Orient Test Stand True North

I would think you could hire a good surveyor to establish an astronomic-north bearing on a line of site that you could visually aim at with a known bearing. I don't know how you turn your "stand" from a known angle to true north, but a surveyor could come off of the known line and establish a new point outside that is a more precise "astronomic north" for instance. There can be a minute (My-noot, not min-ut) difference between geodetic north and astronomic north, but I imagine that difference is immaterial for most applications. You need to explain the purpose of this "stand" a little or at least let someone know the order of precision you need. GPS can come up with extremely accurate bearings as can a good stellar observation with a precise theodolite.

Not sure I said anything helpful, and don't really know exactly what your "stand" and overall setup is.

I appreciate the reply. Astronomic North is something we were exploring as well, and the difference between that and geodetic north would be fine.

The stand rotates on each axis with a motor on a rotary table essentially. Each has position encoders and we could likely rotate and stop accurately based off of a given heading of a known line.

The stand is used to test and calibrate sensors, mostly used in the oil industry to maintain directions underground. For the north heading, we were probably looking on the order of 30-40 arcseconds of accuracy if possible (0.01 degrees), though we could probably be ok with up to 0.05 degrees. So, we want the location data of the stand for magnetic field calculations, and the heading for calibration purposes.

A mockup model of the stand is shown below, each silver disc is the rotation table, and the cylinder represents a sensor. This would be the 'horizontal' position, with the sensor pointing north/south.

Five arc seconds is easy to do for a Polaris shot. Your test stand is going to need some sort of optical sighting apparatus in order to have repeatable accuracy after your reference point (mire) is established. Most common way to do that is with a rifle's telescopic sight. Keeping it indoors will allow it to maintain a long life as long as you don't physically hit it out of zero. Purchasing one that is nitrogen-filled will give some semblance of quality for a rifle scope. Weaver is a good brand, as well as Redfield and Leupold.

A competent surveyor can readily establish a bearing between two intervisible points to well-within your 30-second precision standard. I don't know if they can occupy the actual point where the stand is (or is to be installed), but if not, they can set on a point (monument) and establish a bearing to the stand. Even if they can set over the stand, they may not be able to collect reliable satellite data to utilize GPS. But regardless, someone can establish a bearing to the stand from a point in the field where they can see it from.

Two ways, it might be established could include establish a bearing between two known points in the field using GPS and then turning sets of angles from the established line to the stand. Another is to turn astronomic observations to the stand from a known monument. The good surveyor will find a way to double-check his established bearing (from two different points, or other means) to have a strong confidence level in what he is going to provide you. If wanted, after he has the known bearing established, he can probably put a point on line directly north of your stand if that meets your needs (and if you have a line of sight to that point).

The line of sight to the outdoors would not be line of sight to a north heading in the room with the stand. So, we would get an accurate heading to somewhere with line of sight in the field, and then rotate the stand to its heading? We can do that quite accurately I would think. Is there a way to confirm a north heading?

When a heading is being established, what is a good way to make sure the heading is the actual direction of the sensor (the center of the cylinder in the drawing). Would we use an optical sight? A laser pointer to a monument? I'm just curious what might be best or possible to actually align the system with the heading provided by a surveyor.

Any advice on what to ask for in a surveyor when exploring options for this application? Or what kind of level of detail we should ask for when one is selected?

Also, for information purposes, about what is the difference between true north and astronomic north? A few seconds?

Thanks a lot

What you are calling "true" I think most of use would consider "geodetic" north. Mainly because I would say that someone else might consider "true north" to be the astronomic north. The few seconds (or < 1") difference between the geodetic and astronomic north is just kind of the difference in a line pointing in outerspace and the (curved) line along the surface of the earth going straight to the north pole. Just some calculable difference and the line that the surveyor would provide you would probably be +/- a few seconds anyway making that difference negligible.

I don't know where you are going to put a point (on a wall?) as to where the "exact" north is, but if you don't go very far, the width of your line or mark might make up a few seconds of arc itself, plus the accuracy in the ability to mark that point. Your long line of sight would be the most precise in the different iterations of establishing a line.

I bet others here could give you some better advice as to what to put in your survey requirements. I would give them the leeway on how exactly they are going to establish the bearing; but I would think some kind of stellar observation on a satellite as well as utilizing GPS as a check on each other. A good surveyor can tell you exactly how they established the bearing, and what they did as a check and probably the estimated precision of the final bearing on the line.

The difference between astronomic north and true geodetic north is the Laplace correction. It is non-zero if irregularities in the earth's density cause the gravity in your area to tilt (deflection of the vertical) from the basic earth model. It is typically a few arc seconds but can be much larger in some areas, particularly mountainous areas.

