Entire Forum This board This topic Members Entire Site
 Welcome, Guest. Please login or register.Did you miss your activation email? December 03, 2016, 03:23:59 03:23 1 Hour 1 Day 1 Week 1 Month Forever Login with username, password and session length
 Pages: [1]
 Author Topic: What is the best way to get a reflected pulse off a moving object?  (Read 2627 times) 0 Members and 1 Guest are viewing this topic.
Terry Dactil
V.I.P
Newbie

Offline

Posts: 10

Thank You
-Given: 18

 « on: September 08, 2012, 06:42:49 06:42 »

I want to determine the angular position of a cylinder that has restricted rotation about its axis and regularly and repeatedly passes an observation point where a non-contact or optical sensor can be placed.

(In other words, I want to measure the pitch angle of a blade on a rotating propeller).

I have previously used propeller balancing equipment where an IR transmitter/receiver mounted on the cowling looks at a spot of reflective tape on one of the propeller blades and uses the pulses received to calculate the rpm and a reference datum for a vibration sensor. I thought that some variation of this could be used.

Some measurements showed that when the cylinder (the shank of the blade) rotates its full range (about 50⁰) a spot on its surface had a linear movement of 1.6” as would be viewed by a sensor.  As this spot is also at a 5” radius from the propeller axis, this 1.6” movement becomes about 18⁰ of movement around the circumference of this 5” radius circle.
It is like this:

Therefore, if we get a pulse from a fixed reference spot on this circumference and a pulse from the variable spot on the blade
•   the repetition rate of the two pulses can be used to calculate the rpm
•   the interval between the two pulses compared the repetition rate will give the circumferential  position of the spot on the blade, and then a lookup table can give the blade angle.

I have programmed a 16F685 chip to do the signal processing  and maths required and this is working quite satisfactorily under simulated conditions.

However, I am not convinced that IR sensing a spot of reflective tape in daylight conditions will be totally satisfactory , so I seek your opinions on the best methods I should use to get the required pulses.
 Logged
CocaCola
V.I.P
Senior Member

Offline

Posts: 397

Thank You
-Given: 108

 « Reply #1 on: September 08, 2012, 07:21:51 07:21 »

Possibly a laser module bouncing off the reflective tape?
 Logged
borberk
V.I.P
Senior Member

Offline

Posts: 298

Thank You
-Given: 309

 « Reply #2 on: September 08, 2012, 08:07:04 08:07 »

Can you use something like http://rls.si/en/aksim-off-axis-rotary-absolute-encoder--17584 this?
 Logged
solutions
Hero Member

Offline

Posts: 1446

Thank You
-Given: 587

 « Reply #3 on: September 08, 2012, 08:22:42 08:22 »

I am very curious about why you care about the pitch angle itself?

The beauty of a constant speed propeller is just that - constant speed - and pitch varies with power and altitude.

Even if you're building a prop governor, you don't care about the angle itself in the control laws...you vary pitch and target RPM.

Is the prop oil, ground, or electrically pitched? In those cases, you can determine the pitch angle based on oil pressure, a scribed scale, and an encoder on the motor shaft, respectively.

Posted on: September 08, 2012, 08:19:30 08:19 - Automerged

LOL @ their marketing claims of 10,000 RPM...that's 2.6Gb/s to process off the encoder, assuming their magnetic sensor has that kind of bandwidth (which I doubt)....
 Logged
borberk
V.I.P
Senior Member

Offline

Posts: 298

Thank You
-Given: 309

 « Reply #4 on: September 08, 2012, 11:48:58 11:48 »

As per my informations encoder during rotation make just 3 samples per turn to determine speed and direction. That is calculated in first loop of algorithm. Second loop is intended for precise determination of position. Processor does not need more than 4kb/s. This is absolute encoder and not incremental which would need essentially more bandwidth.
 Logged
Terry Dactil
V.I.P
Newbie

Offline

Posts: 10

Thank You
-Given: 18

 « Reply #5 on: September 08, 2012, 03:53:44 15:53 »

Solutions:
Quote
I am very curious about why you care about the pitch angle itself?
The beauty of a constant speed propeller is just that - constant speed - and pitch varies with power and altitude.
The reason is that it is not a constant speed propeller.

