Jun 12, 2001 Meeting Notes

 

In Attendance:

 

John Carmack

Phil Eaton

Russ Blink

Bob Norwood

 

New Supplies

 

1500 psi regulator

Pressure relief valve

½” quick connects

Etched sheet steel

AMP multimate crimper

AMP CPC connectors for new electronics box

 

Gyros

 

The fiber optic gyros from KVH finally arrived, but unfortunately, the demo kit only included 10hz digital output, which isn’t fast enough for us. They have agreed to swap the units for analog output ones, but I don’t have a firm delivery date.

 

I did learn something important playing with one of the FOGs:

 

One of the things that I had been disappointed about with our previous gyros was that the highly scientific test of wiggling the sensor board around by hand would usually result in a significant drift in the integrated positions when it was set back down in it’s original position.

 

I had been assuming that the problem was due to the low bandwidth of the gyration gyros causing it to integrate poorly as the rate was changing.

 

I was surprised to see the same thing happen with the FOG. If you leave it on a flat surface and rotate it back and forth, the integration performs very nicely. However, if you pick it up and wiggle it around, then set it back down in the original spot, there will usually be a significant error in the integration.

 

I should have realized this at the beginning, but it turns out that you can’t integrate the axis independently. The extreme example is this: with the gyro flat on a surface, rotate it clockwise until the integrator reads 90 degrees. Then tip it up towards you, not changing the integrator at all, because it is an orthogonal axis. Then tip it to the left, again on the orthogonal axis. Then lay it back down in its original position. The integrator still reads 90 degrees, even though it is back in its original position.

 

At small angle changes the effect is much less significant, which is how we have gotten away with it so far, but once we start wanting to tip it aggressively for translational maneuvers, we are going to need to deal with it correctly.

 

The right way to deal with it is to maintain a full coordinate system, and rotate it each sensor sample, instead of integrating. I will also have to give some thought to the engine controls, given that the joystick angles are going to be in “absolute world angles”, which will have to be transformed into the local coordinate system.

 

The FOGs are also limited to 100 degrees / second, down somewhat from the 150 degrees / second of the gyration units. 100 d/s is not all that fast, but should suffice for us. We won’t ever intend to rotate even half that speed, but once we start flying faster, the aero loads may push it around more. I think the trick will be to be correcting very rapidly, so that even if a strong force starts rotating the vehicle, it won’t have much chance to build up rotational speed.

 

 

Manned Vehicle

 

Bob has one quarter of the frame tacked together, and should finish it up in a week or so. It is 13’ in diameter, but the legs disassemble so it can be moved in a normal truck.

 

It seems to be a truism that things always weigh more than you expected. Once everything is all put together, the vehicle is probably going to be 150 pounds dry. With a light pilot, five gallons of peroxide should give 15 seconds of flight.

 

Our second motorized valve for the master cutoff switch is on the way, and Russ is about to start machining the custom manifolds we need, so we should be able to get the complete plumbing together in a couple weeks, with Juan’s motor standing in until we build our big one. I am probably going to need to order some more –6 hose

 

We are still discussing pilot facing tank versus pilot facing away from tank.

 

We are still discussing what the pilot will use for throttle control.

 

We are going to leave the bottom frame cross brace off on the pilot corner, so a bailing pilot won’t have anything to hit. The platform is plenty stiff to not be impacted by removing it, and it will still have the pilot’s standing platform by the tank providing some bracing.

 

 

Motorized Valves

 

The term “servo valve” is a little bit ill-defined. It is sometimes used to refer to any DC motor valve, but seems to be more properly used to refer to a valve that has an internal closed loop control system, like model airplane servo motors.

 

The kzco valves we are using are probably better referred to as “dc motor valves”, because all they have is two power wires, which you can put current through either way, and a potentiometer for feedback. This involves somewhat more work on the flight computer, but is probably better in the long run, because your telemetry stream can have where the motor actually is, and what you are trying to do about it, instead of where you are telling the servo is should be.

