Optimum Altimeter Design (Baro + MAD?)

There are alot of good altimeter designers in this group so...
I am designing a simple dual deployment altimeter, reliability is key. I've looked at baro, accel, baro + accel etc. This is alot of good work on algorithms to detect apogee with all including simple to kalman fiters and the like. It seems like more complexity that is needed.
I am thinking that a baro sensor for altitude deployment, backup apogee and arming coupled with a magnetic apogee detector for apogee is the simplest and most reliable solution.
I am open to arguments contrary to this... Comments?
jd

jdMARS wrote:
  Quoted Text Here  
Whenever someone uses the word optimum, I always remember that there are many different versions of optimum. So you need to define it first. For example: the Kalman filter is optimum in a least squared error sense. There are other filters including one which minimizes the maximum error. It all depends on what you want.
I see that you mention simple and reliable but don't elaborate.
  Quoted Text Here  
While the math behind the Kalman filter is complex, the implementation is not.
  Quoted Text Here  
Using a magnetic attitude indicator has its own problems. One that few think of is the problem of launch rods/rails made from magnetic materials. Even unmagnetized iron pipe will alter the local field significantly.
The biggest problem with using a magnetic sensor is that it requires user adjustment for just about every flight. This is bound to be forgotten or just plain messed up at some point. While the circuit is simple, its operation is not.
Barometric altimeters are pretty simple but they run into problems during high speed flights. The usual workaround is the Mach inhibit timer. This is simple in concept but introduces operational complexity. It requires the flyer to know what time to use. Believe it or not I saw a level 3 certification flight shred spectacularly because the flyer had set the Mach inhibit timer to the burn time of the motor.
So do you want a simple circuit or simple operation? My favorite altimeter is the AltAcc because it is exceedingly simple to use. It also hasn't failed me (except in the inherent limitations of the acceleration algorithm) since I bought it in 1999. Sounds like simple and reliable to me.

By simple I mean the simplest design that peforms the mission.
  Quoted Text Here  
minimum amount of information content will be easier/quicker to develop, minimizes testing required and generally will be more reliable.

(Snip for brevity)
  Quoted Text Here  
Am no expert but what adjustment do you mean? I built one of the Aerocon units and just had to play with one resistor to get the appropriate deploy signal. I haven't launched it yet. I was aware of the problem with magnetized rods but figure I could put a window in the model and have the test led plugged in. Could run it up and down the rod and see if it goes on. This may or may not be accurate if the velocity of the model through the magnetic field might be an issue and I defer to others smarter than myself. It has been written to mount the MAD unit away from the launch lug or the side facing away from the rod.
One way around this would be to just use it with an aluminum rail or if one could find a strong enough aluminum rod.
The Aerocon unit is so small I think one could use it with another altimeter of choice if desired. It is a bear to build but I was prepared with dual magnifiers and a decent soldering station with very fine tips. Was rewarded when it worked right of the bat though.:)
Best regards, Kurt Savegnago

There is no user adjustment for every flight with the Philips KMZ51 or Honeywell HMC1001 devices. Probably degaussing the chip is a good idea.
I was thinking of still using a uP for the altimeter. I would still use the baro sensor to "arm" the device at a safe altitude prior to which the uP would degauss the sensor. Thus we do not have to material about ferro materials around the pad, we won't even look at the magnetic sensor till well after the pad is cleared.
The only problem I see is the case if the rocket doesn't tip over, if it backslides all the way down then the magnetic sensor won't fire, thus using the baro as the backup apogee (past) detector..
jd

  Quoted Text Here  
My 2 cents would be to use accelerometer with baro backup for apogee and use the baro for main.
the majors do this, they must have a reason.