Several friends use converted Subaru car engines in their aircraft, and these are operated in a completely different manner to the “old style” Continental and Lycoming engines.  The throttle is only used when taxying on the ground and on final approach for landing. The rest of the time the throttle is locked in the wide-open position and power is varied by changing the propeller pitch/rpm.

Operating in a this maximum manifold pressure / low rpm regime gives excellent fuel consumption figures, but requires that the propeller pitch angle be known and set to a specific angle before takeoff and on final approach when the throttle is used in the normal manner.  Otherwise a wide open throttle could either result in an rpm overspeed, or not enough rpm and low power.

The standard setup for these engines has the propeller pitch varied by an electric motor fed with DC power via two slip-rings on the spinner backplate. A third slip-ring supplies DC power to a variable frequency oscillator in the propeller hub. The oscillator frequency is controlled by a multi-turn potentiometer on the motor shaft, and the oscillator output is fed back onto its DC power supply. There is therefore an AC signal riding on the oscillator’s power supply, and this is read by an instrument back in the cockpit to display the pitch angle.
Here is the circuit.

Normally this all works very well, but there is one weak point. When the slip-rings get dirty (as they will) the electrical noise produced causes the pitch angle display to become erratic.  Of course, regular maintenance and cleaning will fix the problem, but it is annoying when the display goes erratic in the middle of a flight. It was therefore suggested that it would be nice to have a separate backup indicator that did not depend on slip-rings. Hence this project.

CocaCola:
Quote
Possibly a laser module bouncing off the reflective tape?
Yeah. Considered that as a good idea, but could not find any photo-diodes or photo-transistors to make a receiver that operated at the laser wavelength.  That was some time ago so maybe things have changed now. Any suggestions for suitable components?

We did actually have a laser setup that worked after a fashion.
A laser pointer was aimed at a tiny mirror on the blade shank, and the geometry of the system caused the returned laser spot to move in a radial direction to the propeller disk plane with changes in the blade pitch. This was picked up by a linear array of optical fiber cables and led back to the cockpit where the pitch angle was determined by which optical fibre in the array showed the laser spot.
This had a fundamental flaw in that the mark-space ratio of the laser spot was something like 1:1000 on to off. It worked fine in the dark but was pretty useless at other times. This could have been fixed with a sample-and-hold circuit, but once again the lack of something suitable to respond to the laser was an insurmountable problem at the time. Perhaps better luck this time.

 Logged
RaynW
Newbie

Muted
Offline

Posts: 7

Thank You
-Given: 7

 « Reply #6 on: September 08, 2012, 04:45:28 16:45 »

I don´t know how aggresive the environment is, but optical sensors usually don´t exhibit a long-term mainteinance-free operation. If there is oil, dust, etc probably you will experience similar problems as with slip-rings.
I think a Hall-effect proximity sensor is a better choice. Some links:
http://www.hallsensors.de/Hall-Switch.htm
http://www.instronics.com/sensoronix_hall_effect_proximity_switch_sensor.html

Anyway in case you prefer the optic sensor, can you arrange the mechanical design to use an interrupted light-barrier rather than reflective type? Usually this works better.

Regards.

Edited (because I remembered something, much later. Sorry):

Your solution is very good, indeed. But I can´t figure out how you will know the blade angle if the propeller isn´t rotating (perhaps, it´s not necessary). Anyway, if your problem is bad contact on slip rings, you can read more about these mercury filled ones:

http://www.mercotac.com/

I used them at work, and never had slip ring wearing, dirt or similar problems again. Just one drawback (besides they are really expensive), the model I used can´t be mounted at any position. But perhaps there is one that fits your needs.
 « Last Edit: September 09, 2012, 12:41:04 00:41 by RaynW » Logged
solutions
Hero Member

Offline

Posts: 1446

Thank You
-Given: 587

 « Reply #7 on: September 09, 2012, 02:49:19 02:49 »

Your problem is getting FM encoded position data across the slip rings, not determining the pitch angle. Define the problem, then solve it.

