For a project that I am working on, I need to detect the output of a few IR
leds at about 300 feet (abt. 100 meters) away. The output is pulsed, to
help distinguish it from background light.
I have succeeded in doing this out to about 100 feet, but am having
difficulty beyond that because of noise. I have found appnotes which
explain photodiode noise to some extent, but nothing on phototransistor
noise. Could someone point me to a good reference for both?
Also, does light emitted by an incandescant bulb have a significant noise
due to the discrete nature of the photons (like shot noise in electric
current)? How would one go about calculating the approximate amount of
noise from such a source? I have been using flashlights to simulate intense
background light (like the receiver is trying to pick out the source with
the sun as a backdrop) and the amount of noise I have been seeing is much
larger than I would expect due to electrical effects. I was wondering if I
am seeing real fluctuations in the light intensity.
Sean Breheny wrote:
>
> Hi all,
>
> For a project that I am working on, I need to detect the output of a few IR
> leds at about 300 feet (abt. 100 meters) away. The output is pulsed, to
> help distinguish it from background light.
What is your pulse rate and what is the receiver bandwidth? Of course,
the narrower the bandwidth the better for noise.
>
> I have succeeded in doing this out to about 100 feet, but am having
> difficulty beyond that because of noise. I have found appnotes which
> explain photodiode noise to some extent, but nothing on phototransistor
> noise. Could someone point me to a good reference for both?
I have no references at hand but I suspect that a phototransistor is
noisier than the combination of a photodiode and a low noise transistor.
This would be because of the difficulty of optimizing both photo-gain
and noise in one device in the case of the phototransistor.
You may want to try making your own photodiode/transistor device for
comparison purposes. I think you would want to choose a transistor that
has good noise characteristics at very low currents. A 2N5089 is a part
that comes to mind. I expect that there are better ones.
>
> Also, does light emitted by an incandescant bulb have a significant noise
> due to the discrete nature of the photons (like shot noise in electric
> current)? How would one go about calculating the approximate amount of
> noise from such a source? I have been using flashlights to simulate intense
> background light (like the receiver is trying to pick out the source with
> the sun as a backdrop) and the amount of noise I have been seeing is much
> larger than I would expect due to electrical effects. I was wondering if I
> am seeing real fluctuations in the light intensity.
Again for comparison purposes, maybe try an LED light source to simulate
the background light. Maybe a whole bunch of white LED's ???
I think that sunlight will have a much larger IR content then a flash light
would, so using a flash light bulb will likely not simulate intense background
light.
You could also try using optical filters to filter out light from your receiver
that is outside of the IR diode spectrum.
> Sean Breheny wrote:
> >
> > Also, does light emitted by an incandescant bulb have a significant noise
> > due to the discrete nature of the photons (like shot noise in electric
> > current)? How would one go about calculating the approximate amount of
> > noise from such a source? I have been using flashlights to simulate intense
> > background light (like the receiver is trying to pick out the source with
> > the sun as a backdrop) and the amount of noise I have been seeing is much
> > larger than I would expect due to electrical effects. I was wondering if I
> > am seeing real fluctuations in the light intensity.
Thanks for the response. Please see my answers below:
At 05:09 PM 6/4/00 -0600, Richard Ottosen wrote:
>What is your pulse rate and what is the receiver bandwidth? Of course,
>the narrower the bandwidth the better for noise.
The pulse rate can be whatever I want(I was trying 2kHz because that was
high enough to get away from the 1/f noise of op-amps and low enough so
that op-amps still have significant gain). The bandwidth need only be about
10-20Hz.
>I have no references at hand but I suspect that a phototransistor is
>noisier than the combination of a photodiode and a low noise transistor.
>This would be because of the difficulty of optimizing both photo-gain
>and noise in one device in the case of the phototransistor.
>
>You may want to try making your own photodiode/transistor device for
>comparison purposes. I think you would want to choose a transistor that
>has good noise characteristics at very low currents. A 2N5089 is a part
>that comes to mind. I expect that there are better ones.
That makes lots of sense. When I was able to get to 100 feet, I used a
photodiode combined with a transimpedance amp made from half of a LMC6482,
which has a noise of about 30nV/root Hz around 1kHz. I was actually using a
600Hz pulse freq, so it would probably be a bit easier at 2kHz. I wonder if
it is feasible to do better with a discrete transistor?
>Again for comparison purposes, maybe try an LED light source to simulate
>the background light. Maybe a whole bunch of white LED's ???
Well, yes, you may be correct that an LED would have less noise (no
filament vibrations, etc.), but my question wasn't very clear. What I meant
was: is there some fundamental noise which is inherent in ALL noncoherent
(or perhaps even coherent, for that matter) light sources, simply due to
the fact that the light is emitted as discrete photons which come out at
random times? I think there is, and that it is a similar effect (possibly
different probability distribution) to shot noise in an electric current
(noise which exists in an electric current because it consists of discrete
charges).
Thanks for the reply. Well, both have significant IR content, but you may
be correct that the sun would have more. I was simply trying to explain why
I get so much noise when I shine a flashlight into it.
The phototransistor I am currently using has a built-in IR filter.
