Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss
Cheap Azimuth Indicator ??
9 views
Skip to first unread message
mpm
Feb 12, 2012, 3:17:19 PM2/12/12
to
We have a mobile truck equipped with a pneumatic push-up pole.
On top of this pole, we have a camera and yagi antenna mounted to a
simple pan-tilt-zoom controller.
We would like to know which way the contraption is pointed, regardless
of the truck's orientation.
What is the easiest way to do this? (Surely this requirement has been
beat to death, right??)
The PTZ controller is very simple: left, right, up, down. No azimuth
info. No known starting posiition regardless of azimuth (though we
might be able to cobble something up?).
While the camera could possibly "look at" something on the mount (like
a compass?) in addition to it's normal intended field of view, our
preference would be something other than "visual-only", maybe a 3-
digit RS-232 feed, or something we can at least have the option to
capture via computer. But hey, we'd take visual-only if that's the
only viable way to do it without upgrading to a more capable PTZ
controller. I don't have the camera lens specs handy, but it is an
auto-focus, auto-iris design.
Weight is a factor (it's a 30-foot pole) w/ 8-pounds dead weight
already.
Simplicity is a factor, as is low-cost.
Another possibility is to have the contraption superimpose the azimuth
data on the NTSC 1-volt video feed coming off the camera itself
(somehow). That would be ideal, actually.
Worst case, we'll stick a colored ball on a stick, and use that as a
reference, but then we'll need to know which way the truck is pointing
all the time, and calculate from there. I should mention the mount is
metal, so that could affect a compass. It would have to be mounted in
a way to minimize errors.
Thanks!!
-mpm
On top of this pole, we have a camera and yagi antenna mounted to a
simple pan-tilt-zoom controller.
We would like to know which way the contraption is pointed, regardless
of the truck's orientation.
What is the easiest way to do this? (Surely this requirement has been
beat to death, right??)
The PTZ controller is very simple: left, right, up, down. No azimuth
info. No known starting posiition regardless of azimuth (though we
might be able to cobble something up?).
While the camera could possibly "look at" something on the mount (like
a compass?) in addition to it's normal intended field of view, our
preference would be something other than "visual-only", maybe a 3-
digit RS-232 feed, or something we can at least have the option to
capture via computer. But hey, we'd take visual-only if that's the
only viable way to do it without upgrading to a more capable PTZ
controller. I don't have the camera lens specs handy, but it is an
auto-focus, auto-iris design.
Weight is a factor (it's a 30-foot pole) w/ 8-pounds dead weight
already.
Simplicity is a factor, as is low-cost.
Another possibility is to have the contraption superimpose the azimuth
data on the NTSC 1-volt video feed coming off the camera itself
(somehow). That would be ideal, actually.
Worst case, we'll stick a colored ball on a stick, and use that as a
reference, but then we'll need to know which way the truck is pointing
all the time, and calculate from there. I should mention the mount is
metal, so that could affect a compass. It would have to be mounted in
a way to minimize errors.
Thanks!!
-mpm
lang...@fonz.dk
Feb 12, 2012, 3:38:48 PM2/12/12
to
add a small micro if you want rs232
-Lasse
Don Y
Feb 12, 2012, 5:18:50 PM2/12/12
to
On 2/12/2012 1:17 PM, mpm wrote:
> We have a mobile truck equipped with a pneumatic push-up pole.
> On top of this pole, we have a camera and yagi antenna mounted to a
> simple pan-tilt-zoom controller.
> We would like to know which way the contraption is pointed, regardless
> of the truck's orientation.
By that, I assume your reference to azimuth is wrt magnetic north
> We have a mobile truck equipped with a pneumatic push-up pole.
> On top of this pole, we have a camera and yagi antenna mounted to a
> simple pan-tilt-zoom controller.
> We would like to know which way the contraption is pointed, regardless
> of the truck's orientation.
(or some other geographic reference). Is the truck stationary
at the times you want this information?
I imagine you "stow" the antenna in some configuration while traveling?
(or, *could* do so?). If the procedure you use to deploy the antenna
knows that it will start in this orientation (wrt the truck), you
can arrange for it to pass through a reference point, rotationally.
"Top dead center" (i.e., "front dead center" :> ).
Any sort of incremental encoder could then be used to track the
movement relative to that reference -- reestablishing the reference
any time it passes through it (though I assume you can't go 'round
and 'round and 'round indefinitely -- unless you have a slip-ring-ish
arrangement).
You could also use an absolute position encoder depending on the
resolution required. Have any synchro resolvers lying around?? :>
RVDT's, optical encoders, etc.
I've seen "low resolution" optical encoders "cascaded" (mechanical
gain on the shafts) to make low cost high(er) resolution encoders.
And/or used to augment an "index/reference" position indicator.