A transmitting device (IR LED, magnetic, string and tin can, Tesla coil with a jacob's ladder using the prop blades for electrodes, spark gap, etc) can be attached to the FREQ OUT pin and then be read without going through the power rails or slip rings.

The impedance of freq out looks like it's low enough to drive an LED very nicely, for instance, and you know the prop is not going to turn more than 60 or 70Hz, so a high pass with a cutoff at say 150Hz will get rid of any sunlight or revolutions related artifacts. On the fixed side, you can have an interrogate beam, which triggers a few cycles of pulses coming in,to save power, ease synchronization, and just be a cool way to do it.

I'd even think about putting a low power, low frequency, radio on the prop side and send modulated RF (from freq out) over to an RF receiver less than a foot away.

The EASIEST way, though, is to just crank up a carrier frequency and jump the brush/slip ring using capacitance instead of a DC connection.

In all cases, measure the SNR and illuminate a "clean the f'ing sensor you lazy bast&rd" LED on the panel when it degrades due to dirt or contamination.
 « Last Edit: September 09, 2012, 02:52:10 02:52 by solutions » Logged
borberk
V.I.P
Senior Member

Offline

Posts: 298

Thank You
-Given: 309

 « Reply #8 on: September 09, 2012, 06:28:30 06:28 »

Agree with solutions about what is the problem. It is data transfer between two moving object. Proposed capacitive coupling for RF transfer is one way the other is inductive between two ring shaped coils. They form a sort of transformer with prop axe in center.
 Logged
Terry Dactil
V.I.P
Newbie

Offline

Posts: 10

Thank You
-Given: 18

 « Reply #9 on: September 09, 2012, 08:51:51 08:51 »

RaynW
Thanks for the links to the Hall effect sensor info.  I was not aware that the gap to the magnet could be as much as 20mm or so.  I thought they had to be in very close proximity to work. However, the closest anything fixed to the engine can get to the propeller blade is about 2” (50mm). Therefore the obvious place to place the magnets would be on the rear face of the spinner backplate and with some sort of linkage from the propeller blade to change the position of one of them. That could certainly work.

Solutions:
You have the perfect username. Your analysis certainly found the problem and provided a solution.
There is a terminal block on the front face of the spinner backplate  so there is easy access to power and the oscillator output . I’ll try your ideas. Many thanks.

A bit more of the history of this project may be of interest.
The original design process (if it could be called that) was a conversation over a cup of coffee bitching about the erratic display and this progressed to “…. wouldn’t it be nice if we had….”  which then became a bit more structured, and finally became
1.   Make an alternate separate and self-contained method to measure the prop pitch angle
2.   It does not have to work if the propeller is not rotating
3.   No machining or other alterations to the propeller structure
4.   Don’t use f’ing slip-rings!
As I had just become able to program those newfangled PIC chips, the old adage “if you have a hammer – everything looks like a nail” applied, and this project just had to include a PIC!
Thus the idea of just sticking a couple of spots of reflective tape on the prop and then demonstrating my programming virtuosity was born.

 Logged
Ichan
Hero Member

Offline

Posts: 840

Thank You
-Given: 312

 « Reply #10 on: September 09, 2012, 07:28:04 19:28 »

I have programmed a 16F685 chip to do the signal processing  and maths required and this is working quite satisfactorily under simulated conditions.
However, I am not convinced that IR sensing a spot of reflective tape in daylight conditions will be totally satisfactory , so I seek your opinions on the best methods I should use to get the required pulses.

