Digging up old problems, Relative vs Absolute APC heights

Are you referring to absolute vs relative antenna calibrations to create the antex or ini files. And not antenna heights. ?ÿYour using the term heights itƒ??s throwing me off a bit with antenna relative vs absolute. I could be way off but it seems you are speaking of two of the three or so approaches that obtain a calibration to compute APC position. ?ÿ

I am speaking of Trimble using relative calibration for some receivers and absolute calibration for other receivers in the same ini file and refusing to admit they have a problem.

The APC calibration is the height of the phase center above the antenna reference point. You are welcome to call it something else if you wish. ?????ÿ

I am speaking of Trimble using relative calibration for some receivers and absolute calibration for other receivers in the same ini file and refusing to admit they have a problem.

Unless the operator is forcing TBC to use the Trimble relative values, it pulls directly from either the IGS or NGS absolute antex files:

Those ATX files come directly from the source (IGS or NGS).

I've been processing data with mixed receivers in TBC for nearly 15 years and have never had a problem using the IGS values.

?ÿ

Antenna calibration is not just the height is what i was referring to. Thats where i was getting confused a bit. It is directional in all bands l1 l2 etc and such. Since we cannot physically measure to the APC itself. There are pros and cons and many arguments on which on is best. ?ÿNGS does an conversion from relative antenna calibration to absolute to meet igs standards. They are working to be able to perform an absolute antenna calibration they have the robot arm and possibly will be implementing it this year as a milestone. This has been in development for a while. I was unaware Trimble was using different ones. Of course i have not been in the need to know until recently. Many people have made the assumptions that ngs was doing an absolute antenna calibration now but per their site it is a conversion as of now from relative to absolute. For igs xx compliance. ?ÿEach manufacturer does their own plus most send a sample of each type usually 3 to NGS and to others for them to compute. Ngs does relative.?ÿ

Along these lines, why does the APC value change depending on whether or not you use an extension between the tribrach and antenna?

@lurker the apc value is from a point of reference the arp. In your controller or office software you are giving a value based on bottom of arp or quick connect lever slant height or true vertical height based on a physical point we can measure much like a rod height. ?ÿThe apc is a point that is where the signal comes in there is a technical word for this. ?ÿThat antenna calibration is done to solve the nrp which is north ref point and of course a height. If you open a antenna file example you can on ngs website faq. Will be better at answering then me on my phone walking the farm. Anyway you will see values for north east and up. Doesnƒ??t matter if its absolute vs relative. Itƒ??s not just a height reference. ?ÿAn antenna that uses all constellation vs just gps will have different values because all signals and bands donƒ??t come in at same reference point. Point not being a physical spot. As of the latest ngs release of procedures document 2019 I believe. For absolute values its igs even a web site link to get them. If you wanted your own personal calibration done via absolute values. If people here want i can post a link this evening when i get back home. ?ÿNow if it something different than the antenna calibration files needed then i am off on the wrong tangent here and apologize. But the op is using height and antenna calibration interchangeable based on how i am reading it. I could be wrong but i can document the antenna calibration process and procedures. ?ÿThere are actually more than relative and absolute antenna calibration. They are the ones that get the most attention though.

@olemanriver I'll try to illustrate my question. Why does the APC change when a 10cm extension is used between the tribrach and the gnss antenna?

2 topcon antennae exactly the same

?ÿ1 with out extension and one with

?ÿand different values for the APC calibration.

?ÿThis indicates to me it matters what is below your antenna and will affect the data you collect.

Antenna calibration is not just the height is what i was referring to. Thats where i was getting confused a bit. It is directional in all bands l1 l2 etc and such. Since we cannot physically measure to the APC itself. There are pros and cons and many arguments on which on is best. ?ÿNGS does an conversion from relative antenna calibration to absolute to meet igs standards. They are working to be able to perform an absolute antenna calibration they have the robot arm and possibly will be implementing it this year as a milestone. This has been in development for a while. I was unaware Trimble was using different ones. Of course i have not been in the need to know until recently. Many people have made the assumptions that ngs was doing an absolute antenna calibration now but per their site it is a conversion as of now from relative to absolute. For igs xx compliance. ?ÿEach manufacturer does their own plus most send a sample of each type usually 3 to NGS and to others for them to compute. Ngs does relative.?ÿ

If you go to this website to look at the Antenna.ini file version 8.77:

Trimble Antenna.Ini Viewer (trimbletools.com)

and look up R10 you will find a vertical offset (m) 0.1491

Zephyr Geodetic Model 2 ROHS (57971) is 0.08546

Zephyr Geodetic Model 3 (115000) is 0.06519

Of those three, only the Model 3 (115000) is anywhere close to the IGS14 value. The other two are the typical 2cm off according to my spreadsheet above.