> What is the easiest way to do this? (Surely this requirement has been
> beat to death, right??)
> The PTZ controller is very simple: left, right, up, down. No azimuth
> info. No known starting posiition regardless of azimuth (though we
> might be able to cobble something up?).
So, you can't even reliably determine how much the mechanism
*should* have moved (stepping motors -- operated closed loop -- could
provide that information if you can "count steps").
> While the camera could possibly "look at" something on the mount (like
> a compass?) in addition to it's normal intended field of view, our
> preference would be something other than "visual-only", maybe a 3-
> digit RS-232 feed, or something we can at least have the option to
> capture via computer. But hey, we'd take visual-only if that's the
> only viable way to do it without upgrading to a more capable PTZ
> controller. I don't have the camera lens specs handy, but it is an
> auto-focus, auto-iris design.
reflecting an image of a compass into the video feed?
> Weight is a factor (it's a 30-foot pole) w/ 8-pounds dead weight
> already.
> Simplicity is a factor, as is low-cost.
>
> Another possibility is to have the contraption superimpose the azimuth
> data on the NTSC 1-volt video feed coming off the camera itself
> (somehow). That would be ideal, actually.
having an "opacity gradient". A light passes through the plate and the
sensed signal indicates "how dark" the portion of the disk in the light
path is. (Often, a pair of such detectors operating in a sort of
bridge configuration)
How accurate do you need to be? How fast does the antenna's orientation
change (i.e. via the PTZ controller *and* any motion that the truck
imparts to it)?
> Worst case, we'll stick a colored ball on a stick, and use that as a
field of vision. E.g., like points on a compass referenced to the
front of the *truck*.
John Fields
Feb 12, 2012, 7:35:51 PM2/12/12
to
On Sun, 12 Feb 2012 12:17:19 -0800 (PST), mpm <mpmi...@aol.com>
wrote:
>We have a mobile truck equipped with a pneumatic push-up pole.
>On top of this pole, we have a camera and yagi antenna mounted to a
>simple pan-tilt-zoom controller.
>We would like to know which way the contraption is pointed, regardless
>of the truck's orientation.
---
With respect to what?
---
>What is the easiest way to do this? (Surely this requirement has been
>beat to death, right??)
>
>The PTZ controller is very simple: left, right, up, down. No azimuth
>info. No known starting posiition regardless of azimuth (though we
>might be able to cobble something up?).
>
>While the camera could possibly "look at" something on the mount (like
>a compass?) in addition to it's normal intended field of view, our
>preference would be something other than "visual-only", maybe a 3-
>digit RS-232 feed, or something we can at least have the option to
>capture via computer. But hey, we'd take visual-only if that's the
>only viable way to do it without upgrading to a more capable PTZ
>controller. I don't have the camera lens specs handy, but it is an
>auto-focus, auto-iris design.
>
>Weight is a factor (it's a 30-foot pole) w/ 8-pounds dead weight
>already.
>Simplicity is a factor, as is low-cost.
>
>Another possibility is to have the contraption superimpose the azimuth
>data on the NTSC 1-volt video feed coming off the camera itself
>(somehow). That would be ideal, actually.
>
>Worst case, we'll stick a colored ball on a stick, and use that as a
>reference, but then we'll need to know which way the truck is pointing
>all the time, and calculate from there. I should mention the mount is
>metal, so that could affect a compass. It would have to be mounted in
>a way to minimize errors.
---
How about a magnetic compass used to determine the vehicle's heading,
and an absolute encoder at the PTZ with its zero set to reference the
vehicle's forward looking axis?
--
JF
wrote:
>We have a mobile truck equipped with a pneumatic push-up pole.
>On top of this pole, we have a camera and yagi antenna mounted to a
>simple pan-tilt-zoom controller.
>We would like to know which way the contraption is pointed, regardless
>of the truck's orientation.
With respect to what?
---
>What is the easiest way to do this? (Surely this requirement has been
>beat to death, right??)
>
>The PTZ controller is very simple: left, right, up, down. No azimuth
>info. No known starting posiition regardless of azimuth (though we
>might be able to cobble something up?).
>
>While the camera could possibly "look at" something on the mount (like
>a compass?) in addition to it's normal intended field of view, our
>preference would be something other than "visual-only", maybe a 3-
>digit RS-232 feed, or something we can at least have the option to
>capture via computer. But hey, we'd take visual-only if that's the
>only viable way to do it without upgrading to a more capable PTZ
>controller. I don't have the camera lens specs handy, but it is an
>auto-focus, auto-iris design.
>
>Weight is a factor (it's a 30-foot pole) w/ 8-pounds dead weight
>already.
>Simplicity is a factor, as is low-cost.
>
>Another possibility is to have the contraption superimpose the azimuth
>data on the NTSC 1-volt video feed coming off the camera itself
>(somehow). That would be ideal, actually.
