patiam Posted at 7-26 12:13
LOL I love it... Now we're using ChatGPT to tell us how to do things .

The point is, if the solution you obtained from OPUS was already for the ARP, of course you don't want to add the antenna height AGAIN when entering the coordinates into the D-RTK 2.

Maybe the question was asked incorrectly in ChatGPT.

The answer in itself is not wrong if the same antenna is used, with the same pole, in the same place, without having informed the offset at the time of the capture of the OBS.

That's a lot of parameters that ChatGPT obviously doesn't know.

I find it frightening the number of colleagues and students who use this thing all day long. The worst is for the code. Nobody really knows how it works except Ctrl+V, Ctrl+C, Run.

patiam Posted at 7-26 12:13
LOL I love it... Now we're using ChatGPT to tell us how to do things .

The point is, if the solution you obtained from OPUS was already for the ARP, of course you don't want to add the antenna height AGAIN when entering the coordinates into the D-RTK 2.

Hi again and thanks' for the answers.

Sorry if my question wasn't properly contextualized.

No, by no means am I using ChatGPT to tell you how to do things, I was just trying to find some help to my specific problem. I did not realize you were talking of a known benchmark with known altitude.

My case is different, I put the DRTK2 station on a new (unknown) ground mark, let it there for 4 hours, then took the DAT files, converted to RINEX and sent to a post processing GNSS service (PPP, by ibge.gov.br). The altitude measured by the DRTK2 and sent to PPP was probably at the antenna top level (?). Then, as I additionally had to input the antenna height* at the PPP, I assume that the service gave me back the corrected altitude net of the antenna height, in other words, at the ground level.

Then, if now I have the corrected altitude (HAE) at ground level, I suppose I have to sum the antenna height to input Z at the DRTK2 for surveying, so that the drone will refer to the top of the antenna. Does it make any sense?

* A note on antenna height: because this PPP service does not recognize the DJI DRTK2 as a valid antenna, do I have to sum (measured ARP height + 141.9 mm) to find the actual "antenna height"?

Tks!

LV_Forestry Posted at 7-26 12:49
Maybe the question was asked incorrectly in ChatGPT.

The answer in itself is not wrong if the same antenna is used, with the same pole, in the same place, without having informed the offset at the time of the capture of the OBS.

Hi, just to let you know that I rarely use "this thing"   

In this case, lack of technical content and support from DJI led me to try ChatGPT. But clearly I had to be more specific in my question.

@RDC Agritech- (it looks like your avatar is the kanji for "river", is that right?)-

For all the reasons mentioned by @LV_Forestry and myself, as well as others, you are WAY better off posing well-crafted questions to experts in the field, rather than relying on an AI resource. Also as mentioned, both context/specificity and interpretation can be especially fraught when using such a resource. I get that it is convenient to ask a bot a question, but that convenience comes with considerable risk...

Regarding the topic at hand-  I tried to address this in my initial reply:
When (and where) to apply offsets, including (but not limited to) the Z offset between the benchmark on the ground and the antenna reference point (ARP) at the top of the pole/tripod is a question with multiple possible answers, but the rule governing all of them is that you account for all offsets ONCE, and ONLY ONCE.

If your workflow involves generating a PPP solution for your base station coordinates to be used later for subsequent drone flights, and the PPP solution is for the point on the ground over which the base station is located (the "benchmark"), then the Z offset will need to be applied to the PPP-generated coordinates when setting up the base station over that benchmark for those later flights.

If, however, the PPP solution DOES NOT represent the benchmark, but rather the ARP (in other words the pole & APC offests were never accounted for duting PPP), then the Z offset DOES NOT need to be accounted for when setting up the base station over that benchmark for those later flights ASSUMING THE EXACT SAME SETUP (antenna, pole height, etc.) is used.

Again, while this is all pretty twitchy stuff, this particular part of it is pretty basic: offsets MUST BE accounted for, but ONLY ONCE.

I hope that makes sense.

patiam Posted at 7-26 21:24
@RDC Agritech- (it looks like your avatar is the kanji for "river", is that right?)-

For all the reasons mentioned by @LV_Forestry and myself, as well as others, you are WAY better off posing well-crafted questions to experts in the field, rather than relying on an AI resource. Also as mentioned, both context/specificity and interpretation can be especially fraught when using such a resource. I get that it is convenient to ask a bot a question, but that convenience comes with considerable risk...

