Descent rate
Altimeter


Overview

The descent rate altimeter is primarily used as a backup altimeter.

It monitors the altitude. If it has not changed in 8seconds it stores this in the internal EEPROM as the ground altitude.

Upon power up the ground altitude is read out of EEPROM.

Every second the device calculates the descent rate using the previous seconds readings.

The testing is very simple. There is no "launch detect" as other altimeters have. This device simply looks at the descent rate and tries to reduce it.

If the descent rate is over some value, this preliminary value will be 120ft/sec, it deploys the drogue chute.

If the altitiude is less than the main altitude(compensating for ground altitude) and the descent rate is over some value, the preliminary value will be 20ft/sec, then it will fire the main deployment charge.

Very simple robust logic. It will make for a perfect backup altimeter.


What is a decent descent rate ?

Here is a little snip of a conversation with Dean Roth about descent rate.

I assume you mean, "What would be the descent rate if the rocket failed
to separate into two sections at apogee and came down nose first?" That
is answered by this:

Vt = sqrt( 2W/(Cd * rho * A))

Vt = terminal velocity in feet/second
W = weight in pounds
Cd = drag coefficient - (I assume 0.55)
rho = atmospheric density (about 0.0024 at sea level)
A = surface area in square feet

Vt = sqrt( (2*30)/(0.55 * 0.0024 * 0.196)) = 481 ft/sec

Assuming I've not made any terrible errors, the terminal velocity of my
rocket making a nose dive for the ground would be approximately 480
ft/sec.

A 100 - 120 ft/sec descent rate threshold while above the low altitude
setting is a reasonable value for most cases. Calculations that I have
done in the past indicated that most rockets, if they fall as two
tethered sections, will descend at 40-80 ft/sec without a drogue.


Flow Chart

Main routine

25mS Timer routine

1S Timer routine

8S Timer routine


Images


Connection

This design will take one controller board and a sensor board with just the pressure sensor mounted.



Mail to Robert DeHate