>
>Worst case, we'll stick a colored ball on a stick, and use that as a
>reference, but then we'll need to know which way the truck is pointing
>all the time, and calculate from there. I should mention the mount is
>metal, so that could affect a compass. It would have to be mounted in
>a way to minimize errors.
How about a magnetic compass used to determine the vehicle's heading,
and an absolute encoder at the PTZ with its zero set to reference the
vehicle's forward looking axis?
--
JF
Jim Thompson
Feb 12, 2012, 7:37:10 PM2/12/12
to
...Jim Thompson
--
| James E.Thompson, CTO | mens |
| Analog Innovations, Inc. | et |
| Analog/Mixed-Signal ASIC's and Discrete Systems | manus |
| Phoenix, Arizona 85048 Skype: Contacts Only | |
| Voice:(480)460-2350 Fax: Available upon request | Brass Rat |
| E-mail Icon at http://www.analog-innovations.com | 1962 |
I love to cook with wine. Sometimes I even put it in the food.
brent
Feb 12, 2012, 7:50:29 PM2/12/12
to
On Feb 12, 7:37 pm, Jim Thompson <To-Email-Use-The-Envelope-I...@On-My-
> >On Sun, 12 Feb 2012 12:17:19 -0800 (PST), mpm <mpmill...@aol.com>
> | E-mail Icon athttp://www.analog-innovations.com| 1962 |
Web-Site.com> wrote:
> On Sun, 12 Feb 2012 18:35:51 -0600, John Fields
>
>
>
>
>
>
>
>
>
> <jfie...@austininstruments.com> wrote:
> On Sun, 12 Feb 2012 18:35:51 -0600, John Fields
>
>
>
>
>
>
>
>
>
> >On Sun, 12 Feb 2012 12:17:19 -0800 (PST), mpm <mpmill...@aol.com>
>
> I love to cook with wine. Sometimes I even put it in the food.
A gps will give position. you only get heading if you are moving.
> I love to cook with wine. Sometimes I even put it in the food.
John Fields
Feb 12, 2012, 7:58:27 PM2/12/12
to
Possibly useful for "Where am I?" and maybe, even, for "How high is my
camera off the ground" but, I think, probably problematical for "In
which way is my camera pointing?"
--
JF
lang...@fonz.dk
Feb 12, 2012, 7:59:23 PM2/12/12
to
heading
-Lasse
mpm
Feb 12, 2012, 9:06:01 PM2/12/12
to
>
> - Show quoted text -
Thanks,
I also found this online, but don't know anything about them (yet), or
if they're even available.
I will check out the SparkFun product as well.
-mpm
brent
Feb 12, 2012, 9:09:40 PM2/12/12
to
-----------------------------------------
http://en.wikipedia.org/wiki/Compass
states:
There are other devices which are not conventionally called compasses
but which do allow the true cardinal directions to be determined. They
are said to work "like a compass", or "as a compass", without actually
being a compass. For example, a Global Positioning System (GPS)
satellite receiver determines its own position on the ground, as true
latitude and true longitude. If the receiver is being moved, even at
walking pace, it can follow the change of its position, and hence
determine the compass bearing of its direction of movement, and thence
the directions of the cardinal points relative to its direction of
movement. Some GPS receivers have two antennas, fixed some distance
apart to the structure of a vehicle, usually an aircraft. The exact
latitudes and longitudes of the antennas can be determined
simultaneously, which allows the directions of the cardinal points to
be calculated relative to the heading of the aircraft ........
------------------------------------------
Nico Coesel
Feb 12, 2012, 9:25:22 PM2/12/12
to
"lang...@fonz.dk" <lang...@fonz.dk> wrote:
>> > I love to cook with wine. =A0 =A0 Sometimes I even put it in the food.
>>
>> A gps will give position. =A0you only get heading if you are moving.
(yes, it really is that bad in real life). Useless for determining the
heading. An electronics compass is probably the best solution. But
expect some error. The compass I worked with had an error of about 16
degrees. But there may be better ones available.
--
Failure does not prove something is impossible, failure simply
indicates you are not using the right tools...
nico@nctdevpuntnl (punt=.)
--------------------------------------------------------------
>> > I love to cook with wine. =A0 =A0 Sometimes I even put it in the food.
>>
>> A gps will give position. =A0you only get heading if you are moving.
>
>
>I guess two GPS recievers with sufficient distance could give you
>heading
That won't work. In my experience GPS can be of by 150m / 450feet
>
>I guess two GPS recievers with sufficient distance could give you
>heading
(yes, it really is that bad in real life). Useless for determining the
heading. An electronics compass is probably the best solution. But
expect some error. The compass I worked with had an error of about 16
degrees. But there may be better ones available.
--
Failure does not prove something is impossible, failure simply
indicates you are not using the right tools...
nico@nctdevpuntnl (punt=.)