Sean
At 11:03 PM 6/4/00 -0400, you wrote:
>I think that sunlight will have a much larger IR content then a flash light
>would, so using a flash light bulb will likely not simulate intense
background
>light.
>You could also try using optical filters to filter out light from your
receiver
>that is outside of the IR diode spectrum.
>
A) Power IR LEDs with narrow pulses and LARGE instantaneous
currents. This will make best use of the super-luminant
effect. To aid in getting these large currents delivered,
use fets or transistors that can handle the current pulses.
Drive them HARD, with clean, fast rise and fall times.
If you do not drive the FET/TRANSISTOR hard and fast enough,
then it will spend too much time in the linear region
and overheat.
Keep all wiring between LED and local power source SHORT
and eliminate inductance as much as you can. Use a LOCAL
current source such as a low ESR capacitor close by the
fet/LED combination. Use at least one local storage cap
for each series string of LEDs. Cap must be fully charged
during the LED "off" time. By "fully charged" I mean to about
90% of the nominal power supply voltage.
For best results use IR LEDs that come in metal casings,
and mount the IR LEDs in snug holes drilled through
a bar of aluminum such that the lens of the IR LED
just barely sticks out. Using a heat sink lets you
push the LEDs just a bit more.
B) Use multiple IR LEDs. One very useful technique is to
run several IR LEDs in series. This greatly reduces
average current drain and minimizes wasted current,
since you can now use a smaller resistor value for the
current limiting resistor. More LEDs not only increases
the average amount of light generated, it also widens
the field of view of the light source.
C) Use a condensing lens arrangement to collimate the light
from the LED(s) so that the light is aimed in the direction
of interest.
2) Detect signal and minimize interference
A) Use an IR filter at the receiving end. Do NOT use any
filter at the transmitting end. The narrower the bandwidth
of the filter, and the closer it is matched to the IR
bandwidth of the LED(s), the better.
B) If operating in low light environments, it is sometimes
useful to use a local IR LED to flood the IR sensors with
a very weak DC IR beam that just barely gets the sensor
to operate in its active region.
C) Use a modulated IR beam. You get this for free if you are
pulsing the LEDs. Keep the pulse width ratio LOW. Something
around 5% works well in most cases.
D) An optical lens can help gather light from a larger
area and concentrate it. Large lenses can also help
increase the useable viewing angle.
E) Once the signal has been detected, AC couple the modulated
frequency to a High Q Narrow BandPass Filter. This
helps you differentiate between noise and signal,
and rejects anything outside the band of interest.
I have found LC networks to have the best characteristics
for the bandpass elements. A parallel LC arrangement
in the feedback section of an inverting opamp circuit
works very well.
DC fed bulbs have been used in cinematography and other applications for
many years. These applications are noise-sensitive and there never were
bulb-related problems. The detector and the amplifier were the culprits in
evey case. The best approximation of an incandescent bulb run within
parameters is a black body radiator of the same temperature ;-) Assume
light is corpuscular and calculate the number of photons that reach your
detector from the solid angle of view. From this work out the shot noise.
I suspect that all other things being equal a LED may be quieter than a
blackbody (on account of bandwidth) and a laser even quieter (for the same
reason).
Note that over-volted small bulbs and certain halogen bulbs are not quiet.
This is because the filament is physically boiling (evaporating metal into
the gas filling) and because of strong convection currents in the gas
filling.
Once upon a time IR remote receivers were discrete-built and one way was
to feed a parallel LC tank directly from a PIN photodiode. I once made a
test circuit using a ZN414 and a 60 kHz tape bias tank pulled down to 40
kHz. It worked but the ZN414 is hard to come by. You could try a
monolithic AM radio IC using its IF section only. You can achieve gains of
60 dB easily like this and noise will be low even with a single tank.
I think that the extra noise you saw when shining the flashlight into the
detector is due to the changed input impedance of the circuit when the
photodetector is overloaded. This may change the way your filter works and
inject or amplify noise.
hope this helps,
Peter
PS: If anyone knows where ZN414s can be had, or their equivalents, please
say so. I have a few ideas that would greatly benefit from this chip. Why
did they stop making it ?!
PS2: My postings are becoming longish. If anyone is bothered by this then
please say so.
Peter:
Gateway Electronics (in the U.S.) has some surplus parts at their Denver
store. They have maybe 50 to 100 of them. Their main store in St. Louis
sometimes has more stock.
>
> Hi Sean,
>
> DC fed bulbs have been used in cinematography and other applications for
> many years. These applications are noise-sensitive and there never were
> bulb-related problems. The detector and the amplifier were the culprits in
> evey case. The best approximation of an incandescent bulb run within
> parameters is a black body radiator of the same temperature ;-) Assume
> light is corpuscular and calculate the number of photons that reach your
> detector from the solid angle of view. From this work out the shot noise.
>
> I suspect that all other things being equal a LED may be quieter than a
> blackbody (on account of bandwidth) and a laser even quieter (for the same
> reason).
>
> Note that over-volted small bulbs and certain halogen bulbs are not quiet.
> This is because the filament is physically boiling (evaporating metal into
> the gas filling) and because of strong convection currents in the gas
> filling.