I don't have or use TBC but it is my understanding that the data collector uses the trimble antenna.ini file for RTK.

James

?ÿ

?ÿ

This indicates to me it matters what is below your antenna and will affect the data you collect.

Those differences are mostly sub-millimeter.?ÿ To me that indicates that it's system noise rather than antenna characteristics that are responsible for the differences.

@jim-frame The differences in the "up" values are 5.90mm and 2.30mm for each frequency modeled. I'm speculating the NGS doesn't consider that noise. I'm also speculating they didn't publish two different calibrations for the same antenna because the system noise was different each time. If in fact the system noise cannot be distinguished in the values they are publishing, the values shouldn't be published at all.

The differences in the "up" values are 5.90mm and 2.30mm for each frequency modeled.

Sorry, I was referring at the underlined numbers.?ÿ I agree that 6 mm in the up is significant (though I wouldn't call it substantial), and it may reflect -- as you indicated -- the difference in mounting hardware.

I note that I've conducted thousands of 1-hour sessions (GPS-only) on 7-10 km baselines and gotten as much as 4 cm relative difference in vertical between sessions, using the same groundplane geodetic antennas.?ÿ The system does contain a fair bit of noise.

@jim-frame I certainly don't understand all this stuff but I'm curious about these calibrations. From what I can tell, if you are using the same antennae the calibrations don't matter in your results because they are the same. The problem comes in when you include different models of antennae. If the calibrations don't relate accurately across antennae models then your results will be incorrect. It troubles me that what material sits beneath your antennae has an effect. I believe I saw a white paper discusssing this but I can't seem to find it again.

Okay I believe some are joining the term antenna offset and antenna calibration files with what is not 100% technically correct. The antenna offset from diagrams are the measurements from the APC (Which is the antenna reception point) to the physical locations that we can all measure. NGS refers to a specific location in which they call this ARP Antenna reference point. They also use measurements provided by the manufactures or measure themselves like from center of bumper, bottom of a notch an additional lever if requested.

All I ever said was that the antenna calibration is more than a vertical offset. Sure we have the template 9antenna calibration files From manufacture, NGS, GEO++ AKA kinda IGS, we have absolute antenna calibrations we have relative Lets just leave it at that because other methods are not discussed here)

From NGS:

1) What is an "antenna calibration"?

An antenna calibration is the act of determining the point of reception of the Global Navigation Satellite System (GNSS) carrier phase signals. Antenna hardware such as the antenna elements and pre-amplifiers create signal phase advance and delay before passing the signal to the receiver. The phase advance/delay changes the range measurement, and will introduce error to position solutions. The point of signal reception of an antenna is not a physically measurable location, and varies depending upon the direction of the satellite signal being received. Therefore antenna calibrations create a map of phase advance and delay which depends on the satellite position in an antenna-centric frame (based on a specific?ÿARP?ÿand?ÿNRP).

2) Why do I need to use an antenna calibration?

To obtain accurate coordinates at a point, the signal phase advance and delays (calibration values) that occur at the antenna must be applied when processing collected GNSS data. Non or misapplication of antenna calibration values can introduce errors at the centimeter level. Furthermore, because the calibrations are specific to a reference frame, processing GNSS data with the correct values is an essential step to accurately define and access positions consistent with the National Spatial Reference System (NSRS).

10) What is a relative calibration?

In a relative calibration, all antenna offsets (PCO) and phase center variations (PCV) are computed with respect to a reference antenna which is normally assigned zero PCV values. For?ÿNGS relative calibrations, the reference antenna is the Dorne Margolin choke ring antenna, type T (AOAD/M_T NONE). A relative calibration is therefore biased by the phase advance/delay experienced by the reference antenna. A file with NGS relative calibrations conducted to date is available on request, please contact ngs.antcal @ noaa.gov.