--------------------------------------------------------------
mpm
Feb 12, 2012, 9:22:33 PM2/12/12
to
On Feb 12, 5:18 pm, Don Y <t...@isnotme.com> wrote:
> By that, I assume your reference to azimuth is wrt magnetic north
> (or some other geographic reference). Is the truck stationary
> at the times you want this information? [...snip..]
> By that, I assume your reference to azimuth is wrt magnetic north
> (or some other geographic reference). Is the truck stationary
Yes - position relative to true north is preferred, but magnetic north
will do.
The truck will be stationary. I can almost picture the insurance
adjuster's face when we explain how 30-foot extended mast meets
highway overpass. :)
Yes - travel is not round and round. It's actually only about 300
degrees.
More than enough for what we do with it, which is interference
abatement.
And we don't need it every job. Maybe 5%-10% of the tough ones, out
in the boonies usually with no natural elevation opportunities, and by
"natural" I mean multi-story parking garages of course.
Resolution - Realistically, if we can get 3 to 5 degrees I think we'd
be happy.
Within that range, repeatability is probably more important that
absolute accuracy.
As to rotation speed, the mast is straight-up/down. The PTZ is fairly
slow, though I've not measured it.
My guess is it takes 15-20 seconds (at least) to travel 300 degrees.
You gave me an idea. (..and I'm not sure it's a good one yet, but it
qualifies for low-tech..?)
Since we always have the mast full-up when deployed, maybe we could
use a wire, cobble up some kind of tension gear, and just measure the
length of wire travel with something "widget" mounted back at the
truck?
It doesn't necessarily solve the initial "which way is North", but I
haven't flushed it out yet either. Maybe a quick pre-calibration
with a Boy Scout compass before raising the mast?
Then again, 30-feet of wire flapping in the breeze might be
problematic to measure (unless we route it inside the mast?)
The camera prism thing sounds messy, though I agree it would likely
work. Would just have to review the camera focus specs, etc...
brent
Feb 12, 2012, 9:29:20 PM2/12/12
to
On Feb 12, 9:25 pm, n...@puntnl.niks (Nico Coesel) wrote:
If two are used and compared the absolute error should be the same for
both. I think you can get heading with two receivers. However, it
may fail the requirement for "cheap".
both. I think you can get heading with two receivers. However, it
may fail the requirement for "cheap".
mpm
Feb 12, 2012, 9:23:55 PM2/12/12
to
>
> - Show quoted text -
Oops - did not include the link.
> - Show quoted text -
Here it is:
http://www51.honeywell.com/aero/common/documents/myaerospacecatalog-documents/Space-documents/Digital_Compass_Solution_HMR3400.pdf
mpm
Feb 12, 2012, 9:35:22 PM2/12/12
to
On Feb 12, 9:25 pm, n...@puntnl.niks (Nico Coesel) wrote:
may even be a requirement) that differential GPS setups like we're
describing here only work when both GPS units are using the exact same
constellation of satellites in calculating their position fixes. If
one GPS uses a different set of satellites, it can really throw off
the calculated results - even if the sets differ by as little as a
single satellite.
That could be complete BS. I'm not an expert on differential GPS.
I know my Garmin Montana 650t often shows a position error estimate
nearly as long as the truck.
Great GPS by the way - terribly pricey, but it does everything I need
it to.
16 degrees is too much error, or too limited a resolution.
Don Y
Feb 12, 2012, 10:35:50 PM2/12/12
to
On 2/12/2012 7:22 PM, mpm wrote:
> On Feb 12, 5:18 pm, Don Y<t...@isnotme.com> wrote:
>
>> By that, I assume your reference to azimuth is wrt magnetic north
>> (or some other geographic reference). Is the truck stationary
>> at the times you want this information? [...snip..]
>
> Yes - position relative to true north is preferred, but magnetic north
> will do.
So, anything that doesn't report orientation against this reference
> On Feb 12, 5:18 pm, Don Y<t...@isnotme.com> wrote:
>
>> By that, I assume your reference to azimuth is wrt magnetic north
>> (or some other geographic reference). Is the truck stationary
>> at the times you want this information? [...snip..]
>
> Yes - position relative to true north is preferred, but magnetic north
> will do.
would require having something in the truck that does so.
> The truck will be stationary. I can almost picture the insurance
> adjuster's face when we explain how 30-foot extended mast meets
> highway overpass. :)
remembered to secure the antenna (narrow edge to the wind). But,
you didn't always remember to turn off the transmitter! :-/
> Yes - travel is not round and round. It's actually only about 300
> degrees.
> More than enough for what we do with it, which is interference
> abatement.
> And we don't need it every job. Maybe 5%-10% of the tough ones, out
> in the boonies usually with no natural elevation opportunities, and by
> "natural" I mean multi-story parking garages of course.