>
> Once upon a time IR remote receivers were discrete-built and one way was
> to feed a parallel LC tank directly from a PIN photodiode. I once made a
> test circuit using a ZN414 and a 60 kHz tape bias tank pulled down to 40
> kHz. It worked but the ZN414 is hard to come by. You could try a
> monolithic AM radio IC using its IF section only. You can achieve gains of
> 60 dB easily like this and noise will be low even with a single tank.
>
> I think that the extra noise you saw when shining the flashlight into the
> detector is due to the changed input impedance of the circuit when the
> photodetector is overloaded. This may change the way your filter works and
> inject or amplify noise.
>
> hope this helps,
>
> Peter
>
> PS: If anyone knows where ZN414s can be had, or their equivalents, please
> say so. I have a few ideas that would greatly benefit from this chip. Why
> did they stop making it ?!
>
> PS2: My postings are becoming longish. If anyone is bothered by this then
> please say so.
Thanks to all who responded to this question. As usual, I didn't make
myself clear enough, so many of the proposed solutions are not applicable
in this case. Let me explain.
I am a model rocket enthusiast (as are many fellow piclisters, as I have
seen in previous threads), and for a long time I have wanted to try a
project where some kind of control was added to a model rocket. The most
obvious thing (of any significant challenge level) is to try to get the
rocket to hit or pass near to a target suspended in the air (by several
helium balloons).
So, what I have decided to do is to make the target consist of an array of
IR leds, pulsed any way we want (however is best). A four-channel receiver
goes into the rocket's nose and by placing light baffels in the right
positions, the onboard PIC (yes, it involves a PIC ;-) can get a rough idea
of its angle relative to the target by seeing which of the four receivers
is receiving light from the target. Two micro servos would then operate
control fins in an attempt to guide it to the target.
Because of the large number of variables, I can't guarantee the orientation
of the rocket with respect to the target, so I can't collimate the beam
from the target or put all the LEDs pointing in the same direction, etc. In
addition, because of size constraints in the rocket(four receivers have to
fit in a 1.8 inch diameter tube), I don't think it is feasible to use
lenses on the receiver photodiodes/phototransistors.
As I said before, I was able to get the receiver to detect the target at
about 100 feet (perhaps a bit more), so I don't need a really huge
improvement (about a factor of 8 or so in signal to noise ratio should
probably do it).
So, as you can see, this is not a communications system, and I can't put
the receiver and transmitter in tubes with lenses ;-)
I haven't exhausted all my own ideas on this yet, but I thought I would ask
to save myself the trouble if someone had a really good idea. Please do
give me any suggestions you have.
By the way, I have been working on this for quite some time, but I
hesitated to post it for fear that people would think I was trying to work
on a real guided missile. Rest assured, it carries no explosive, has a
ceiling of about 300 feet, and it's guidance system is not practical for
military applications.
Sean
At 12:30 AM 6/6/00 -0700, Arthur Brown wrote:
>Try putting tx & rx in a tube so as to shade from sunlight make sure you use
>lens with the right focal lenth.
>Firm I worked for used this over 250ft.
>
>Regards Art
>
Sean,
Now that we know more details we can be more helpful
in our responses. :)
Since you are using four phototransistors and they will be
very close to one another, you want to maximize the
signal differential. This can be attained by making sure
that each of them sees a different view. You would want the
detectors to be aiming ever so slightly AWAY from the
center, -OR- make sure that there is some optical separation
based on the presence of an optical barrier between
each of the sensors.
This helps enhance the difference signal significantly.
It does not in any way *boost* the received signal,
but it does help make the four received signals more
differential, so to speak. The barrier must be IR opaque
and just high enough to cast a shadow when illuminated
off-axis. You can use the rocket itself as the barrier.
If possible, mount the four sensors in the nose cone,
separated as far apart as you can. In this case the
nose cone itself becomes the optical barrier.
I'm really interested in the servo mechanisms you use on this. What servo is
small enough yet fast enough? How much deflection on the fins? What size
engines?
I thought it would be helpful if I explained it in more detail, considering
the responses I got (they were good, but this isn't exactly a
communications system ;-)
Yes, what you describe is exactly what I plan on doing. I was planning on
using cardboard, painted black, as the barrier. You did remind me, though,
to test this for IR opacity.
I derived a formula for the range of angles for which each sensor will be
illuminated for given barrier dimensions and used it to tweak the spacing
for the right null zone width (some null zone is probably necessary to
prevent the thing from going nuts when it is close to being pointed
correctly).
>Sean,
>Now that we know more details we can be more helpful
>in our responses. :)
>
>Since you are using four phototransistors and they will be
>very close to one another, you want to maximize the
>signal differential. This can be attained by making sure
>that each of them sees a different view. You would want the
>detectors to be aiming ever so slightly AWAY from the
>center, -OR- make sure that there is some optical separation
>based on the presence of an optical barrier between
>each of the sensors.
>
>This helps enhance the difference signal significantly.
>It does not in any way *boost* the received signal,
>but it does help make the four received signals more
>differential, so to speak. The barrier must be IR opaque
>and just high enough to cast a shadow when illuminated
>off-axis. You can use the rocket itself as the barrier.
>If possible, mount the four sensors in the nose cone,
>separated as far apart as you can. In this case the
>nose cone itself becomes the optical barrier.
