Shot differential

We are currently using RS3 base and rover using NTRIP. The equipment is fairly new to us but we completely understand the Emlid workflow. Here are my issues. I can establish a control point by averaging a fix point using NTRIP and set that point as my base marker. I am using Lora to my rover and when I take a shot on a check nail (different from base point) I can take the shot for 10seconds and every time I get a different Northing Easting and Elevation ( more than 0.2 differential) without touching or moving my rover.Today I imported known points from a surveyor. I set my base over the point and chose the correct point that my base was on top of as my base marker and same result. Bottom line I get a different result every time I take a shot. I need my drone data at a tenth or less.

What would cause the results of each shot to be that much different without moving or touching my rover?

I was able to compare against a Trimble rover using our 3rd party VRS today and we were spot on. I just don’t understand what we are doing wrong to literally set the rover over a check nail take a shot store the point not touch the rover at all take another shot on the same point and get such a difference…

2 Likes

Have you tried the Emlid on your VRS? Rule out your base and configuration.

2 Likes

What do you mean exactly?

If I manually enter our base marker of a known point or establish a point using eGPS (VRS) when I check into our check nail I get a completely different shot each time I shoot that check nail…

2 Likes

Your post is a little confusing as to what solutions you are using for each the Trimble and Emlid setups. You say NTRIP to set the base, got that, but you then say broadcasting corrections via LoRa but you don’t mention and base station or whether or not you are using radio for both models or NTRIP on one and a local base on the other. Do you have a site calibration in the Trimble setup? If so, did you localize the Emlid to it or are you just running with whatever the NTRIP/local base is telling you? When you start switching between brands and corrections solutions you are going to get deviations unless you make sure all datums are configured exactly the same.

On the fact of the Emlid shooting inconsistently have you been watching for multipath and the magnetic storms? What kind of environment are you in? The Emlid has many less channels and does not handle multipath and canopy near as well as the Trimble and there have been a lot of high Kp-index days over the last couple of weeks. We were unable to converge our drone with NTRIP yesterday and that has never happened before.

Have you submitted a full system report to Emlid support yet? That would be the first step.

2 Likes

I just used the NTRIP to an average shot using Trimble and then signed into the same provider using Emlid. After two or three shots I did get the same result as the Trimble. I set that shot as our base marker and then moved to a check nail and just for verification I shot the same check nail twice and got completely different results so that was confusing so as I continued to shoot that same point I would get a different result each time… I sent a email to Emlid yesterday and they replied stating they were on vacation and would not be available until 10/17…

I am using a 2m rod. Emlid automatically sets the antenna adjustments correct?

2 Likes


This morning.

1 Like

Yes, the app should auto-rec what model of receiver and apply the appropriate offset. So you shot points with the Emlid until you received on that matched what you got with the Trimble and then you dropped NTRIP and setup an Emlid base and corrected via LoRa?

What method did you use to acquire the point with the Trimble from NTRIP, average fix? Or an instant shot? You backed off at least 10m and then staked it out to verify it right?

How much was the Emlid floating? Was this also backing off and reoccupying? Or just sitting in the same place?

There must be an understanding of the differences in the equipment. One costs $15K and the other $3K. One has 600-700 channels and the other has 184 so it is going to be a little slower and less resilient in challenging environments. It kills me to see people shooting points with Trimble for 1-3 seconds. Just because it is a little bit better receiver doesn’t mean that its 5Hz is any different than any other brand’s 5Hz. They are relying on 5-15 samples… I never shoot less than 10 seconds with our Topcon and 30 seconds with the Emlid on anything than requires +/- 0.10ft. If it’s a control point I shoot it for 2 minutes.

4 Likes

In addition to @michaelL input also say that tilt compensation dont like static shots in RTK mode.

0.2, is that feet or metric meters?
Anyways. TC might corrupt static shots if you had it enabled

6 Likes

Do what @TB_RTK recommends and turn OFF tilt on the ROVER to see what you get. Do what @michaelL says and sample a little longer.

Also turn OFF tilt on the BASE. Was it off anyways?

Your Antenna Height appears correct… but not sure if rounding is that critical of 7.001’ vs 7.002’, but may if tilt is enabled?

Does the Trimble have tilt and was that enabled or not?

4 Likes

Thank you to all above! I agree the tilt compensation should be turned off. With more testing today I believe that shooting the points longer and having NTRIP turned off to our base after the base maker is established definitely helped . I was able to establish 2 points and check back into each with .005 difference. I was taking 10 second shots before and upped to 1:30 shots with tilt turned off on both.

2 Likes

.05

1 Like

Tilt is for faster convenience where lower tolerance is acceptable.

Was tilt enabled on the base when you had these issues?

Sounds like turning off tilt no matter what solved your issue?

Once you establish your FIXED known point via NTRIP CORS at the start, you can turn it off anyways on the BASE, then use LoRa and ROVER.

1 Like

I never had tilt turned on the base. I only turned tilt on the rover for two or three shots to see what the difference is (if any)

Turning off NTRIP to our base and using LoRa to the rover I get a difference of +/- .05 now. Again that is occupying a point taking a shot not moving the rover take another shot on the same point. Prior to doing that I would get 1-3 tenth difference. Longer measurements as well.

3 Likes

Awesome! I’ve used 5 different brands of GNSS gear and they all have their quirks. Glad you got it working!

2 Likes

Thank you yall for jumping in and helping us on this!

2 Likes

I had a similar problem a few months ago. i have a ntrip correction service -30km away- survey a square. turn off rs3. turned on, stake out the same points and all of them were off by 0.15 m. turned off and on the rover a couple of times, staking and measuring the same points with the same results, off by 0.15m. chatting with a colleague i mention this issue and he told me: all your measuremes are good, as long as you are in the same GNSS session, if you want to stake out some points the next day or resume with a survey you will be off about 0.2 m more or less. Therefore the importance of control points to “calibrate” or “localice” each GNSS session to your working site.
This is not a problem with Emlid recievers, all brands have this issue when using a ntrip service from far way.
Idk, may this has to be with almanacs or ephemerides that are constantlly in change, i know that are people really smart on this forum so please, correct me if im wrong.

3 Likes

Not with Javad. Part of our procedures doing boundary or any kind of control work, we always verify the previous days control points. Using either RTN or local RTK, our checks are usually within 2-3 cm.

I’m seeing the same with Emlid’s machines if all measurements procedures are observed. Possibly an error in user methods ? As professional land surveyors, we are supposed to be experts in the measurement processes.

2 Likes

Thanks for this info.
I’m going to start adapting 2 mins for control too.
Already been employing 30 secs for 0.1’ accuracy.

3 Likes