>
> Resolution - Realistically, if we can get 3 to 5 degrees I think we'd
> be happy.
> Within that range, repeatability is probably more important that
> absolute accuracy.
the antenna) POINTED UP. Feed this to a PC and superimpose a set of
compass bearings on it -- from knowledge of the direction the truck is
facing. The antenna would always be a fixed distance away from the
camera (or, in a very narrow range). Paint a big red line on the
bottom of the antenna. A piece of software superimposes a *rotated*
N-S-E-W (marked in 5 degree increments) on the image, centered at the
mast. Big red line looks like the needle of the "compass" on the
screen. (i.e., use your eyes to resolve the position)
> As to rotation speed, the mast is straight-up/down. The PTZ is fairly
> slow, though I've not measured it.
> My guess is it takes 15-20 seconds (at least) to travel 300 degrees.
degrees to get an idea as to which general direction it is facing.
"OK, Bob, looks like we just spotted something at around 245 degrees.
Why don't you stop and back up a bit... easy... easy... *there*!"
> You gave me an idea. (..and I'm not sure it's a good one yet, but it
> qualifies for low-tech..?)
> Since we always have the mast full-up when deployed, maybe we could
> use a wire, cobble up some kind of tension gear, and just measure the
> length of wire travel with something "widget" mounted back at the
> truck?
a mast-mounted bobbin, an LVDT driven by a rack-and-pinion, etc.
You could also just fasten a big red *pointer* (broomstick?)
on the bottom of the mast INSIDE the truck (if the whole mast were
to rotate).
Or, a large NSEW ring mounted on the roof of the truck that you
"spin" to face North (or wherever) when you set up.
> It doesn't necessarily solve the initial "which way is North", but I
> haven't flushed it out yet either. Maybe a quick pre-calibration
> with a Boy Scout compass before raising the mast?
> Then again, 30-feet of wire flapping in the breeze might be
> problematic to measure (unless we route it inside the mast?)
>
> The camera prism thing sounds messy, though I agree it would likely
> work. Would just have to review the camera focus specs, etc...
antenna is looking at? (so you know whose door to kick down?)
Jon Elson
Feb 13, 2012, 1:23:25 AM2/13/12
to
mpm wrote:
> We have a mobile truck equipped with a pneumatic push-up pole.
> On top of this pole, we have a camera and yagi antenna mounted to a
> simple pan-tilt-zoom controller.
> We would like to know which way the contraption is pointed, regardless
> of the truck's orientation.
The way I saw it done years ago was the azimuth was manual, and there
> We have a mobile truck equipped with a pneumatic push-up pole.
> On top of this pole, we have a camera and yagi antenna mounted to a
> simple pan-tilt-zoom controller.
> We would like to know which way the contraption is pointed, regardless
> of the truck's orientation.
was a compass rose built into the ceiling of the station wagon.
The mast was a telescoping air cylinder with a key in the sections,
so the top stayed aligned with the bottom. Once on site, they
used a big surveyor's magnetic compass a distance from the
vehicle, and sighted through the sight slots on it through the
mast, and somebody inside the vehicle turned the compass rose to
align to the compass sights. The compass rose was locked, and
now you had your north reference. This was used for FCC field
strength surveys, so it had to be pretty accurate.
The mast was elevated by 12 V DC compressor, and if the air relief valve
was closed or the power to the compressor was turned on, a HUGE
red light started flashing inside the car and a buzzer went off.
That pretty much made sure the mast stayed down when it was in
motion. (When I say huge, it was an after-market truck taillight.)
Jon
Rich Webb
Feb 12, 2012, 5:38:31 PM2/12/12
to
On Sun, 12 Feb 2012 12:17:19 -0800 (PST), mpm <mpmi...@aol.com> wrote:
>We have a mobile truck equipped with a pneumatic push-up pole.
>On top of this pole, we have a camera and yagi antenna mounted to a
>simple pan-tilt-zoom controller.
>We would like to know which way the contraption is pointed, regardless
>of the truck's orientation.
>
>What is the easiest way to do this? (Surely this requirement has been
>beat to death, right??)
A small fluxgate magnetic compass. There are commercial marine-grade
>We have a mobile truck equipped with a pneumatic push-up pole.
>On top of this pole, we have a camera and yagi antenna mounted to a
>simple pan-tilt-zoom controller.
>We would like to know which way the contraption is pointed, regardless
>of the truck's orientation.
>
>What is the easiest way to do this? (Surely this requirement has been
>beat to death, right??)
instruments that spit out NMEA sentences that, with a correction for
your local magnetic variation, can be used to give a pretty good
indication of true north. If you already have a GPS in your truck, there
are units that have an internal "map" of the global magnetic variation
so it can be pretty much automatic; feed NMEA position in from the GPS
and get NMEA (or displayed) heading out from the compass.