>
>Fr. Tom McGahee
>
|
Well, you are much more the expert on servos than I am, and the fact that
you ask that question worries me :-(. I haven't fully thought out every
aspect of the project yet. I bought and tested (in a PIC circuit) a HiTech
HS-50, which claims to be able to do a 60 deg rotation in .09 sec, and have
a torque of 0.6kg/cm ( I assume they mean 0.6 kg * cm ?!). It is only 5.8
grams and its largest dimension is 0.87 inch. If it seems promising, I am
planning on buying a second one (the rocket needs two).
What do you think? BTW, do you know about how much current one of these
needs? They don't specify.
>Sean,
>
>I'm really interested in the servo mechanisms you use on this. What servo is
>small enough yet fast enough? How much deflection on the fins? What size
>engines?
>
>This sounds like a fun project!
>
>Andy
>
|
>Yes, what you describe is exactly what I plan on doing. I was planning on
>using cardboard, painted black, as the barrier. You did remind me, though,
>to test this for IR opacity.
You can get some real suprises. Also note that unless it's a metal cased
transistor, it's also sensitive to light coming into the base, and every
other angle. It's best at the front, but the other parts work too.
>I derived a formula for the range of angles for which each sensor will be
>illuminated for given barrier dimensions and used it to tweak the spacing
>for the right null zone width (some null zone is probably necessary to
>prevent the thing from going nuts when it is close to being pointed
>correctly).
Be prepared for some fun, what you're doing is very much like the
sidewinder missile control system. They used a segmented spun mirror to
chop the target IR, different rates for each band of mirrors. Apparently
there was some degree of vane movement dependent on the frequency. Pilots
later tapped into this signal to give them an indication that the
sidewinder could see the target before launching, which improved the
results significantly.
At 08:03 PM 6/7/00 -0700, David VanHorn wrote:
>You can get some real suprises. Also note that unless it's a metal cased
>transistor, it's also sensitive to light coming into the base, and every
>other angle. It's best at the front, but the other parts work too.
Yes, I realize that. I am not planning on firing much off boresight (at
least at first), but I will keep it in mind.
>Be prepared for some fun, what you're doing is very much like the
>sidewinder missile control system. They used a segmented spun mirror to
>chop the target IR, different rates for each band of mirrors. Apparently
>there was some degree of vane movement dependent on the frequency. Pilots
>later tapped into this signal to give them an indication that the
>sidewinder could see the target before launching, which improved the
>results significantly.
HAHAHAHAHAHA, I was wondering how the real missiles did it (I figured it
couldn't be TOO sophistocated considering the AIM-9 was introduced in the
early 50's). I got the idea of light barriers from a web page of an EE
school project, in which the students built a rocket to seek the sun. They
used LDRs and light barriers. It would be REALLY neat to use a lens and a
PSD, but I think that would be expensive, and perhaps not sensitive enough.
I have also dropped the idea of a gimballed seeker or lead pursuit, both
because of complexity, and the fact that it is tracking a stationary target
:-)
I had always wondered what the origin of the "growl" was. It is kinda neat
that the missile says "lemme at 'em!" by an audible indication.
>
>You know how the sidewinder got it's name, right?
I have fired many a sidewinder in simulation, but I can only guess that it
got its name from the similarity of its flight path to the sidewinding
motion of certain snakes. Is that right, or is there more to it?
It is truly amazing what you end up discussing on the PICLIST <VBG>
For those just joining the thread, I think I should add the standard
disclaimer: no, the PICLIST has not just turned into a terrorist
think-tank! This is just a project to build a small model rocket which can
guide itself to a balloon-suspended target. No explosives involved.
>Yes, I realize that. I am not planning on firing much off boresight (at
>least at first), but I will keep it in mind.
You might want to paint the backs and sides so that they don't get any
other "input".
I don't have a good picture of the arraingement other than it's four quadrant.
>HAHAHAHAHAHA, I was wondering how the real missiles did it (I figured it
>couldn't be TOO sophistocated considering the AIM-9 was introduced in the
>early 50's). I got the idea of light barriers from a web page of an EE
>school project, in which the students built a rocket to seek the sun. They
>used LDRs and light barriers. It would be REALLY neat to use a lens and a
>PSD, but I think that would be expensive, and perhaps not sensitive enough.
>I have also dropped the idea of a gimballed seeker or lead pursuit, both
>because of complexity, and the fact that it is tracking a stationary target
>:-)
The sidewinder chops the incoming IR with the mirror, I understand you are
modulating the target. Much the same, except that the mirror works with
uncooperative targets, and is more difficult to jam.
>I have fired many a sidewinder in simulation, but I can only guess that it
>got its name from the similarity of its flight path to the sidewinding
>motion of certain snakes. Is that right, or is there more to it?
The seeker is always in error to some degree, so the flight path tends to
wobble.
As the missile rolls, the inputs automatically translate though (cute)
Another fellow at NARAM two years ago built a finless rocket that sensed
the horizon and gimballed the engine. I saw it, it flew well.
Watch the gain though, excessive will be "interesting". :)
>For those just joining the thread, I think I should add the standard
>disclaimer: no, the PICLIST has not just turned into a terrorist
>think-tank! This is just a project to build a small model rocket which can
>guide itself to a balloon-suspended target. No explosives involved.