@patiam, yes! RDC = Rio da Cachoeira in portuguese means "waterfall river", or 滝川 in japanese. It's a river nearby my farm, in Atibaia, a town with strong japanese influence, 90 Km from São Paulo.

So again thank you for your patience. The answer in the end is pretty obvious: account once for the antenna, do not double account. My first approach to the problem was a bit erratic. As a physicist I shouldn't make those basic mistakes

What was not so obvious (although now it's clearer) is that the starting point of the reference system for the drone should be the top of the RTK antenna. And that the antenna top point Z means (from my undesrtanding so far) <<HAE + measured ARP height + 0.1419 m>>.

[]s

Nice. I have strong Japanese influence as well

Nearly correct; when entering coordinates into the D-RTK 2 interface the Z value needs to be for the APC. We are telling the software the XYZ of the APC.

If the Z value you have already is for the APC, you just use it (assuming the exact same setup is used).

If the Z value you have is for the benchmark on the ground, you add the [pole height (from ground to antenna base) + 0.1419 m]:

APC Z = (benchmark Z + pole height + 0.1419 m)

Thanks for the info in this thread, was very useful reading.

I measured my D-RTK2 with a tape measure, not exactly hardcore science but the totals seem to check out.

Base Pole = 1.24m / 1240mm
D-RTK2 Extension Pole = 0.42m / 420mm
D-RTK2 Phase Center = 0.1419m / 141.9mm

Base Pole + Extension Pole + Phase Centre = 1.8019m / 1801.9mm

I using the D-RTK2 in a known position, and to make it work I use a Sparkfun Express RTK (solid bit of kit, thanks for suggesting LV_Forestry) and a TOP106 antennae to get the sub-inch reading.

With the base pole sitting in the tripod secured and touching the baseplate, I put the TOP106 on top of it and take a reading, then switch over to the D-RTK2 antennae from the TOP106, I then adjust the reading from the Sparkfun before giving it to the software by adding in the difference between the PC's of the TOP106 and the D-RTK2, which I've recorded as 0.509m / 509mm, this brings the TOP106 PC from 1.293m to 1.8019m to match the D-RTK2's PC.

I want to remove human factors a little more, so I will need to make a new extension pole that can couple with the D-RTK2 base pole, and elevate the TOP106's PC to the exact same height as the D-RTK2 like-for-like millimetre-wise, then there would be no adjustments needed.

Can someone sanity check that for me?

The TOP106's Phase Centre is referenced here: https://docs.sparkfun.com/SparkF ... ication/#field-trip

Sounds like a lot of fiddling, but should work.

You realize you could just ditch the D-RTK 2 and use the SparkFun as an NTRIP server for your aircraft, right?

patiam Posted at 12-8 22:04
Sounds like a lot of fiddling, but should work.

You realize you could just ditch the D-RTK 2 and use the SparkFun as an NTRIP server for your aircraft, right?

I could ditch it, but it cost a bit so I'll keep it in use as part of my PPK workflow.

I can live with doing the reading on the Sparkfun for now.

DJI seemed to nerf the D-RTK2 at a software level before I bought it, so that it cannot talk to an NTRIP server and GNSS satellites to get centimetre level readings.

From what I can read it was possible with the Phantom 4 and the Pilot app at the time. If its capable hardware-wise but doesn't because of a little bit of coding, which won't come because they have moved on already, that's really annoying.

I see. I understand the desire to get something out of kit that you paid $$ for. You could do PPK with the SparkFun as well ;)

No, the D-RTK 2 won't do NTRIP for itself.

But you should be able to do NTRIP for whatever aircraft you are using without the D-RTK 2 involved at all. You set up the SparkFun on a known location, get a fix, and then use it an an NTRIP caster (set up a local WAN with a hotspot or portable router). Then you use "Custom Network RTK" (NTRIP) on your aircraft for the RTK correction source. Of course if you want to PPK, you don't need any base at all as long as there is a good CORS station close enough. But if you need locally logged data for PPK, then I understand the need to set up a local base.

BTW, now that I re-read your methods, I realized that it's not clear where the SparkFun is getting its corrections from? Remember, no GNSS (including the D-RTK 2 and SparkFun) can get cm accuracy without either some form or corrections (RTCM), or being told exactly where it is a priori. That is why you need to give the D-RTK 2 a precise position to begin with. Usng the SparkFun instead for a moment and then using its position to adjust that of the D-RTK 2 will only work if you're providing it corrections via NTRIP or something. Is that what you're doing? Otherwise your'e just using a second "single" mode fix to change the first one.