There is no user adjustment for every flight with the Philips KMZ51 or Honeywell HMC1001 devices. Probably degaussing the chip is a good idea.
I was thinking of still using a uP for the altimeter. I would still use the baro sensor to "arm" the device at a safe altitude prior to which the uP would degauss the sensor. Thus we do not have to worry about ferro materials around the pad, we won't even look at the magnetic sensor till well after the pad is cleared.
The only problem I see is the case if the rocket doesn't tip over then the magnetic sensor won't fire, thus using the baro as the backup apogee (past) detector would be prudent.
jd

Kurt wrote: > (Snip for brevity) > > > >> Using a magnetic attitude indicator has its own problems. One that few >> think of is the problem of launch rods/rails made from magnetic >> materials. Even unmagnetized iron pipe will alter the local field >> significantly. >> >> The biggest problem with using a magnetic sensor is that it requires >> user adjustment for just about every flight. This is bound to be >> forgotten or just plain messed up at some point. While the circuit is >> simple, its operation is not. >> > > Am no expert but what adjustment do you mean? I built one of the > Aerocon units and just had to play with one resistor to get the > appropriate deploy signal. I haven't launched it yet.
I am pretty sure that all variations on the MAD unit use one of the magnetoresistive sensors. When exposed to a high strength magnetic field they will be effected. Therefore you have to activate the set/reset coil built into the sensor to remove this effect. After you do this you should then adjust the trip point so that the unit activates at whatever angle you deem appropriate. As the sensor offset and sensitivity can change with time, this should be checked before flight. Another problem is that the angle of the magnetic field varies depending on your location.
The only circuits I have seen use a variable resistor to set trip point.
> I was aware of the problem with magnetized rods but figure I could put > a window in the model and have the test led plugged in. Could run it > up and down the rod and see if it goes on. This may or may not be > accurate if the velocity of the model through the magnetic field might > be an issue and I defer to others smarter than myself. It has been > written to mount the MAD unit away from the launch lug or the side > facing away from the rod. > > One way around this would be to just use it with an aluminum rail or if > one could find a strong enough aluminum rod. >
The data I have was recorded with an HMC2003 three axis sensor. It was located in the center (more or less) of a 4" airframe. The launch rail was the Blacksky model rail reinforced with iron pipe brought to the launch by one of the flyers. Why this person decided to build the rail that way rather than just buy the high power rail is beyond me.
> The Aerocon unit is so small I think one could use it with another > altimeter of choice if desired. It is a bear to build but I was > prepared with dual magnifiers and a decent soldering station with very > fine tips. Was rewarded when it worked right of the bat though.:) > > Best regards, > Kurt Savegnago

David Schultz wrote: > I am pretty sure that all variations on the MAD unit use one of the > magnetoresistive sensors. When exposed to a high strength magnetic > field they will be effected. Therefore you have to activate the > set/reset coil built into the sensor to remove this effect. After you > do this you should then adjust the trip point so that the unit > activates at whatever angle you deem appropriate. As the sensor offset > and sensitivity can change with time, this should be checked before > flight.
All true, in principle, but in practice I've rarely had to hit the reset button, and between flights at the same field have never changed the trip angle. (This is with the MAD which uses a Phillips KMZ11 sensor.)
... which is not to say I ignore these issues. I do test the instrument before each flight, using an LED in place of an electric match and tilting the payload compartment and observing where the LED lights. I also take care to avoid launching from a steel launch rod.
> Another problem is that the angle of the magnetic field > varies depending on your location.
Yes. When I go to a launch at a significantly different latitude I find I need to adjust the resistance. (With the MAD units this means resoldering a new resistor, unfortunately.) However, I've only had to do this twice in the past six years.

David Schultz wrote:
  Quoted Text Here  
This potential problem is discussed in the instructions that come with my MAD kit. Perhapd it should be emphasized a bit more.
There's a good chance that iron pipe is magnetized, although not intentionally.
  Quoted Text Here  
My MAD units do not need adjustment for every flight. Simply hitting the reset button will bring the unit back into full sensitivity and original offset.
  Quoted Text Here  
Neither my original design (as shown in September 1999 Sport Rocketry) nor the current kit (as sold by Aerocon and others) uses a variable resistor. I have never seen the offset and sensitivity vary from their initial values.
The angle of the Earth's magnetic field does vary with location on the Earth, but, with the recommended calibration, the unit should function fine in all locations. What will vary with location is the angle off of horizontal that the unit triggers when launched to the north or south. To the east or west, the unit will trigger at horizontal in all places.
  Quoted Text Here  
The MAD operation does not depend on velocity, so sliding the model along the rod would show if there is a magnetic launch rod problem.
  Quoted Text Here  
I could see where that could cause a local magnetic field anomaly...
  Quoted Text Here  
- Robert Galejs