11) What is an absolute calibration?

To conduct an?ÿabsolute calibration, the antenna being tested is moved via a robot so that a particular satellite is received at different angles by the test and reference antennas. This angular separation enables cancellation of the reference antenna effects, leaving behind only the antenna offsets (PCO) and phase center variations (PCV) of the test antenna.

12) What is the source of the NGS absolute antenna calibration values?

NGS sources calibration values from a combination of IGS absolute calibrations and NGS relative calibrations. NGS starts with the most recent published IGS ANTEX file and incorporates all the antennas listed in this file. For additional antennas with NGS relative calibrations that are not in the IGS ANTEX, NGS has converted these additional antennas using the absolute antenna calibration values for the reference antenna AOAD/M_T NONE to create an "absolute from relative" absolute calibration. NGS is in the process of completing development of an absolute antenna calibration technique; this is being developed by the Geosciences Research Division and conducted by the Instrument and Methodologies Branch of the Geodetic Services Division at NGSƒ??s Testing and Training Center in Corbin, VA.

?ÿ

From me: ARP NRP APC Antenna Reference point North Reference Point These two are physical locations something we can measure it is exactly what they say a reference point.?ÿ APC?ÿ Antenna Phase center it is the reception point.?ÿ

From NGS This is a format example:

<ant_info.003> <TYP:REL SRC:NGS-Database>                     <CBL-22/12/14=732>

ANTENNA ID        DESCRIPTION                  DATA SOURCE (# OF TESTS) YR/MO/DY
                                                             |AVE = # in average
 [north]  [ east]  [  up ]                                   | L1 Offset (mm)
 [90]  [85]  [80]  [75]  [70]  [65]  [60]  [55]  [50]  [45]  | L1 Phase at
 [40]  [35]  [30]  [25]  [20]  [15]  [10]  [ 5]  [ 0]        | Elevation (mm)
 [north]  [ east]  [  up ]                                   | L2 Offset (mm)
 [90]  [85]  [80]  [75]  [70]  [65]  [60]  [55]  [50]  [45]  | L2 Phase at
 [40]  [35]  [30]  [25]  [20]  [15]  [10]  [ 5]  [ 0]        | Elevation (mm)

The APC can change in two identical antennas on with a ground plane one without. One with a RADOM one without. Because it is a reception point. Even two identical antennas lets change that to 3 the reason is NGS requires at minimum of 3 when doing these antenna calibrations (As of Now only Relative) No one antenna even the same on in the same environment are exactly the same. some are closer than others. I know a particular brand that all 3 or 4 that were sent identical all fell under a dew mm of each other. I imagine they have all become better maybe this was years ago. Out of 10 that were tested not by NGS 3 were chosen because they were so close to each other.

?ÿ

Yes it matters what's under an antenna. unfortunately we cannot at this time account for every variable it is not cost effective for anyone. And even if we could just like a brand new set of tires all at the same psi will fluctuate sitting in your driveway a little based on sun temperatures slope of the ground you are parked on or that whisky bottle setting on one side versus the other. The OP asked about absolute vs relative heights and also used the term calibration. two different things in vocabulary and I am the worst on grammar I know.?ÿ Do not confuse the offsets with antenna calibrations the antenna calibration numbers will change based on how it was done absolute vs relative. What all signals it tracks GPS only Glonass and GPS or full on GNSS. The ini files aka antenna calibration files different formats all will be different based on what and how it was done. the APC can be changed by so many things but it is not only in the height. The height changes based on what you are measuring too. Lets speak Trimble since he asked specifically about Trimble. We have a 2m fixed rover rod. the height is 2m to what ARP (NGS REFERENCE POINT) bottom of antenna (COULD BE EQUEL TO ARP)?ÿ or bottom of quick release or did we physically measure to the bumper on say an R8 which is a different animal because now it is a slant height not true height measure up. Those are all physical locations that the antenna calibration file and software can use to get you to and from the APC (AGAIN RECEPTION POINT) it is the locations model that all signals are coming into and we need to get from that to our point on the ground.