Or, if you really need it, a FOG (fiber-optic gyrocompass) can give a
reasonably accurate heading reference that's immune from the effects of
nearby metal and (maybe) won't break the bank.
--
Rich Webb Norfolk, VA
Tauno Voipio
Feb 13, 2012, 8:51:26 AM2/13/12
to
please check with a magnetic compass how much disturbance
comes from the auto structure. To get absolute directions,
you'll need some kind of magnetic reference system (i.e.
a fancy compass).
--
Tauno Voipio
Rich Webb
Feb 13, 2012, 8:53:48 AM2/13/12
to
On Mon, 13 Feb 2012 02:25:22 GMT, ni...@puntnl.niks (Nico Coesel) wrote:
>"lang...@fonz.dk" <lang...@fonz.dk> wrote:
>
>>> > I love to cook with wine. =A0 =A0 Sometimes I even put it in the food.
>>>
>>> A gps will give position. =A0you only get heading if you are moving.
>>
>>
>>I guess two GPS recievers with sufficient distance could give you
>>heading
>
>That won't work. In my experience GPS can be of by 150m / 450feet
>(yes, it really is that bad in real life). Useless for determining the
>heading. An electronics compass is probably the best solution. But
>expect some error. The compass I worked with had an error of about 16
>degrees. But there may be better ones available.
Actually, there are "satellite compasses" that use two (or more) GPS
>"lang...@fonz.dk" <lang...@fonz.dk> wrote:
>
>>> > I love to cook with wine. =A0 =A0 Sometimes I even put it in the food.
>>>
>>> A gps will give position. =A0you only get heading if you are moving.
>>
>>
>>I guess two GPS recievers with sufficient distance could give you
>>heading
>
>That won't work. In my experience GPS can be of by 150m / 450feet
>(yes, it really is that bad in real life). Useless for determining the
>heading. An electronics compass is probably the best solution. But
>expect some error. The compass I worked with had an error of about 16
>degrees. But there may be better ones available.
antennas. They essentially track the carrier phase difference (similar
to a conventional RDF antenna but done with DSP). The ones I'm familiar
with spec a heading accuracy of 1.0 to 0.5 degrees; cost goes up with
accuracy, of course.
They'll run from 12 or 24 VDC; should be a decent fit for a truck/van.
I'm most familiar with the Furuno products but there are probably others
out there.
<http://www.furunousa.com/Products/ProductList.aspx?category=Products+%3a+Satellite+Compasses+%26+Heading+Sensors+%3a+Satellite+Compasses>
Jim Thompson
Feb 13, 2012, 9:48:52 AM2/13/12
to
>
>A gps will give position. you only get heading if you are moving.
You mean the truck go there without moving ?:-)
>A gps will give position. you only get heading if you are moving.
...Jim Thompson
--
| James E.Thompson, CTO | mens |
| Analog Innovations, Inc. | et |
| Analog/Mixed-Signal ASIC's and Discrete Systems | manus |
| Phoenix, Arizona 85048 Skype: Contacts Only | |
| Voice:(480)460-2350 Fax: Available upon request | Brass Rat |
Jim Thompson
Feb 13, 2012, 9:51:40 AM2/13/12
to
quite well.
VWWall
Feb 13, 2012, 12:06:55 PM2/13/12
to
mpm wrote:
> We have a mobile truck equipped with a pneumatic push-up pole.
> On top of this pole, we have a camera and yagi antenna mounted to a
> simple pan-tilt-zoom controller.
> We would like to know which way the contraption is pointed, regardless
> of the truck's orientation.
>
> What is the easiest way to do this? (Surely this requirement has been
> beat to death, right??)
>
I was in the Army Signal Corp in North Africa and Italy during WW2 and
> We have a mobile truck equipped with a pneumatic push-up pole.
> On top of this pole, we have a camera and yagi antenna mounted to a
> simple pan-tilt-zoom controller.
> We would like to know which way the contraption is pointed, regardless
> of the truck's orientation.
>
> What is the easiest way to do this? (Surely this requirement has been
> beat to death, right??)
>
we used a passenger car with a steerable loop antenna mounted on the
roof to locate suspicious radio signals. When we took a bearing we
manually turned the loop approximately to the expected signal direction.
With a standard siting compass, standing away from the car we measured
the azimuth to which the loop was pointing. This "calibrated" the
azimuth indicator in the car which showed loop's rotational position.
If there is no azimuth indicator in the vehicle, a simple synchro
repeater could be attached to the pan axis.
I would expect the yagi antenna would be sufficient as an alignment
element to sight from the compass. Probably good to +-5 degrees.
