2 Engines Versus 6 Engines

Adam

Well-Known Member
5PLS
I did some testing yesterday with an LS Base sending out corrections via UHF (D8PSK) and the RTCM MSM3 Short message format. The Base is in a good open sky location. The test units were 2 LS's in a deep creek bottom with a lot of Holly trees (thick inpenatrable leaves) and various mature hardwoods without leaves. Also there was a 25 or 30' tall bank about 10' to the South. The test consisted of 2 different action profiles which were the same on both receivers. I will show the results of each test. The tests were selected to test the performance of each rtk firmware package and to see if a 2 engine fix will yeild an incorrect fix.

First test - started at 10:11 am - finished at 1:57 pm
The settings were 2 epochs required and 2 engines fixed. Duration was turned off.


On LS1 the 2 engine multi constellation package was selected. On LS2 the 6 engine multi constellation package was selected. The units were about 3' apart.

LS1 - The LS stored 151 points in a little less than 3 hours. This is a little bit misleading though. Below I will list the periods of time when points were collected.
10:11 Point 1
10:30-10:38 Points 2-48
10:41-10:48 points 49-84
10:57-10:59 points 85-110
11:45-11:49 points 111-115
11:54-11:59 points 116-126
12:12-12:13 points 127-133
1:00-1:01 points 134
1:05-1:06 points 135
1:10-1:11 points 136-137
1:15-1:16 points 138
1:27-1:28 points 139
1:37-1:38 points 140
1:48-1:50 points 141-151

I would count these blocks of collection as individual points. So really I would count this as 14 points. The point count is not important for this test. I basically want to find out if a bad fix can slip thru when 2 engines are fixed with the 2 engine rtk firmware and when 2 engines are fixed with the 6 engine rtk firmware.


NONE OF THESE SHOTS WERE BAD FIXES. 2 ENGINES 2 EPOCHS. They are also pretty tight with one another for the area.
2-REVIEW-COLLECT_20200202-09.25.14.png

I am in no way telling you to willy nilly accept a 2 epoch point. (yet). I want others to post their experience and tell me if you see the 2 engine rtk lie when both engines are fixed. Imagine the time we may be able to save if it proves true. And, I still see a need to use one engine and use verification for your preferred duration. My hope is that at some point we can have the settings configured so you run the normal boundary profile with 1 min rtk engine but when 2 engines fix the shot is over (regardless of time spent there).


LS2 - 6 engine multi rtk...Same settings as above 2 epochs and 2 engines.
134 points stored

10:11-10:12 point 1
10:17-10:31 points 2-30
10:41-10:55 points 31-60
11:07-11:13 points 68-77
11:30-11:31 points 78-79
11:56-11:58 points 80-86
12:37-12:38 points 87-88
12:46-12:49 points 89-93
12:54-1:08 points 94-105
1:14-1:20 points 106-119
1:28-1:30 points 120-121
1:47-1:51 points 125-134

Looks like some bad fixes got thru with the 6 engine multi constellation setup. The correct fix is the cluster at 67.
2-REVIEW-COLLECT_20200202-10.02.51.png


Here we have the correct group of points. We will count 33 in as well. 116 of 134 were correct with spread of about .19'
2-REVIEW-COLLECT_20200202-10.04.22.png



Second Test - started at 2:11 finished at 4:40
The settings for this test was total duration of 240, min duration of 180, min rtk engines 1 and validate with 1


I'm not going to run thru the time each point was selected with this test because it was pretty much a wash except for a little more precision on the 2 engine unit

LS1 - 2 engine multi constellation rtk

11 points All Good and within .09'
2-REVIEW-COLLECT_20200202-10.13.47.png


LS2 - 6 engine multi constellation rtk - 10 points all good and within 0.13'
2-REVIEW-COLLECT_20200202-10.15.25.png
 

Attachments

Last edited:

Nate The Surveyor

Well-Known Member
Can you tell us more about the 2 engine setup?
We know it has 864 channels. Does this mean that half the channels are using GPS, and GLONASS, and half are using bediu, and Galileo, and the rest of them?
Does this mean that there is a divide coming, in the software, for those of us with a T-2, and those with t-3's?
I don't have 13-14k laying under the bed. So, I'll have to wait. But, it'd be nice to kinda know "the plan".
Nate
 

Darren Clemons

Well-Known Member
Wow! Extremely detailed report Adam. Kudos for taking the time and effort to do this for us.

I did a similar test myself last Saturday in an even worse spot than you have pictured.

I sat my two LS's directly up against two 14" Maple trees in a drain.....probably about 60' below my base...

I, however, had my profiles set up very different. I had the two engine firmware to accept 1 engine of course, and to spend a minimum of 60 seconds in phase 1 and a minimum of 240 seconds shooting overall.

I had the 6 engine firmware set to my "standard" boundary profile I have used for years.....

They both ran for about 5 hours and the 2 engine firmware "won" storing 19 shots to 16 shots for the 6 engine firmware.....NEITHER stored a bad shot but with both at a minimum of 240 seconds I wouldn't expect them to.

The only noticeable difference to me was the "tightness" of the collected points. The 6 engine firmware had a 0.36' horizontal spread between the two farthest apart shots. The 2 engine firmware only had a 0.17' difference between the farthest.