<old man mode> NAR and tripoli forbid this of course, and the technology
could be extended to laser "painted" targets </old man mode>
That sounds small and fast, but low torque. Usually the small ones are .15-.20
secs/90 deg.
Servos can draw up to 1A if heavily loaded! You should make sure that the
servos and PIC don't share the power supply, or if they do, that the PIC power
is well isolated (LC) and you bypass it well.
I also suggest you drive the pulses out it at 100Hz or more (you need about a
6mS delay between pulses, which range from 1-2 mS, so you could theoretically be
running on an 8 mS period=120Hz). This will increase the apparent torque (since
the servo relaxes and doesn't hold position well between refreshes) and give you
a better response to control inputs.
Also I suggest you do some aerodynamic studies on the movable fin design. At
high speed, flat control surfaces flutter. You should have airfoil surfaces on
the fins. Also you will want to do some dynamic balancing of them, which can
also help prevent flutter.
Subject: Re: [EE]: Noise in
phototransistors/photodiodes/light
Hi Andy,
Well, you are much more the expert on servos than I am, and the fact that
you ask that question worries me :-(. I haven't fully thought out every
aspect of the project yet. I bought and tested (in a PIC circuit) a HiTech
HS-50, which claims to be able to do a 60 deg rotation in .09 sec, and have
a torque of 0.6kg/cm ( I assume they mean 0.6 kg * cm ?!). It is only 5.8
grams and its largest dimension is 0.87 inch. If it seems promising, I am
planning on buying a second one (the rocket needs two).
What do you think? BTW, do you know about how much current one of these
needs? They don't specify.
>Sean,
>
>I'm really interested in the servo mechanisms you use on this. What servo is
>small enough yet fast enough? How much deflection on the fins? What size
>engines?
>
>This sounds like a fun project!
>
>Andy
>
|
At the high velocity that a model rocket flies at you do
not need much fin movement to cause steering.
What you DO need is FAST RESPONSE! In fact, you can keep
the fins fixed in place and do the actual "steering" using
changes in control surfaces along the sides. These
control surfaces need only move a small amount, and
it is fine to just control them with a simple ON/OFF
control. Consider it to be a form of pulse width modulation
control. A simple electromagnetic solenoid would probably
be just fine.
The control surfaces actually cause the airflow to change
and that causes the rocket to change direction. These
control surfaces "spoil" the airflow. Four of them would
be needed. A good spot for the control surfaces is at the
front of the rocket.
They can take various forms. One form is a mini airfoil
that sticks out the side of the rocket. Looks like a very
small fin.
The other form is a small hinged plate whose bottom can be
pushed outward a small amount. When it is swung out it changes
the aerodynamics on that side, thus causing a change in direction.
For various reasons, it is best if the "hinged plate" style
has is more three dimensional in nature, so that as it
swings in and out the airflow is kept directed only along
the surface. A simple flat plate hing arrangement would
not do as well, since airflow would create more of a turbulance
effect, which does not give good control.
Opposite facing control elements would normally work in
unison (though opposite), so the number of servos or
solenoids would be 2.
If using a solenoid, you may have to use a spring to
insure the control surface closes properly.
I would think that a servo might not respond fast enough.
What I would do is bypass the existing servo electronics and
control the motor directly. A simple opamp circuit
could use the pot feedback to achieve two known states.
Commands to this modified servo would be simply
Normal & Change Direction. If you keep the angle of travel
small you should be able to get reasonable response times.
> Yes, what you describe is exactly what I plan on doing. I was planning
> on using cardboard, painted black, as the barrier.
I'd suggest the opposite, at least for the "guides". Use silvered
foil to maximise gain. The four photocells mount at the very inside
corners, with a black barrier behind:
As for lensing; I suspect that a custom nosecone could be moulded to
function as a lens.
An alternate sensor mounting of course could be on the four sides of
the fuselage behind the nosecone, using the fuselage itself as the
baffle.
--
Cheers,
Paul B.
Actually, I guess I didn't read the previous suggestion well enough,because
what you suggest (except for the foil) is what I was planning on doing, the
sensors in the corners (actually, slightly displaced radially from the
corners because we want some null zone (a couple of degrees each way)).
I really like the foil idea, thanks! I'm afraid I won't be able to do the
custom molded nosecone (unless you know of a cheap way of doing custom
plastic molding!)
Sean
At 08:50 AM 6/9/00 +1000, Paul B. Webster VK2BZC wrote: {Quote hidden}
>Sean Breheny wrote:
>
>> Yes, what you describe is exactly what I plan on doing. I was planning
>> on using cardboard, painted black, as the barrier.
>
> I'd suggest the opposite, at least for the "guides". Use silvered
>foil to maximise gain. The four photocells mount at the very inside
>corners, with a black barrier behind:
>
>
> XXXX|XXXX
> XXXXXX|XXXXXX
> XXXXXX,|,XXXXXX <-- sensors
> --------+-------- in
> XXXXXX`|`XXXXXX <-- corners
> XXXXXX|XXXXXX
> XXXX|XXXX
> ^--------- reflective baffles
>
> As for lensing; I suspect that a custom nosecone could be moulded to
>function as a lens.
