called sequencers, which come in many
forms, plus remote activation charges and other devices. Sequencers come
in several basic forms:
7.2.1 Timer - If you are fairly comfortable with the projected flight profile
of your rocket, timers are a relatively inexpensive way to control the
flight. The timer is started by some sort of signal on the pad. Sometimes
this can be the ignition signal, but more often it's some sort of "first
motion" detector which can be a microswitch that senses the launch rod
or a fine wire that is broken as the rocket leaves the pad. As the flight
progresses the timer executes the various functions such as staging,
switching on a payload, firing the ejection charge(s).
The down side of timers is that they are "open loop." This is engineering
term which means that they work independently of the events surrounding
them. If, for example, you get a motor that burns a little "hot" the amount
of coast time you programmed into the timer might not be sufficient and the
rocket might still be traveling at a high rate when it fires the ejection
charge. You can partially "close the loop" by having the timer be started
by a flight event (e.g. a recovery timer started by an inertia switch at
burnout) but you are still stuck with the pre set timing values.
7.2.2 Altimeter - Altimeters can be both a payload and a control system. The
simple ones only record and playback altitude information. These are
described in the "Payloads" section 5.2.2.3. The more sophisticated ones
can actually control events based on the rocket's altitude. This can be more
effective than timers since it's fully "closed loop," i.e. it operate's
based on information coming from actual flight events rather than a rigid
timed sequence.
As an example, you could program the altimeter to turn on a payload at a
certain altitude on the way up, note the maximum altitude and fire the
drouge ejection when the rocket had fallen back 100 feet (to make sure it
had enough velocity to deploy the 'chute). Finally, you could have the
main 'chute deploy when the rocket was back down to 500 feet above the
pad altitude; plenty of time to have the main 'chute deploy but not drift
too far.
7.2.3 Radio Control - There's no substitute for the Mark I eyeball :-) R/C
controllers allow an operator on the ground to execute flight events based
on observations of how the flight is progressing. The most common use for
this is as a backup recovery system activator. If the standard recovery
system doesn't deploy when you expect it to, you can hit the button yourself.
Some folks even use this for the primary recovery activation. While it sounds
good in theory to have this kind of ground control over your rocket (very
James Bond-like) it takes nerves of steel to allow it to fall and resist
the temptation of "punching out" early.
An R/C system is relatively simple consisting of a receiver and an actuator
of some sort. This actuator can be either a mechanical servo to physically
activate a recovery system or perform other functions, or an electrical
signal initiator (such as used with the timers and altimeters) to fire
pyrotechnics. The ground based transmitter completes the system.
7.2.4 Equipment Sources - Since most of these control systems have made their
way into the rocketry mainstream, the products are much more highly developed
than the Guidance System parts previously discussed. Thus, the addresses of
the following can be found in the main address section (in Part 1). These
companies are listed alphabetically:
Adept Rocketry - Adept has probably the widest selection and best developed
line of hobby rocketry controllers available. They have both altimeter and
timer based controllers in addition to payload style altimeters.
Countdown Hobbies - Countdown is not a manufacturer, but is probably the
most varied retailer of hobby rocketry supplies. You can find everything
listed in this section and, as they say, much, much more!
Designex Corp - Bill Schaffer has just introduced a small (2" x 4") timer
with a range of 3 to 60 seconds.
Pratt Hobbies - Doug Pratt has recently introduced his ECS-2 radio controlled
recovery system. Intended primarily as a backup, it can also be used for
primary recovery. The advertised range is 5,000 ft and it will fit in body
tubes 2" dia and up.
Robby's Rockets - Robby's doesn't make control systems, but they are one
of the largest suppliers of the secondary items needed to make them work;
specifically flashbulb ignitors, thermalite and stand-alone ejection charges.
crystal controled, 4 channel timer which fits in a 29mm tube. The timer was
flown on a "P" powered flight at LDRS and worked perfectly. Previously
sold through MicroBrick (now MRED).
Transolve Corp. - While concentrating mostly on payload style altimeters,
their high end models have expansion ports which allow integration into
other control systems.
7.3 I keep launching my rockets completely out of sight! Are there any kinds
of Recovery Aids to get them back?