 

If you just apply power to the motor, it gets up to speed very rapidly, but if you remove current, it takes a considerable time (quarter second or so) to coast to a stop. If instead of just removing power, you short the leads together, the motor acts as it’s own brake and stops as rapidly as it starts.

 

I burned out four solid state relays and two motor driver boards working with the motor, due to a lack of respect for the inductive issues.

 

The driver board has instructions like “for maximum life, slowly increase the motor speed over several seconds, and wait several seconds before changing directions”. Well, when trying to match a valve position to a moving joystick throttle, you wind up running full blast all the time and making a few reverses a second. I fried some things (reminding me why I like software again…).

 

I eventually got it all working reliably by using electromechanical relays, but Phil is going to try building a proper transistorized driver that can do what we need.

 

 

Software

 

RemotePilot.exe now has starts making a loud beep after a certain amount of time has gone by, as a crude “propellant warning”.

 

An updated Lander3D.exe is available at: media.armadilloaerospace.com/misc/Lander3d.exe

 

GraphLog is updated for three-axis display.

 

 

Next Vehicle

 

We talked a bit about where we are going after the first manned vehicle. The basic goals will be:

 

Streamlined.

 

Mass ratio of two or better. This will be the first time we are actively trying to build light.

 

Supine pilot. It will pull several Gs.

 

Pressurized cabin. It won’t be going to space, but it will be going high enough to want pressure.

 

One question is if we are going to make another small vehicle, or if all future vehicles will be capable of holding a pilot (they will all still be flown by remote control during testing, of course). Small vehicles would be easier to transport and would use less peroxide, but they won’t really be much easier to build, and even if they are built, they won’t answer all the questions that manned versions will need to address.

 

The plumbing will probably be identical to the current manned vehicle, it will just be different tankage, structure, and a somewhat larger lifting engine.

 

The FAA issues will have to be addressed.

 

 

Master To Do List

 

Slightly drill out demonstrator engines to account for the extra weight of the new electronics box.

 

Experiment with slower pulse rates on the VTVL, because the bigger valves may need to be slowed down.

 

Finish new electronics box

FOG integration

Make a milled mounting corner for FOGs?

PC104 SSR board

Better 5V power regulation?

 

 

Manned vehicle issues

Closed cell foam for base

Motorcycle throttle potentiometer

Three custom manifolds

Tank cleaning and passivation

get 35% peroxide locally

Pump for peroxide loading instead of vacuum

Buckle the pilot in?

Pilot communication headset

 

 

50lb thrust engines

Plug nozzles?

Injector plates?

Better O-ring seals?

External threads?

Long burn tests.

Pack wear tests.

Good attack/decay tests for pulses.

High pressure tests

 

 

NPT burst discs, both 1/4" and 1/2". We may want both 600 psi burst and 1500 psi burst.

Or are we ok with the relief valve?

 

Gyro angle initialization by accelerometer gravity vector

 

Metal line for test stand chamber pressure tap

Pack transducer with brake fluid?

 

Cooling vests for hazmat suits

 

Continue pushing Porvair about pure metal foams

 

Follow up with FAA if no response after 30 days

 

Tank level measurement by pressure / temperature calculation.

 

600lb thrust engine

 

Drill ball valve vents

 

Drill out Juan's engine for 150 lb thrust?

 

Flash a light on vehicle every second for telemetry video sync

 

Lifting motor valve driver

Make a PCB for it.

 

Master cutoff valve driver

Basic stamp and flightcom development

Make a PCB for it.

 

Vertical test stand

Clean out Neil's tank, or get another big one from Bob for the test stand

 

Integrate test stand electronics into a single box and cable.

 

55 gallon plastic drums for water at new site

 

 

Ballistic demonstrator flight

Parachute canisters.

Fins.

Ultrasonic / radar altimeter testing

GPS / accelerometer altimeter testing

 

-----------------------------

 

longer term:

 

Ballute experiments

802.11 range extending devices

Hybrid / biprop development

Investigate separate pressurization tanks

Pressurized cockpit capsule

Really fat tanks

Unrestricted GPS

Satellite phone telemetry systems