>
> An alternate sensor mounting of course could be on the four sides of
>the fuselage behind the nosecone, using the fuselage itself as the
>baffle.
>--
> Cheers,
> Paul B.
>
|
Subject: Re: [EE]: Noise in
phototransistors/photodiodes/light
Hi Paul,
Actually, I guess I didn't read the previous suggestion well enough,because
what you suggest (except for the foil) is what I was planning on doing, the
sensors in the corners (actually, slightly displaced radially from the
corners because we want some null zone (a couple of degrees each way)).
I really like the foil idea, thanks! I'm afraid I won't be able to do the
custom molded nosecone (unless you know of a cheap way of doing custom
plastic molding!)
Sean
At 08:50 AM 6/9/00 +1000, Paul B. Webster VK2BZC wrote: {Quote hidden}
>Sean Breheny wrote:
>
>> Yes, what you describe is exactly what I plan on doing. I was planning
>> on using cardboard, painted black, as the barrier.
>
> I'd suggest the opposite, at least for the "guides". Use silvered
>foil to maximise gain. The four photocells mount at the very inside
>corners, with a black barrier behind:
>
>
> XXXX|XXXX
> XXXXXX|XXXXXX
> XXXXXX,|,XXXXXX <-- sensors
> --------+-------- in
> XXXXXX`|`XXXXXX <-- corners
> XXXXXX|XXXXXX
> XXXX|XXXX
> ^--------- reflective baffles
>
> As for lensing; I suspect that a custom nosecone could be moulded to
>function as a lens.
>
> An alternate sensor mounting of course could be on the four sides of
>the fuselage behind the nosecone, using the fuselage itself as the
>baffle.
>--
> Cheers,
> Paul B.
>
|
Thanks to all who responded to this thread, both on and off list. I have
been very busy these last few days, so I haven't had a chance to respond in
detail. What I can say is that I have decided to switch the method of
target detection to four phototransistors or photodiodes placed behind a
convex lens, a little closer than focal length away, and mold my own
nosecone, using either epoxy as Andy suggests, or acrylic.
I have also almost completed a MATLAB simulation of the system and I will
try various control scenarios to determine if the servos are fast enough
and if the detection system is adequate. Bear in mind that this is a heavy
rocket with lots of drag and will reach less than 150 mph.
Lastly, I must ask Dave why the NAR and Tripoli ban this? What is it that
they specifically ban? The only clause I could find in the NAR safety code
was something prohibiting firing a rocket so that it hits a target. Is this
what you are referring to?
I am actually hoping that I DON'T hit the target(instead, I want a near
miss, say about 5 feet), because I want to preserve the rocket and target
(lots of work will have gone into them!). To accomplish this, I planned on
reducing the maximum control authority and widening the null zone enough to
reduce the probability of actually hitting, but still demonstrating control.
True, the NAR and Tripoli rules are not legally binding, but it would be
nice to be able to say that the project abides by them :-)
Thanks again, and I will be back to you on this later,
Sean
At 07:47 AM 6/9/00 -0400, Andrew Kunz wrote:
>Sean,
>
>You can do pretty good with West Systems epoxy and "clear" hardener. I
have a {Quote hidden}
>clear boat that I use to show the insides to people.
>
>Andy
>
>
>
>
>
>
>
>
>
>Sean Breheny <EraseMEshb7spamspamBeGoneCORNELL.EDU> on 06/09/2000 12:15:05 AM
>
>Please respond to pic microcontroller discussion list
>
>
>
>
>
>
>
>
> To: PICLISTSTOPspamspam_OUTMITVMA.MIT.EDU
>
> cc: (bcc: Andrew Kunz/TDI_NOTES)
>
>
>
> Subject: Re: [EE]: Noise in
> phototransistors/photodiodes/light
>
>
>
>
>
>
>
>
>Hi Paul,
>
>Actually, I guess I didn't read the previous suggestion well enough,because
>what you suggest (except for the foil) is what I was planning on doing, the
>sensors in the corners (actually, slightly displaced radially from the
>corners because we want some null zone (a couple of degrees each way)).
>
>I really like the foil idea, thanks! I'm afraid I won't be able to do the
>custom molded nosecone (unless you know of a cheap way of doing custom
>plastic molding!)
>
>Sean
>
>At 08:50 AM 6/9/00 +1000, Paul B. Webster VK2BZC wrote:
>>Sean Breheny wrote:
>>
>>> Yes, what you describe is exactly what I plan on doing. I was planning
>>> on using cardboard, painted black, as the barrier.
>>
>> I'd suggest the opposite, at least for the "guides". Use silvered
>>foil to maximise gain. The four photocells mount at the very inside
>>corners, with a black barrier behind:
>>
>>
>> XXXX|XXXX
>> XXXXXX|XXXXXX
>> XXXXXX,|,XXXXXX <-- sensors
>> --------+-------- in
>> XXXXXX`|`XXXXXX <-- corners
>> XXXXXX|XXXXXX
>> XXXX|XXXX
>> ^--------- reflective baffles
>>
>> As for lensing; I suspect that a custom nosecone could be moulded to
>>function as a lens.
>>
>> An alternate sensor mounting of course could be on the four sides of
>>the fuselage behind the nosecone, using the fuselage itself as the
>>baffle.
