MAD alternative?

[vjp]

I've been thinking about building a MAD detector for a while now, but was
wondering if it wouldn't be cheaper and just as reliable to use a pair of CDS
photoresistors, once facing up, the other facing rearwards, in a voltage
divider configuration. Feed this into the op-amp (configured as a comparator)
with the reference voltage set to half V+.

When the rocket reaches apogee and begins to turn downward, the rear-
facing CDS sensor now "sees" more light than the forward facing CDS, and
the voltage at the op-amp input goes low (or high) compared to the reference
voltage, triggering the ejection circuit.

Obviously, there are probably some unusual weather/terrain lighting conditions
that could cause it to "see" a brighter ground than sky (i.e. a dark cloud
overhead and sunlit sand or snow on the ground), but I would think these
conditions would be the exception rather than the rule, in most parts of the
country and in most launch conditions.

Anyone ever tried this?

 

[rstaff3]

Rather than get MAD, I prefer to get even!

I've never heard of your approach, but it might work. Sounds like some experimentation is in order. Make sure you have an interlock so it doesn't fire the charge on the way to the pad

 

[vjp]

Originally posted by rstaff3
Rather than get MAD, I prefer to get even!

I've never heard of your approach, but it might work. Sounds like some experimentation is in order. Make sure you have an interlock so it doesn't fire the charge on the way to the pad

Definitely... first though, I think I'll build a test article, a piece of BT with the circuit connected to a piezo buzzer. Then, take it outside and test all sorts of different orientations to find out just how reliable it is, under various sky/lighting conditions.

 

[rstaff3]

Yup, definitely what needs to be done. Gotta test for both the ability to reliably sense the different light levels and avoid false alarms.

 

[bobkrech]

VJP

Futaba makes a 4 photodiode horizon sensor looking sideways to recover RC airplanes into a straight and level configuration.

Someone (Gassaway?) has used this unit to keep a scale model rocket vertical in powered flight by sending the output to servos that moved a gimbeled motor. The article appeared in Sport Rocketry about 3 years ago.

A potential problem you might encounter is if there was a reflective object below the rocket and you get a reflection or glint off that object into the lower sensor. That would trigger your deployment system.

You will have to do a lot of testing on this approach to make it failsafe. It may work but I think you will need a fair bit of extra electronics to prevent false triggering. I don't think that the finished electronics package would have a significantly different cost than a MAD detector.

Bob Krech

 

[rstaff3]

People have also made those 'sun seeker' rockets that would head towards the sun. You might be able to get by with one sensor as long as its a sunny day and your firing at high noon


BTW, I was being somewhat facetious...

 

[vjp]

I'm familiar with futaba horizon sensing autopilot, but I haven't looked at one to see what sort of sensors it uses. Do you know if this unit uses CDS photocells? Something else? Regarding the sensitivity to brief glints, etc. - yes, I 'm using a capacitor to filter out transients, so I don't think that will be a major problem.

So far, I have been somewhat successful in my breadboard testing. It works amazingly well in low light, but yesterday when I tested it in bright sunlit conditions, I couldn't get it to "flip". I think I know the problem, though - I'll do some more tinkering this weekend if I get a chance.

I don't think the PAD could replace the MAD as the apogee detector of choice, but it might be a practical alternative in some situations. The main disadvantage that the optical 'window' is an integral part of the airframe, making it impractical to move it from one rocket to another (unless it was built as an interchangeable "payload section" that could be swapped among different bodies based on the same BT). But the cost advantage (total parts cost <$10) might make it attractive, nonetheless.

 

[bobkrech]

Tower Hobbies sells the Futaba PA-2 Pilot Assist Link Auto Pilot which is a "Horizontal Stabilizer For Model Aircraft" and incorporates "the latest Optical Sensor Technology to read the difference between light (sky) and dark (horizon), then automatically adjust pitch and roll to stabilize the plane".

https://www2.towerhobbies.com/cgi-bin/wti0001p?&I=LXNA35&P=7

https://www.futaba-rc.com/radioaccys/futm0999.html

It doesn't say what type of photosensor it uses but I assume its either 4 silicon photodiodes or a quadrant photodiode sensor.

Check out the Q&A links on the above webpages for more info and restrictions.

Bob Krech

 

[vjp]

Thanks for the links, Bob - I looked at the specs, and I think I understand how it works. The sensor is mounted looking down, and compares brightness in 4 different directions, and probably interprets a brightening in one direction vs. the opposite, as meaning that the airplane is tilting "up" (i.e. seeing sky) in that direction, to apply a corrective force to the plane. So, it's not usable under certain lighting conditions (over water, snow, etc.), same as the limitations I'm expecting for the PAD.

Dick - Since you're into micro hybrids, I suspect you probably use a MAD yourself? If so, do you know if the 1000mfd cap in the design will fire xmas-bulb ejection charges? I don't have any flashbulbs or e-matches (and no LEUP, which rules out e-matches for now anyway). I can't find anything over 1000mfd at Radio Shack without going WAY up in size.

 

[rstaff3]

Vince, Sorry, I don't have a MAD. I've only flown my micro hybrid in saucers.

 

[vjp]

Originally posted by rstaff3
Vince, Sorry, I don't have a MAD. I've only flown my micro hybrid in saucers.

Doh! Darn - oh well, I'll eBay up some flashbulbs then! I know they'll work (besides, my wife will shoot me if I raid the xmas lights for experiments).

 

[Ray Dunakin]

One possible problem you should consider is the glare from the motor's exhaust flame, especially if you use an Aerotech White Lightning motor or any similar propellent formula. They produce a great deal of light. I've had problems with my camera payloads because of this --- with auto-exposure, the camera often over-compensates for the glare of the flame, and everything else comes out much too dark. You sure wouldn't want you ejection charge to fire when the motor ignites.

 

[Missileman]

To overcome the "rockets red glare" as it were
Perhaps you could use some type of delay device.
Such as maybe a g switch like the ones used in some staging timers.
After motor burnout the rocket experiences rapid deceleration and as a result payloads will, at least momentarily, be in a negative G situation.
Some type of simple latch circuit would then keep the system active after the initial neg. g event is detected.
And gates, nor gates oh oh here comes the headache!!OUCH