Upon looking through all the screenshots, there was a definitive difference in the "spread" (both H & V) of the screen as the shot was being collected to verify these results.

Three weeks in and there's no doubt in my mind the 2 engine firmware is "better" with using all constellations, BUT, it is very different than what we've all gotten used to - mainly having to allow it to use 1 engine epochs. This is appearing to give me a ton of jump/fail episode trying to validate in phase 3. What I've done to try and minimize this is change my confidence to only 1 so after it does a jump/fail the very next epoch that goes back to the "good" group, it will immediately try and validate again....

This 2 engine firmware, in my use so far, seems to give you something in the range of 20% to 30% of the epochs I'm getting with the 6 engine firmware, BUT, they'll almost ALL be "good epochs". The 6 engine firmware is popping and clicking all over the place (similar to what we're used to) and it'll finally find the correct group. The 2 engine firmware starts out with the right group and rarely strays away from it.....
 

Adam

Well-Known Member
5PLS
Can you tell us more about the 2 engine setup?
We know it has 864 channels. Does this mean that half the channels are using GPS, and GLONASS, and half are using bediu, and Galileo, and the rest of them?
Does this mean that there is a divide coming, in the software, for those of us with a T-2, and those with t-3's?
I don't have 13-14k laying under the bed. So, I'll have to wait. But, it'd be nice to kinda know "the plan".
Nate
I used the default engine setup for the 2 engine and 6 engine methods. For the 2 engine, engine 1 is gps glonass galileo and engine 2 is glonass galileo beidue.
I don't know how the channels are utilized in the processing of these engines. I did not use the Auto Setup button during this test because I just left them there all day. Auto Setup seems to help when things get stale.
 

Joe Paulin

Active Member
Nice testing and reporting Adam, thanks! If you have time, please setup your old T2 as a base and go back to the same spot and report your findings using the "old" V6 conventional tracking with just GPS and Glonass. I would be curious to see how the new compares to the old. P.S. it will help sell new receivers if we can see the difference from old to new.

Thanks!
 

Adam

Well-Known Member
5PLS
Nice testing and reporting Adam, thanks! If you have time, please setup your old T2 as a base and go back to the same spot and report your findings using the "old" V6 conventional tracking with just GPS and Glonass. I would be curious to see how the new compares to the old. P.S. it will help sell new receivers if we can see the difference from old to new.

Thanks!
I did quite a few tests of this before the release. With the standard firmware, 1 engine provides many bad fixes, the 2 engine 2 epoch test would be a total fail. For verification, I can say that more points were collected with the 2 and 6 engine multi constellation firmware then the standard firmware and the spreads were tighter.
 

Jim White

Member
I had an opportunity today to set up about 1000 feet off a NYSRTN. I spent a little time to try and determine the performance of the 3 engines (old 6, new 6 and new 2). I am interested not only in the quality of the points, but also the time required to get it.

I set the LS between two trees under leaf-off canopy, which I believe is a difficult location, but typical for trying to locate a point in the intersections of two hedge rows.

On the first try, the old 6x engine moved out of Phase 1 with 5 groups and a confidence of 10 after 450 seconds.

Both the new 6x and 2x engines created more than 6 groups in the same time period, but the best confidence was only a 2.

The fourth and last try was again with the old 6x, and it resolved in less than 30 seconds.

The first and fourth point are acceptably close.

I tried another location, 20 feet away at the edge of the canopy and all 3 engines took less than 30 seconds with 1 group each.

I assume that I have some settings wrong, does anybody have suggestions as to how to get the 2 new engines working as well as the old one?

Thanks!
 

Adam

Well-Known Member
5PLS
What is the rtn using? Galileo and Beidou? I have talked to a couple of folks in Georgia that use egps that have said the performance is bad when they use that network. I think it is using Galileo, not sure about Beidou. I talked to someone yesterday on a network that was using Galileo and GPS but not Glonass. It would use only 2 constellations at a time. I think this is built into the networks. I recommend setting the LS up over night and connect it to RAMS and the network and let someone look into it.
 

Darren Clemons

Well-Known Member
The 2 engine firmware doesn’t seem to be performing as well as it was when it was first released. There was an update a couple weeks ago to both the gnss and Jfield and I updated one of our LS’s and seemed to notice a drop in performance the 1st day so I didn’t update the rest.

Has anyone seen any similar results or have any info on what’s going on? It’s always so difficult to know whether it’s something with the software or just “one of those” spots or times of day where it just is the satellites and/or geometry. The 6 engine “regular” GPS/Glonass has been performing fine and we’ve just went back to using it exclusively.
 

Adam

Well-Known Member
5PLS
It's getting bogged down after a few minutes. The development team knows about it. For now switching firmware to the standard is the fix. Also if you switch to another firmware it reboots the gnns and fixes temporarily. I have been switching g between all three.
 

T.Guisewhite

Active Member
The 2 engine firmware doesn’t seem to be performing as well as it was when it was first released. There was an update a couple weeks ago to both the gnss and Jfield and I updated one of our LS’s and seemed to notice a drop in performance the 1st day so I didn’t update the rest.
Definitely seeing the same thing...
It gets stuck and bogged down. You go to the satellite screen and it says “No Data”.
I am switching to 6 engine again for now. Please let us know when this is attempted to be resolved and I would like to try and test the two engine theories again.
 
Top