>>--
>> Cheers,
>> Paul B.
>>
>|
>| Sean Breheny
>| Amateur Radio Callsign: KA3YXM
>| Electrical Engineering Student
>\--------------=----------------
>Save lives, please look at http://www.all.org
>Personal page: http://www.people.cornell.edu/pages/shb7
>spamBeGoneshb7STOPspamEraseMEcornell.edu ICQ #: 3329174
>_____________________________________________
>NetZero - Defenders of the Free World
>Click here for FREE Internet Access and Email
>http://www.netzero.net/download/index.html
>
|
>Lastly, I must ask Dave why the NAR and Tripoli ban this? What is it that
>they specifically ban? The only clause I could find in the NAR safety code
>was something prohibiting firing a rocket so that it hits a target. Is this
>what you are referring to?
Active guidance twoard a target.
I don't have the regs in front of me, but I remembered that both forbid it.
I could be wrong.
- --
Are you an ISP? Tired of spam? http://www.spamwhack.com A pre-emptive strike against spam!
With how your government is so up itself with potential terrorist threats, I
can see you going to jail over this, after all you won't let Sony export
play stations to Iraq and other places.
Did you also see that thread on the other company selling FM bug kits.
I think that the people that speak into their wrists will jump on you as
soon as they hear about it (Perhaps they have!) Yet another terrorist weapon
disposed of.
Oh yes Sean, they will also wreak your carrier and any potential job
opportunity, take away your SSN, and ensure that you have a bad life!
I was under the impression that they banned blasting stuff away, not
almost-targeting balloons. I think this automatically includes "ballistic"
operations (mortar rounds), explosives, and aircraft. Shoot, don't we all do
the "drop the rocket in the circle" routine using a parachute? That's targeting
right there!
Subject: Re: [EE]: Noise in
phototransistors/photodiodes/light
Thanks to all who responded to this thread, both on and off list. I have
been very busy these last few days, so I haven't had a chance to respond in
detail. What I can say is that I have decided to switch the method of
target detection to four phototransistors or photodiodes placed behind a
convex lens, a little closer than focal length away, and mold my own
nosecone, using either epoxy as Andy suggests, or acrylic.
I have also almost completed a MATLAB simulation of the system and I will
try various control scenarios to determine if the servos are fast enough
and if the detection system is adequate. Bear in mind that this is a heavy
rocket with lots of drag and will reach less than 150 mph.
Lastly, I must ask Dave why the NAR and Tripoli ban this? What is it that
they specifically ban? The only clause I could find in the NAR safety code
was something prohibiting firing a rocket so that it hits a target. Is this
what you are referring to?
I am actually hoping that I DON'T hit the target(instead, I want a near
miss, say about 5 feet), because I want to preserve the rocket and target
(lots of work will have gone into them!). To accomplish this, I planned on
reducing the maximum control authority and widening the null zone enough to
reduce the probability of actually hitting, but still demonstrating control.
True, the NAR and Tripoli rules are not legally binding, but it would be
nice to be able to say that the project abides by them :-)
Thanks again, and I will be back to you on this later,
Sean
At 07:47 AM 6/9/00 -0400, Andrew Kunz wrote:
>Sean,
>
>You can do pretty good with West Systems epoxy and "clear" hardener. I
have a {Quote hidden}
>clear boat that I use to show the insides to people.
>
>Andy
>
>
>
>
>
>
>
>
>
>Sean Breheny <.....shb7spam_OUTCORNELL.EDU> on 06/09/2000 12:15:05 AM
>
>Please respond to pic microcontroller discussion list
>
>
>
>
>
>
>
>
> To: TakeThisOuTPICLISTKILLspamspamMITVMA.MIT.EDU
>
> cc: (bcc: Andrew Kunz/TDI_NOTES)
>
>
>
> Subject: Re: [EE]: Noise in
> phototransistors/photodiodes/light
>
>
>
>
>
>
>
>
>Hi Paul,
>
>Actually, I guess I didn't read the previous suggestion well enough,because
>what you suggest (except for the foil) is what I was planning on doing, the
>sensors in the corners (actually, slightly displaced radially from the
>corners because we want some null zone (a couple of degrees each way)).
>
>I really like the foil idea, thanks! I'm afraid I won't be able to do the
>custom molded nosecone (unless you know of a cheap way of doing custom
>plastic molding!)
>
>Sean
>
>At 08:50 AM 6/9/00 +1000, Paul B. Webster VK2BZC wrote:
>>Sean Breheny wrote:
>>
>>> Yes, what you describe is exactly what I plan on doing. I was planning
>>> on using cardboard, painted black, as the barrier.
>>
>> I'd suggest the opposite, at least for the "guides". Use silvered
>>foil to maximise gain. The four photocells mount at the very inside
>>corners, with a black barrier behind:
>>
>>
>> XXXX|XXXX
>> XXXXXX|XXXXXX
>> XXXXXX,|,XXXXXX <-- sensors
>> --------+-------- in
>> XXXXXX`|`XXXXXX <-- corners
>> XXXXXX|XXXXXX
>> XXXX|XXXX
>> ^--------- reflective baffles
>>
>> As for lensing; I suspect that a custom nosecone could be moulded to
>>function as a lens.