--
Virg Wall
Peter Bennett
Feb 13, 2012, 1:23:47 PM2/13/12
to
On Mon, 13 Feb 2012 02:25:22 GMT, ni...@puntnl.niks (Nico Coesel)
wrote:
wrote:
>"lang...@fonz.dk" <lang...@fonz.dk> wrote:
>
>>> > I love to cook with wine. =A0 =A0 Sometimes I even put it in the food.
>>>
>>> A gps will give position. =A0you only get heading if you are moving.
>>
>>
>>I guess two GPS recievers with sufficient distance could give you
>>heading
>
>That won't work. In my experience GPS can be of by 150m / 450feet
>(yes, it really is that bad in real life). Useless for determining the
>heading. An electronics compass is probably the best solution. But
>expect some error. The compass I worked with had an error of about 16
>degrees. But there may be better ones available.
It probably won't work using two stand-alone GPS receivers, but there
>
>>> > I love to cook with wine. =A0 =A0 Sometimes I even put it in the food.
>>>
>>> A gps will give position. =A0you only get heading if you are moving.
>>
>>
>>I guess two GPS recievers with sufficient distance could give you
>>heading
>
>That won't work. In my experience GPS can be of by 150m / 450feet
>(yes, it really is that bad in real life). Useless for determining the
>heading. An electronics compass is probably the best solution. But
>expect some error. The compass I worked with had an error of about 16
>degrees. But there may be better ones available.
are GPS-based compasses. I've seen one - from Furuno, I think - that
had two or three antennas in a single housing about 3 ft diameter.
--
Peter Bennett, VE7CEI
peterbb (at) telus.net
GPS and NMEA info: http://vancouver-webpages.com/peter
Vancouver Power Squadron: http://vancouver.powersquadron.ca
Message has been deleted
Jim Thompson
Feb 14, 2012, 12:39:16 PM2/14/12
to
On Tue, 14 Feb 2012 09:07:34 -0800, Fred Abse
<excret...@invalid.invalid> wrote:
>On Mon, 13 Feb 2012 07:51:40 -0700, Jim Thompson wrote:
>
>> My car (Infiniti Q45) has some kind of flux-gate compass that works
>> quite well.
>
>So does my Jeep. Not sure about its being fluxgate, though, there are some
>magnetoresistive devices (Honeywell, I think), that respond to geomagnetic
>field levels.
Yep. I designed such a chip for Honeywell Sensor Division
(Minneapolis) in 2007-2008... took 18 months... lots of bells and
whistles.
>
>What resolution does the Q45 compass have? The Jeep only does
>half-quadrant.
In the mirror text, half-quadrant, on the map screen fairly smooth, so
I haven't taken note of the granularity.
<excret...@invalid.invalid> wrote:
>On Mon, 13 Feb 2012 07:51:40 -0700, Jim Thompson wrote:
>
>> My car (Infiniti Q45) has some kind of flux-gate compass that works
>> quite well.
>
>magnetoresistive devices (Honeywell, I think), that respond to geomagnetic
>field levels.
Yep. I designed such a chip for Honeywell Sensor Division
(Minneapolis) in 2007-2008... took 18 months... lots of bells and
whistles.
>
>What resolution does the Q45 compass have? The Jeep only does
>half-quadrant.
In the mirror text, half-quadrant, on the map screen fairly smooth, so
I haven't taken note of the granularity.
Paul Hovnanian P.E.
Feb 14, 2012, 2:20:50 PM2/14/12
to
mpm wrote:
> We have a mobile truck equipped with a pneumatic push-up pole.
> On top of this pole, we have a camera and yagi antenna mounted to a
> simple pan-tilt-zoom controller.
> We would like to know which way the contraption is pointed, regardless
> of the truck's orientation.
>
> What is the easiest way to do this? (Surely this requirement has been
> beat to death, right??)
>
> We have a mobile truck equipped with a pneumatic push-up pole.
> On top of this pole, we have a camera and yagi antenna mounted to a
> simple pan-tilt-zoom controller.
> We would like to know which way the contraption is pointed, regardless
> of the truck's orientation.
>
> What is the easiest way to do this? (Surely this requirement has been
> beat to death, right??)
>
> The PTZ controller is very simple: left, right, up, down. No azimuth
> info. No known starting posiition regardless of azimuth (though we
> might be able to cobble something up?).
>
> While the camera could possibly "look at" something on the mount (like
> a compass?) in addition to it's normal intended field of view, our
> preference would be something other than "visual-only", maybe a 3-
> digit RS-232 feed, or something we can at least have the option to
> capture via computer. But hey, we'd take visual-only if that's the
> only viable way to do it without upgrading to a more capable PTZ
> controller. I don't have the camera lens specs handy, but it is an
> auto-focus, auto-iris design.
>
> Weight is a factor (it's a 30-foot pole) w/ 8-pounds dead weight
> already.
> Simplicity is a factor, as is low-cost.