7.3.1 Tracking Powder - The simplist and cheapest (although messiest) way
to help spot your bird is to use tracking powder. The most commonly used
materials are carpenters' chalk (available at building supply and large
hardware stores) and tempura paint (available at art supply outlets).
The use of tracking powder is required for all NAR altitude events.
The concept behind tracking powder is simple. At ejection the charge pushes
out the powder along with the recovery device where it disperses into a
large cloud that is much more visible from the ground than your tiny rocket.
Of course the same turbulence that disperses the powder also smears it all
over the body tube and fins, but it's a small price to pay for getting your
model back!
The most popular colors, because they seem to be the most visible under the
greatest variety of conditions, are bright red, orange and pink. You might
think of just using an extra large slug of the talc you're already using to
keep the chute from sticking, but white is only really visible in a dark
blue sky (polarized sunglasses help a lot) and most skys have some degree
of haze, high clouds or smog to lighten them to the point where white is
useless. Dark colors don't do so well under most conditions. Blue is out,
for obvious reasons, but black could be used on an overcast day or against
a smog brightened sky. OTOH, red works pretty well against overcast and
there's no need to carry a bunch of different colors in your range kit.
The amount of powder to use varies by the application. A typical model
rocket, like an Estes Commanche III, might only need an ounce or two to
make a cloud visible from it's 2,000 foot max altitude, but HPR birds going
for altitude records at 25,000 feet or more can use several pounds! Some
flyers just pour the stuff down over the recovery system just before popping
the nose cone on. Others, trying to hold down on the mess, wrap the powder
in little pouches of recovery wadding or some such in the hope that it will
be blown free before releasing the messy cloud. The risk here, of course,
is that the pouch might not open at all if there isn't enough turbulence
(say, if it's ejected right at apogee). Some even use the parachute itself
as the pouch, in a sort of compromise.
7.3.2 Sounders - The next step up in sophistication is an audio sounder which
usually some sort of piezoelectric device which puts out an incredibly
shrill, piercing tone at the upper end of the human hearing range. This
is chosen because 1) the higher frequency carries farther, and 2) it's
not likely to be confused with any other background sound.
A simple beeper can easily be designed and assembled from materials found
at any electronic hobbiest store (e.g. Radio Shack) and some may even
have kits available that are easily adaptable. Those that want a purpose
designed device, however, still have several to choose from. Estes has the
"Transroc II" which is a combination of a sounder (sized to fit in BT-20)
and hand held directional microphone/amplifier with a narrow-pass audio
filter. Note: this item has been dropped from the current catalog, although
some should be available in hobby stores and other outlets for a while yet.
Going up the scale in size and volume, Adept has a beeper for 1.5" and
larger body tubes which is available in either kit or pre-assembled form.
And finally, for the truly impatient, LOC/Precision sells a "snap 'n go"
beeper that's completely pre-assembled and designed to attach to the back
of a nose cone or recovery lanyard. It will fit in their 2.14" and larger
airframe tubing.
7.3.3 Transmitters and Homing Beacons - This is the top of the line for
sophistication in getting your rocket back. If you check back in Trans-
mitter Payloads, Section 5.2.1, you'll see that many data transmitters
also have a simple "beacon" mode for helping locate your bird with the aid
of a directional antenna. To date there's only been one manufacturer who's
come up with a highly optimized system for rocket location; and even at
that, it's an adaptation of one used in the R/C airplane hobby successfully
for years.
Built by Walston Retrieval Systems, it features a 20 mile air range, 2
mile ground range, 4 gram launch weight (including battery) and a 3 element
directional antenna on the ground receiver. This specialization doesn't
come cheap, though, as these systems are in the $200 range. Still, if you
are going to sink a Kilobuck into that altitude record bird, it's nice to
be able to find it later!
7.3.4 Equipment Sources - The addresses for Estes, LOC and Adept can be
found in the main address section. The address for Walston (which offers
a free catalog) is:
Jim Walston Retrieval Systems
725 Cooper Lake Rd., S.E.
Smyrna, GA 30082
Ph: (404) 434-4905
or (800) 657-4672