>>
>> An alternate sensor mounting of course could be on the four sides of
>>the fuselage behind the nosecone, using the fuselage itself as the
>>baffle.
>>--
>> Cheers,
>> Paul B.
>>
>|
>| Sean Breheny
>| Amateur Radio Callsign: KA3YXM
>| Electrical Engineering Student
>\--------------=----------------
>Save lives, please look at http://www.all.org
>Personal page: http://www.people.cornell.edu/pages/shb7
>.....shb7RemoveMEcornell.edu ICQ #: 3329174
>_____________________________________________
>NetZero - Defenders of the Free World
>Click here for FREE Internet Access and Email
>http://www.netzero.net/download/index.html
>
|
-----Original Message-----
From: Plunkett, Dennis <spamBeGonedplunkett@spam@spam_OUTAIRINTER.COM.AU>
Date: Wednesday, June 14, 2000 12:04 AM
Subject: Re: [EE]: Noise in phototransistors/photodiodes/light
>With how your government is so up itself with potential terrorist threats,
I
>can see you going to jail over this, after all you won't let Sony export
>play stations to Iraq and other places.
>> Subject: Re: [EE]: Noise in phototransistors/photodiodes/light
>>
>> Thanks to all who responded to this thread, both on and off list. I have
>> been very busy these last few days, so I haven't had a chance to respond
>> in
>> detail. What I can say is that I have decided to switch the method of
>> target detection to four phototransistors or photodiodes placed behind a
>> convex lens, a little closer than focal length away, and mold my own
>> nosecone, using either epoxy as Andy suggests, or acrylic.
I would have suggested using a diode switched doppler antenna and a
super-regenerative detector, like in my FUZZBUSTER-HARM design. But let us
know how it goes. The local cops have started using lidar now, so my old
design needs to be updated 8^)
> >Lastly, I must ask Dave why the NAR and Tripoli ban this? What is it that
> >they specifically ban? The only clause I could find in the NAR safety code
> >was something prohibiting firing a rocket so that it hits a target. Is this
> >what you are referring to?
>
> Active guidance twoard a target.
> I don't have the regs in front of me, but I remembered that both forbid it.
> I could be wrong.
> - --
Really get em going...
Make it shoulder launched, complete with lock on tone....
_____________________________
Lance Allen
Technical Officer
Uni of Auckland
Psych Dept
New Zealand
hehehe, Actually, I was planning on having a lock-on tone <VBEG>!
BUT, I'm not stupid enough to try to make it shoulder-launched! I don't
want a face-full of rocket exhaust! I think the real shoulder launched SAMs
use compressed air or something similar to boost the missile out some
distance from you before rocket ignition. In addition, due to the advice of
the people on this list (including those who emailed me off list), I'm
going to try to arrange my launcher so that I can be several hundred feet
away, in case it just goes nuts after leaving the rail.
Sean
At 11:51 AM 6/16/00 +1200, Lance Allen wrote:
>> >Lastly, I must ask Dave why the NAR and Tripoli ban this? What is it that
>> >they specifically ban? The only clause I could find in the NAR safety code
>> >was something prohibiting firing a rocket so that it hits a target. Is
this {Quote hidden}
>> >what you are referring to?
>>
>> Active guidance twoard a target.
>> I don't have the regs in front of me, but I remembered that both forbid it.
>> I could be wrong.
>> - --
>Really get em going...
>
>Make it shoulder launched, complete with lock on tone....
>_____________________________
>
>Lance Allen
>Technical Officer
>Uni of Auckland
>Psych Dept
>New Zealand
>
>http://www.psych.auckland.ac.nz
>
>_____________________________
>
|
I have not seen anyone that has had success yet with even a sun tracking
amateur rocket. Seems there are lots of problems in the control mechanisms.
But it does sound like a fun project.
>Lastly, I must ask Dave why the NAR and Tripoli ban this? What is it that
>they specifically ban? The only clause I could find in the NAR safety code
>was something prohibiting firing a rocket so that it hits a target. Is this
>what you are referring to?
Active guidance twoard a target.
I don't have the regs in front of me, but I remembered that both forbid it.
I could be wrong.
- --
Are you an ISP? Tired of spam? http://www.spamwhack.com A pre-emptive strike against spam!
> hehehe, Actually, I was planning on having a lock-on tone <VBEG>!
>
> BUT, I'm not stupid enough to try to make it shoulder-launched! I don't
> want a face-full of rocket exhaust! I think the real shoulder launched SAMs
> use compressed air or something similar to boost the missile out some
> distance from you before rocket ignition. In addition, due to the advice of
> the people on this list (including those who emailed me off list), I'm
> going to try to arrange my launcher so that I can be several hundred feet
> away, in case it just goes nuts after leaving the rail.
You might also want to try a couple of static test "firings" (probably
without ignition, at least at first) with the rocket clamped in a test
stand to observe the behavior of the fins. Anything over a very small
amount of travel would probably be a Bad Thing.
Dale
---
The most exciting phrase to hear in science, the one that heralds new
discoveries, is not "Eureka!" (I found it!) but "That's funny ..."
-- Isaac Asimov
cornell.edu ICQ #: 3329174