Maybe the IR sensor can be used in linear mode (not on/off), the output from it feed to an adc. This way we can have both "dark" value and "light" value, so as long as there is still a good contrast between dark and light then the environment effect (ambient light, dust, etc) can be significantly reduced. Apparently it will need a high speed adc, some PIC has it - up to 2Msps I think. Just a thought.

-ichan
 Logged

There is Gray, not only Black or White.
cranor
Newbie

Offline

Posts: 17

Thank You
-Given: 8

 « Reply #11 on: October 14, 2012, 01:39:04 13:39 »

As dungelectronic, I would recommend you a modulated IR beam.

The infrared emitter modulation is easiest part.
For reception, there are many receivers (for remote control) using 38kHz (or more) modulated infrared light.
But you could encounter problems with the AGC and other in-chip integrated filtering depending on the RPM.
For example : too slow, the AGC will have time to adjust its gain and detect the signal even if the IR beam is pointing to the "hole".

Maybe you could adjust emitting IR's power to be near the low threshold of receiver sensitivity (or add a filter in front of the receiver to reduce receiving signals power).
Or, maybe, you can just have to consider signal losses.

The output of a receiver like the TSOP1838 is pull-up'ed and would go high when no signal and low when the 38kHz carrier is detected but may need some filtering (or maybe not.. you would easily have an answer if you try with a scope..)
I recommend you to try this solution, as TSOP family receivers costs nothing.. take a look at eBay. You may find equivalent parts (VS1838B..)

But you will have to take into consideration that 38kHz burst have a minimum length and a minimum delay between bursts (due to AGC's fixed parameters. the datasheet will give you all information).

I first talked about the TSOP1838 because I already used them for robot detection.

But a more powerful solution could be to use a Hamamatsu sensor.
They have the proximity sensor S6986 that modulates the emitting diode and, if I remember, well, it gives you an analog output so you can set the detection threshold by yourself using a comparator (standalone or in-µC like a PIC18F2550 or other).
As the emission and reception are synchronized, this sensor could give better results than the TSOP family which is not really made for this kind of applications).
I used this component once, but with static objects.

I am sure that read both components' datasheets will help you.

Regards,

cranor
 « Last Edit: October 14, 2012, 01:41:25 13:41 by cranor » Logged
Gallymimu
Hero Member

Offline

Posts: 577

Thank You
-Given: 98

 « Reply #12 on: October 14, 2012, 04:13:58 16:13 »

I would use the magnetic prox sensors.  As was mentioned, depending on the type of metal used in the sensed element range can be a few mm to 10 or 20 mm.  Be sure to check some of the industrial grade prox sensors.  These will all be a little more expensive than as they are meant for drop in high reliability industrial controls applications.  You may want to use something simpler and have more of your own electronics behind it but here are some industrial solutions that get used in manufacturing lines.

http://www.turck-usa.com/Products/Sensors/
look under the tab for inductive sensors.  They have a lot of choices!

http://www.keyence.com/products/sensors/proximity/proximity.php

http://www.ia.omron.com/product/48.html
 « Last Edit: October 16, 2012, 03:50:54 03:50 by Gallymimu » Logged
solutions
Hero Member

Offline

Posts: 1446

Thank You
-Given: 587

 « Reply #13 on: October 14, 2012, 07:13:10 19:13 »

You cannot use magnets on an aircraft, or you'll be going north no matter which way you point it
 Logged
Gallymimu
Hero Member

Offline

Posts: 577

Thank You
-Given: 98

 « Reply #14 on: October 14, 2012, 07:26:43 19:26 »

You cannot use magnets on an aircraft, or you'll be going north no matter which way you point it

Not sure what you are talking about, are you referring to the use of prox sensors?  If so you should read up on modern prox sensors as there isn't a magnet involved in the sensors I use (or suggested).  If you were talking about something else my apologies

Anyway of course you can take a magnet on an aircraft.  Have you ever seen TSA signs that say NO MAGNETS.  HAHAHAHA