Be sure that your surveyor knows (as most would realize) that you do NOT want state plane coordinate azimuth. If they send a technician to the job who is used to working in SPC he might not think about the (possibly large) difference.

Your surveyor should check the National Geodetic Survey database for the nearest monument that has the Laplace and deflection information, or use the tool linked below, to see if it is significant. If he is sure he can get the sign right, he should correct for it even though it may be small. An estimate can be obtained here from approximate lat-lon coordinates:
http://www.ngs.noaa.gov/GEOID/DEFLEC09/

You need a reasonably long sight to the bearing reference point in order to maintain angular accuracy. Consider how accurately you can identify the axis of rotation of your test stand (centering error) and the uncertainty in the reference mark (width of a scribed line?). To maintain 30 second accuracy you need almost 7000:1 distance to total sideways uncertainty, for example 57 feet with 0.1 inch.

For best accuracy, you should also specify whether you want astronomic latitude and longitude or the more commonly used geodetic value in NAD83(latest flavor), as they have similar differences due to gravity. I suspect, though, that for your purposes a recreational grade GPS would be accurate enough for lat-lon at tenths of an arc second (tens of feet). You wouldn't be able to get a good north bearing using it, though.

DEFLEC12a

The latest vesrion of the deflec program is DEFLEC12a:

http://www.ngs.noaa.gov/cgi-bin/GEOID_STUFF/deflec12A_prompt.prl

Loyal

> I have an equipment test stand that I would like to have accurate lat/long coordinates of, and calibrate so that I know in a certain position, it's heading is true north (as accurately as possible).

This sounds vaguely like something like a service I provided about twenty years ago to an aerospace company that was using a rotatable fixture to test or calibrate navigation gyros. At the time, the most difficult part of the work was getting a latitude and longitude for the test stand by conventional survey methods from distant visible objects of known position. These days, using survey-grade GPS methods, that part of the exercise would be trivial.

The rest of the exercise was quite straightforward and at the time consisted of transferring an azimuth determined by astronomical methods to points indoors. Depending upon what that outdoors space looks like (and how long a line of sight is possible between two convenient outdoors reference points), it may still be best to use astronomical methods over GPS to derive geodetic North. As others have mentioned, probably the critical element of the test stand design is having some fiducial marks on it that reference its axes in a way that can be positioned with survey equipment.

If the angle encoder is accurate on the horizontal turntable, the simplest method to calibrate it for azimuth is to rotate it until a reference line passing through the axis of the turntable's rotation is collinear with the axis of collimation of the telescope of a theodolite or total station oriented to a known geodetic azimuth. Then, rotate the turntable 180 degrees to verify collinearity or to determine the angle encoder reading that is collinear and take the mean of the angle encoder readings (after reducing the second by 180 degrees) as having the nominal geodetic azimuth.

I can't recall whether I had that manufacturer make a special fixture with a long axis, a fixture that attached to the turntable as the sensors did, but that sounds like a good idea.

Just reviewing the post, I had some questions.

It's is one thing to set up a north on a 400 foot line with a very high level of accuracy. But I imagine that you are looking for something smaller. What are you asking for? A north line that is 50, 40, 30, 20, 10, or even five feet will all require a different level of effort to get within your accuracy requirements. Are you asking that the surveyor provide you this north line to a level of accuracy relative to a confidence interval, for example 66% or one sigma, 95% or two sigma, or even 99%. The more you ask for the more it will cost. A surveyor would need to set up an accurate baseline, traverse into your building (no gps inside), traverse through the line and then close out his or her traverse. To provide the accuracy with a confidence interval the surveyor will need to adjust his traverse, most likely using a least squares adjustment, software packages that include least squares reduction can be expensive. I would be interested to hear what you are asking of the surveyor, exactly.

I wish you luck.

For use in confined spaces, collimation is indeed the only way to do it. Just don't disturb the collimator!

I appreciate all of the advice and insight a ton here, it is definitely helping me get a direction on all of this and kind of know what to maybe expect. Just for more of an image, this is a mini schematic of what I am dealing with. The image isn't to any sort of scale, just a quick visual of what I mean by line of sight to the outdoors and where North is. There are two rooms that there is sight through a couple of tall (15') roll up doors. The line of sight can continue outside for a couple hundred feet or so through a parking lot before there is another building.

My idea and how I have been imagining it, is information could be established outside and somehow resolved indoors with line of sight? Is that on track with what might actually happen?

> I appreciate all of the advice and insight a ton here, it is definitely helping me get a direction on all of this and kind of know what to maybe expect. Just for more of an image, this is a mini schematic of what I am dealing with. The image isn't to any sort of scale, just a quick visual of what I mean by line of sight to the outdoors and where North is. There are two rooms that there is sight through a couple of tall (15') roll up doors. The line of sight can continue outside for a couple hundred feet or so through a parking lot before there is another building.
>
> My idea and how I have been imagining it, is information could be established outside and somehow resolved indoors with line of sight? Is that on track with what might actually happen?