?ÿ

Notice the header in this one format example above. North East Up

?ÿ

I hope this helps. explain what I was trying to figure out and such. I will call a Trimble person next week and find out if Trimble fixed whatever issue (the OP question on some receivers using Absolute and some using relative) if he will be working. But all of this truly absolute vs relative is not enough of an issue really especially in RTK and most Land surveying projects because Land Surveyors are more about precision than accuracy or maybe better way to say it is relative accuracy of everything on a site. It could be an issue if the base and rover where wrong by a lot I guess but it would all still be relative. Now if you are doing high end very accurate datum work then we need to get in the weeds more so we understand terms like PCO PCV and things that can affect them. But we will not be using a R10 or R12 for that type of work. LOL..

?ÿ

?ÿ

?ÿ

@rover83 No issues for me either its all about the reference frame which one should be used. I have used TGO GPS Survey a little with TBC. I have processed geodetic networks with every kind of manufacture antenna astech trimble leica and more. Choosing the correct reference frame is more important to match the antenna calibration files to as they are datum specific but that gets into the weeds of accuracy vs precision very easily and quickly. For RTK relative surveying would be hard pressed to see it making a huge difference. It can based on reference frame but then were are talking cm or so.

For background on antenna calibrations at the US NGS see: https://geodesy.noaa.gov/web/science_edu/presentations_library/files/antcal_NGSwebinar_2021.pptx

They do have a robotic arm to perform absolute calibrations. It is shown in the webinar slides.

The GFZ has an visualization tool to display absolute calibration data here: http://wox.geopp.de/gnpcv2pdf/index.html

Another good resource is: http://www.epncb.oma.be/_newseventslinks/workshops/EPNLACWS_2019/pdf/2.01_gpp_cal125_euref19_p.pdf

@lurker I will try and get you the correct terms but near narrow far bands in RF fields all cause issues with GNSS signals. There are studies showing different materials fabrics that cut down on noise for CORS stations around the world on different buildings and such. Nest is a term academia has used for fabric to lay down around a roof where a station is mounted. I was in this world for a while and because I am just an old country boy I asked them all once if my beard was sufficient material to use to stop some noise interference LOL. So my large beard at the time was nickname the GPSnested beard multipath noise reduction filter. everyday surveying i would not lose sleep over this if we truly discussed every thing that affected or could affect GNSS signals we all would never use GNSS for anything. Build good field and office procedures and let those with the PHD's do that chicken scratch. Manufactures can make mistakes but the all truly have some very intelligent people on the design of these things and science behind it. if you want to dive in deep go to gps.gov and start reading all the PNT papers all the technical standards tech note 36 etc etc, ocx information and specifically where they address the manufactures so they can build the equipment and software to be compatible with GPS. Once you get that chewed up and digested I will give you $100 if you can find one just one manufacture that meets every spec. And we are only discussing GPS standards. I left that world it made my brain hurt lol. I cannot lie I enjoyed it for a while and really learned a lot and was involved with a lot and seeing the benefits of things that effect the surveying community today was all fun. But nothing beats the smell of fresh cow pies in the morning and looking at or finding an old corner.

@geeoddmike The robotic arm is awesome. anyone in the area should go by and see it for sure it is a piece of work. I was around when they installed it. On other side of the building is where they do the relative antenna calibrations. I cannot wait until they start doing the absolute ones very possible it will happen this year but who knows. Dr. Andria and Mader are some very smart people they have working on this absolute thing for some time.

and look up R10 you will find a vertical offset (m) 0.1491

That's the same offset as for the R10-2 and the R12, both of which had processor/power unit changes, but left the antenna and much of the other internals the same. I was working Trimble tech support and training for a dealer at the time the R10 and R10-2 were released, and saw no tech support issues during that time which would indicate that the R10 had an incorrect vertical offset versus R8 series or the R10-2.

The R12i has a 0.1291 offset because room had to be made for the IMU.

?ÿ

In any case, as @olemanriver has correctly pointed out, the vertical offset in the antenna.ini file is NOT to the L1/L2 phase center, but to the physical antenna phase center, which is what is used for APC-APC computations and reductions in Access:

The offset from the ARP to the mechanical APC is not the same as that to the L1 or L2 PCs.

@olemanriver?ÿ

I had thought from the linked article below that there would have been faster progress.This was presented at a 2010 ION meeting.

https://geodesy.noaa.gov/pub/abilich/papers/ABilich_ION2010official.pdf