The earth's magnetic field is about 0.3-0.6 gauss near the surface of the planet (http://en.wikipedia.org/wiki/Gauss_(unit)).  Let's assume we have an N35 neodymium magnet of approx 0.25insq.  This would have a magnetic force of around 2000 gauss at 0.1" from the magnet surface) (http://www.kjmagnetics.com/calculator.asp).  Since the magnetic field falls as at least an inverse square of the distance (at least this much assuming best case monopole magnetic fields which it wouldn't be but just for the sake of argument), you would only have to be 18" away (on axis which is worst case) to have less than 1% influence from the magnet.
 « Last Edit: October 14, 2012, 07:47:46 19:47 by Gallymimu » Logged
cranor
Newbie

Offline

Posts: 17

Thank You
-Given: 8

 « Reply #15 on: October 15, 2012, 11:06:18 11:06 »

An automotive wheel speed hall effect sensor would maybe do the job ?
Easy to find and the signal is often already shaped for "direct reading".
 Logged
solutions
Hero Member

Offline

Posts: 1446

Thank You
-Given: 587

 « Reply #16 on: October 16, 2012, 02:51:50 02:51 »

Not sure what you are talking about, are you referring to the use of prox sensors?  If so you should read up on modern prox sensors as there isn't a magnet involved in the sensors I use (or suggested).

YOU WROTE "I would use the magnetic prox sensors."

Quote
Anyway of course you can take a magnet on an aircraft.  Have you ever seen TSA signs that say NO MAGNETS.  HAHAHAHA

The earth's magnetic field is about 0.3-0.6 gauss near the surface of the planet (http://en.wikipedia.org/wiki/Gauss_(unit)).  Let's assume we have an N35 neodymium magnet of approx 0.25insq.  This would have a magnetic force of around 2000 gauss at 0.1" from the magnet surface) (http://www.kjmagnetics.com/calculator.asp).  Since the magnetic field falls as at least an inverse square of the distance (at least this much assuming best case monopole magnetic fields which it wouldn't be but just for the sake of argument), you would only have to be 18" away (on axis which is worst case) to have less than 1% influence from the magnet.

Dear Dr. Math Theory:

You CANNOT.

TSA does not set the regulations for Aviation or potential interference with navigation.

"Air-freight forbidden (magnetic field too large)
If the package with the magnets (including magnetic shielding packaging) creates a magnetic field of more than 0,00525 gauss at a distance of 4,6 metres (in any direction), the transport of the package via air freight is generally forbidden.
Alternatively, a compass can be placed at this distance, and if the needle diverts by more than two degrees, it failed the test and the transport is forbidden.

* * *

The second and more strict test requires that at a distance of 2,1 metres from the package a magnetic field of less than 0,002 gauss is created. (Alternative: a compass at this distance diverts by less than 0,5 degrees).
* * *

Shipment as a declared dangerous good (for all other cases)
If the package falls between the two categories, meaning the first test was passed but not the stricter second one, the package has to be declared a dangerous good, which will lead to higher shipping costs and additional work. Also, there are certain airports that cannot be used in this case."

http://www.supermagnete.de/eng/faq/airfreight

Yeah - 4.6 meters, 5 MILLI-gauss (1 percent of your claimed Earth magnetism)

So, your 2x2x0.25 magnet will generate 3.4 milligauss at 2m (on a small plane, as in this app the distance from compass to spinner is more like about 4 feet, not to cargo hold as these regs intend to apply), exceeding the 2 milligauss compass swing threshold

Posted on: October 16, 2012, 02:15:24 02:15 - Automerged

An automotive wheel speed hall effect sensor would maybe do the job ?
Easy to find and the signal is often already shaped for "direct reading".

Magnetic...
 Logged
Gallymimu
Hero Member

Offline

Posts: 577

Thank You
-Given: 98

 « Reply #17 on: October 16, 2012, 03:50:13 03:50 »

HA, I did say magnetic prox sensor.  So sorry for the confusion.  I'll edit the post for clarity!