As I'm imagining things, the sensor is aligned with the test fixture by some sort of carefully-machined cradle or mounting yoke that allows it to be repeatably positioned in the fixture within some acceptable angular tolerance. The test procedure will involve (among other things) mounting the sensor on the fixture, turning it to some nominal azimuth orientation, and verifying that the sensor outputs a heading that is acceptably close to the nominal value.

Obviously, the mounting method for the sensors is critical to high-accuracy calibration, but assuming that you have that detail in hand, you could use the housing of one of the sensors to make a test fixture for calibrating the reading on the horizontal turntable to a known geodetic azimuth. If the test fixture is unstable enough that the circle reading will need to be recalibrated to geodetic azimuth quite often (such as daily or more frequently), then a more elaborate fixture may be warranted. However, for just an occasional check on an essentially stable setup, I'd think just a couple of small (1mm), needle-pointed alignment pins would work fine in conjunction with a theodolite.

The theodolite would need to be at close range (under 10 ft.) to observe the alignment pins. For repeated use, you'd install a mark in the floor that defined one point on a line of known azimuth to some other mark outdoors. Depending upon the environment and structures, it might be operationally superior to install a target on a building wall more than 80 ft. distant.

Probably the critical detail is how well the sensor housings are machined and how repeatably they can be mounted to the test fixture in the first place. If they can't be mounted with a repeatability significantly better than the accuracy of the calibration you're shooting for, then I'd think the exercise needs to be rethought.

The stand and mounting interface for the gyros is very well machined, very consistent, and very stable. It is all around solid and it treated with the utmost care when being used or touched, so for how we calibrate the gyros, we basically just have to assume it's "perfect" once initially calibrated and perform a calibration check on a regular basis. We are just looking for a better way to do an initial calibration on this as far as the azimuthal heading goes.

We could mount something either in the same mounts that the gyros would go into, or just on to the mounts (we could manufacture some sort of fixture). I guess my question was more of, what would we put on there, or how could that be used to align the system to a heading.

> We could mount something either in the same mounts that the gyros would go into, or just on to the mounts (we could manufacture some sort of fixture). I guess my question was more of, what would we put on there, or how could that be used to align the system to a heading.

Well, depending upon the project budget, you could have an adapter machined that would allow you to mount an alignment telescope such as this made by Brunson to the test fixture to check alignment:

Brunson 2062 Alignment Telescope

Brunson makes adapters for the telescope, so you'd just need to figure out how to attach the adapter to the test fixture.

Then, all you'd need would be to use the alignment telescope in the fixture to establish two ground marks, one near the test fixture and the other more than 100 ft. distant, and have a surveyor determine the geodetic azimuth of the line between them.

And then we would basically know that we could calibrate to ~60deg East of North (because that get's us outdoors), and knowing uncertainties in the rotation base, could rotate to "North" and zero it there?

We can machine what we'd need here with an in house machine shop, but are there sensors or equipment that a surveyor may have that we would not need to purchase? And then we could somehow mark our systems for future calibration checks?

> And then we would basically know that we could calibrate to ~60deg East of North (because that get's us outdoors), and knowing uncertainties in the rotation base, could rotate to "North" and zero it there?
>

Well, if you know that the test fixture is oriented to some specific geodetic azimuth when the turntable angle reads some specific value, then it seems reasonable (if the turntable is well levelled) to assume that there is merely an addition constant that is needed to convert horizontal turntable readings to geodetic azimuths.

One thing that I think I'd want to do is to verify that an alignment telescope actually is perpendicular to the axis of the vertical turntable. Any mounting for an alignment telescope would presumably be offset from the axis of the horizontal turntable (according to the sketch), so one way to do this that comes to mind to check for non-perpendicularity would be to set up two reference lines by an alignment telescope, but with the horizontal turntable rotated 180 degrees between the two settings, reversing the telescope in the holder, but otherwise keeping everything the same. If the telescope is perpendicular to the axis of the vertical turntable, the two telescope positions should describe parallel lines of sight. Any divergence should be measurable by readings on scales near and far, i.e the difference in scale readings will show non-parallelism and give a means to correct for it.

> We can machine what we'd need here with an in house machine shop, but are there sensors or equipment that a surveyor may have that we would not need to purchase? And then we could somehow mark our systems for future calibration checks?

Yes. I'd turn the determination of geodetic North over to a surveyor and have him or her also examine the rest of the alignment system to see if any obvious improvements or alterations would be in order.