>
> Another possibility is to have the contraption superimpose the azimuth
> data on the NTSC 1-volt video feed coming off the camera itself
> (somehow). That would be ideal, actually.
>
> Worst case, we'll stick a colored ball on a stick, and use that as a
> reference, but then we'll need to know which way the truck is pointing
> all the time, and calculate from there. I should mention the mount is
> metal, so that could affect a compass. It would have to be mounted in
> a way to minimize errors.
>
> Thanks!!
> -mpm
http://www.parallax.com/Store/Sensors/CompassGPS/tabid/173/List/0/CategoryID/48/Level/a/SortField/0/Default.aspx
> info. No known starting posiition regardless of azimuth (though we
> might be able to cobble something up?).
>
> While the camera could possibly "look at" something on the mount (like
> a compass?) in addition to it's normal intended field of view, our
> preference would be something other than "visual-only", maybe a 3-
> digit RS-232 feed, or something we can at least have the option to
> capture via computer. But hey, we'd take visual-only if that's the
> only viable way to do it without upgrading to a more capable PTZ
> controller. I don't have the camera lens specs handy, but it is an
> auto-focus, auto-iris design.
>
> Weight is a factor (it's a 30-foot pole) w/ 8-pounds dead weight
> already.
> Simplicity is a factor, as is low-cost.
>
> Another possibility is to have the contraption superimpose the azimuth
> data on the NTSC 1-volt video feed coming off the camera itself
> (somehow). That would be ideal, actually.
>
> Worst case, we'll stick a colored ball on a stick, and use that as a
> reference, but then we'll need to know which way the truck is pointing
> all the time, and calculate from there. I should mention the mount is
> metal, so that could affect a compass. It would have to be mounted in
> a way to minimize errors.
>
> Thanks!!
> -mpm
Issues with local magnetic field distortion can be handled if you can raise
the camera into position and then rotate it several times continuously. A
microprocessor (not included) can sore a map of the local field anomalies
and find magnetic north (my watch's compass works the same way).
A uP in the camera head can handle all of this (Parallax sells these as
well) plus encode the resulting data into the NTSC blanking lines. Or you
could use a coupe of low power RF modules to send the data down sans wires.
--
Paul Hovnanian mailto:Pa...@Hovnanian.com
------------------------------------------------------------------
Quantum Mechanics: The dreams stuff is made of.
Baron
Feb 14, 2012, 4:29:49 PM2/14/12
to
John Fields Inscribed thus:
Doesn't Garmin make a hand held compass !
--
Best Regards:
Baron.
--
Best Regards:
Baron.
Nico Coesel
Feb 14, 2012, 6:29:47 PM2/14/12
to
Peter Bennett <pet...@somewhere.invalid> wrote:
>On Mon, 13 Feb 2012 02:25:22 GMT, ni...@puntnl.niks (Nico Coesel)
>wrote:
>
>>"lang...@fonz.dk" <lang...@fonz.dk> wrote:
>>
>>>> > I love to cook with wine. =A0 =A0 Sometimes I even put it in the food.
>>>>
>>>> A gps will give position. =A0you only get heading if you are moving.
>>>
>>>
>>>I guess two GPS recievers with sufficient distance could give you
>>>heading
>>
>>That won't work. In my experience GPS can be of by 150m / 450feet
>>(yes, it really is that bad in real life). Useless for determining the
>>heading. An electronics compass is probably the best solution. But
>>expect some error. The compass I worked with had an error of about 16
>>degrees. But there may be better ones available.
>
>It probably won't work using two stand-alone GPS receivers, but there
>are GPS-based compasses. I've seen one - from Furuno, I think - that
>had two or three antennas in a single housing about 3 ft diameter.
Those probably only work in a rural area with a clear view to the sky.
>On Mon, 13 Feb 2012 02:25:22 GMT, ni...@puntnl.niks (Nico Coesel)
>wrote:
>
>>"lang...@fonz.dk" <lang...@fonz.dk> wrote:
>>
>>>> > I love to cook with wine. =A0 =A0 Sometimes I even put it in the food.
>>>>
>>>> A gps will give position. =A0you only get heading if you are moving.
>>>
>>>
>>>I guess two GPS recievers with sufficient distance could give you
>>>heading
>>
>>That won't work. In my experience GPS can be of by 150m / 450feet
>>(yes, it really is that bad in real life). Useless for determining the
>>heading. An electronics compass is probably the best solution. But
>>expect some error. The compass I worked with had an error of about 16
>>degrees. But there may be better ones available.
>
>It probably won't work using two stand-alone GPS receivers, but there
>are GPS-based compasses. I've seen one - from Furuno, I think - that
>had two or three antennas in a single housing about 3 ft diameter.
As soon as you get close to buildings or an airforce base (radar) GPS
is very inaccurate.
0 new messages