Thanks for all the great information Solutions.

I'm kinda honored that you can tell that I'm a doctor.  I must sound pretty smart . I'm really more hands on and pretty unmathy if you can believe it.  I think it's harder to be credible in a forum environment without quantitative references.  I certainly don't believe anything I see on forums.  There are a lot of people who haven't gotten a clue as to what they are talking about but go spouting off with great authority.

Try not to be condescending, we're all friends here (honestly it makes me feel like you don't like any of us sometimes ).  That said I'm glad were keeping each other honest .  I'm really not here to bust balls and am sorry if it comes across that way.  It's always good practice to go and look up things that don't seem right and make sure.  It makes us better educators.  I hope that everyone who follows this stuff learns from the debate.

So the prox sensors I was thinking of, as I mentioned can detect metal.  They are great when used with a screw head sticking out of a shaft.  They tend to work better and have better ranges (closer to that 10-20mm) when you use ferrous metals but they also work okay with non-ferrous.  They are actually pretty cool but I don't have a sense as to whether they have the bandwidth for the rotation speeds that will be seen in this application.  I simply never looked into it.
 « Last Edit: October 16, 2012, 04:06:31 04:06 by Gallymimu » Logged
solutions
Hero Member

Offline

Posts: 1446

Thank You
-Given: 587

 « Reply #18 on: October 16, 2012, 04:39:31 04:39 »

Nobody knows what prox sensors you were thinking of (we're not mind readers). Mfr, part number, datasheet?

Every one I know of that detects metal has a magnet in it
 Logged
Gallymimu
Hero Member

Offline

Posts: 577

Thank You
-Given: 98

 « Reply #19 on: October 16, 2012, 06:55:27 06:55 »

Nobody knows what prox sensors you were thinking of (we're not mind readers). Mfr, part number, datasheet?

Every one I know of that detects metal has a magnet in it

I provided links here (http://www.sonsivri.to/forum/index.php?topic=47029.msg139113#msg139113) but I can be more specific if it helps.

So these style don't require the use of a magnet for detection
http://www.turck-usa.com/Products/Sensors/Inductive/Barrel_Sensors/Barrel_with_Potted-in_Cable/Barrel_with_Potted-in_Cable_3-wire_TTL.htm

Here is an overall guide for the turck sensors, there is some theory of operation in there.
http://stevenengineering.com/tech_support/PDFs/46SENMSPECS.pdf

Similar guide for keyence sensors.
http://www.sensorcentral.com/photoelectric/proximity01.php

Previously I thought the discussion was use of a magnet as the sensed element and now I think you are suggesting that the magnet is inside the sensor?  Certainly these work by induction.  I wouldn't call that a magnet no more than I'd call a pot core on a small switching power supply a magnet (if my mind reading correct).  As most of these sensors draw only about 1/10th of a watt I don't think any appreciable magnetic field could be generated.  Don't worry I won't go and try to math everyone to death on that one!

So is your position that: 1). These sensors work via induction therefore there is a magnetic field induced.  2). Magnets may not be used on aircraft because they will interfere with the aircraft compass. 3).  Therefore these types of sensors can not be used in an aircraft?

OR am I completely misunderstanding?
 « Last Edit: October 16, 2012, 07:14:15 07:14 by Gallymimu » Logged
solutions
Hero Member

Offline

Posts: 1446

Thank You
-Given: 587

 « Reply #20 on: October 16, 2012, 12:22:24 12:22 »

Induction = magnetic field.

Reluctance = magnetic field

Even a professor, like yourself, knows that

Your inductive sensors must have an exciter coil if they are variable reluctance sensors, or they must sense an external mag field. There is no magic in physical laws.

And yes, magnets are avoided at all costs in aircraft components, and if they are absolutely required (magneto, for instance), they are shielded. The compass is the last safety net for navigation/control and you cannot legally take off without one.

