Hot shoe and RTK/PPK work flow

Reach Module / M+ / M2 Hardware integration
1

Hi,

I have a drone set up with the reach RS2 as base. The R2 rover is set up with RTK corrections and the autopilot triggers the camera based on a flight plan. A hot shoe is connected back to the autopilot (cube) so that GPS raw coordinates can be logged precisely. The hot shoe needs to be connected to the reach GNNS directly in order to follow the suggested emlid-PKK workflow. My questions:

  • Is it advised to have the hot shoe connected to both the autopilot and the reach? I prefer also logging to the autopilot. I can possibly modify the hot shoe cable.
  • Can one accomplish similar accuracy using the RAW GPS that is logged from the autopilot, if so, does anybody have a suggested workflow for PPS where the autopilot records the CAM-events?

Any help appreciated, thanks!

2

Electrically, yes, recommended, nop.

It cannot, unless it has a PPK capable GNSS module inside.

3

Thanks for you quick reply, Christian

Can you elaborate on this? Are you suggesting this is not a good idea? One could add diodes downstream the hot-shoe split. I was thinking just to solder the signal and gnd wires and add to the reach unit with no additions.

Thanks!

5

Hi @atob,

Let the Reach do all time marking job. It’s designed for that.

Reach M2 records time marks from the camera to the raw data log. Recorded time marks don’t have any coordinates in the log, only time. Coordinates can be acquired during post-processing only.

The event marking with the modified cable will be unstable. You can lose time marks in the raw data with the described connection. So the camera should be connected straight to the Reach M2 without any additional connections.

3 Likes
7

2 posts were split to a new topic: PixHawk 2 with Reach M2

10

Hi, thanks for the response. I have purchased an emlid hotshoe cable and have received it now. Will start testing this workflow shortly.

11

Hi @atob,

Good news! Don’t forget to share the results of your tests :slight_smile:

1 Like
13

Hi Andrew,
I have been testing with the hot shoe adapter. So I disabled the previous setup (events stored in the raw logs of the flight controller). So fully running the emlid-work flow.

It is working, however, it seems to miss events (?). In particular, it has missed one single event for several of my tests. I am running emlid studio beta 10, with the Drone data PPK feature in kinetic mode. I tested a run twice, with 144 images taken, and there are only 143 events. Quite frustrating.

Is there something I have missed in setting it up, work flow or hardware related? The old hot-shoe adapter always registred the correct number of events.
thanks!

15

Hi @atob,

Can you share the dataset with me by sending it to support@emlid.com? I’d like to check it by myself. We have a 20 Mb limit. You can send the link to the cloud folder.

Raw data logs from both units and photos are required.

I have some other questions down below:

  • What is the firmware version on the Reach M2?
  • Do you process logs in the Emlid Studio or another software?
16

Hi,
Thanks for the reply.
Yes, I did process using Emlid Studio. Ills send what I have in a separate email.
Regards

17

OK, sent over now, thanks!

1 Like
18

Did you ever get an answer as to why it was missing one event? We are about to start on a massive project and we can’t afford to be losing events.

Thanks

19

Hi Francis,

I should have posted an update here earlier :slight_smile:

We’re currently investigating the cause of lost time marks with @atob. Once we have the results, I’ll post the conclusion here.

Right now, I can recommend checking that your units are updated to the latest firmware version and set up accordingly to our guide about mapping with the RTK modules and antenna placement guide in the docs.

If any questions arise, don’t hesitate to contact me.

20

Hi @fpelletier,

Just a minor update on this topic.

@atob reported that lowering the update rate from 10 Hz to 5 Hz helped to get all time marks in the data. The unit currently has V27.0 on board.

However, we don’t have the results of the tests with the latest firmware version — we improved the time marking feature in the V27.1.

I’ll return to this thread if we have an update from @atob.

22

This topic was automatically closed 100 days after the last reply. New replies are no longer allowed.