Anyway, it was established a long time ago in this thread that the sensor was not the problem that needed to be solved, so all of this alternative sensor stuff has been a fun, somewhat educational (I hope...), but now getting boring, pi\$\$ing contest
 Logged
Gallymimu
Hero Member

Offline

Posts: 577

Thank You
-Given: 98

 « Reply #21 on: October 16, 2012, 04:02:35 16:02 »

For the OP.  Here are two companies that sell metal sensing induction based prox sensors specifically for avionics applications specified to detect ferrous metals.

Crane Aerospace: http://www.craneae.com/Products/Sensing/ProximitySensors.aspx

and Goodrich Sensors: http://www.goodrich.com/gr-ext-templating/images/Goodrich%20Content/Business%20Content/Sensors%20and%20Integrated%20Systems/Products/Literature%20Listing/4210%20-%20Proximity%20Sensing%20Systems%20-%20Cover%20A30(low).pdf

I wouldn't write off this sensor type as a solution quite yet.  It looks like Goodrich has these deployed on fairly modern small and larger aircraft (F-35, Boeing 787).

 Logged
Terry Dactil
V.I.P
Newbie

Offline

Posts: 10

Thank You
-Given: 18

 « Reply #22 on: October 17, 2012, 04:46:22 04:46 »

Quote
And yes, magnets are avoided at all costs in aircraft components, and if they are absolutely required (magneto, for instance), they are shielded.
That's completely wrong.

Aircraft are full of unshielded systems that produce magnetic fields and /or contain magnets. Also, everyone on this forum would know that a cable carrying a current produces a magnetic field and aircraft are full of electrical wiring.
The point is that they are installed components and their effect on the compass system is known and allowed for.

A ‘compass swing’ is a maintenance procedure where the aircraft is pointed in various directions and the local magnetic field near the compass system is ‘tweaked’ with small adjustable magnets to correct the compass readings as much as possible. Any remaining errors are listed on a ‘deviation card’ (e.g. for heading 090 steer 092). Deviation card adjustment up to 10 degrees for a standby compass is acceptable, but it is generally much less.

Here are some of the Australian regulations:

6. OCCASIONS FOR CALIBRATION
6.1   Each installed compass or compass system should be calibrated:
(a) prior to the issue of an initial Certificate of Airworthiness; or
(b) at intervals specified in the aircraft’s Maintenance Program.

6.2   Each compass or compass system should also be calibrated:
(a) when initially installed or reinstalled in an aircraft;
(b) when permanently locating an aircraft from northern hemisphere operations to southern hemisphere operations;
(c) after an engine change, unless the aircraft manufacturer prescribes otherwise;
(d) whenever a magnetic sensing element has been changed or relocated;
(e) after modification of an electrical  or avionic system installation of the aircraft, unless the certifying engineer is satisfied that the modification will not affect the compass;
(f) after a suspected lightning strike, unless an in-flight comparison on at least four headings, checked 90 degrees apart proves that no change of deviation has occurred between the standby compass and compass system(s) installed on the aircraft;
Note: A heading check may only be made during the flight on which a suspected strike has occurred if this procedure is documented in the appropriate aircraft flight manual.
(g) after any maintenance involving the addition, removal, or relocation of magnetic materials likely to influence compass deviation; Note: Manufacturers maintenance manuals may indicate the components that, if changed, would require the compass to be swung.
(h) following any operational occurrence, such as an accident, or heavy landing, that is likely to affect compass deviation; and
(i) after long-term storage of the aircraft whenever there is reason to suspect that a change of deviation may have occurred.

The general prohibition on magnetic materials being carried as cargo mentioned in previous posts is because the effect would be unknown.
It is permissible in certain situations to carry magnetic materials as cargo by loading it onto the aircraft and then doing a new compass swing so the cargo’s effect is now known and allowed for.

So basically, putting magnets on aircraft is perfectly OK if you know what you are doing and follow the clearly defined procedures to allow for the effects.
 Logged
Gallymimu
Hero Member

Offline

Posts: 577

Thank You
-Given: 98