PX4 Sapog, Zubax Orel, Zubax GNSS 2 status update

This post announces recent noteworthy developments in some of our projects.

The observations we made since the announcement of Zubax Orel 20 (a Sapog-driven ESC) led us to believe that people tend to have serious misconceptions about power capabilities of ESC. A possible reason for that is the abundance of Chinese low-quality ESC that tend to have much lower actual power capability than advertised. Here at Zubax we termed it power capability inflation. In order to prevent our customers from having incorrect understanding of the true power capabilities of Zubax Orel 20 (pictured on the right), we strongly suggest to model and verify your application using ECalc: http://www.ecalc.ch/xcoptercalc.php.

As was promised a couple months back, the reference hardware design for PX4 Sapog has been released under CC BY SA 3.0. The design files (in Eagle format) are available in the official PX4 Hardware repository on GitHub. We're encouraging third parties to develop their own hardware ESC designs based on this reference. Feel free to stop by our brand new Gitter channel if you have questions.

Besides the above, Sapog had another development: we've released a number of important improvements, mostly concerning stability of the spinup algorithms. The release binaries can be downloaded from files.zubax.com and installed via UAVCAN or DroneCode Probe.

Zubax GNSS 2 has received a firmware upgrade too. A few users reported that they're having difficulties calibrating the compass with PX4 or APM. An investigation uncovered that the difficulties were caused by 2 factors:

1. The measurement range was not sufficiently wide to account for all valid use cases. This was fixed in the firmware.

2. APM and PX4 somewhat mishandle external compasses by making invalid assumptions about their measurement ranges. We added a configurable scaling coefficient that allows to reduce the magnitude of the output magnetic field vector, silencing the warning from the autopilot. Learn more from the docs at docs.zubax.com.

The fix is available in firmware version 3.1, you can get it from files.zubax.com. If you're not using the compass, there's no need to upgrade.

Useful links:

Pavel.

Views: 1576

Comment by Rob_Lefebvre on October 31, 2016 at 11:16am

How does eCalc help to understand the claimed over-statement of the power handling of ESC's?

Comment by Zubax Robotics on October 31, 2016 at 3:36pm

Hi Rob,

It doesn't help with that directly. Instead, it allows you to evaluate the required power capability of ESC for your application, providing you with relatively accurate numbers. We believe that it is likely that the real power requirements of your application will turn out to be lower than what you expected.

Comment by Brandon Robinson on October 31, 2016 at 3:39pm

I was under the impression eCalc does empirical testing to determine the numbers they use rather than simply using the listed specs from a supplier.

Comment by Johnatan on October 31, 2016 at 5:25pm

We need an 80amp ESC not a 20amp!

Comment by Rob_Lefebvre on October 31, 2016 at 6:16pm

I need 160 Amp.

eCalc is not terribly accurate, and misses quite a lot of issues.  It is only one tool in the engineering toolbox. 

Comment by Olivier on October 31, 2016 at 11:07pm

Just replace with an octo and divide by 8.

Kidding!

Comment by Rob_Lefebvre on November 1, 2016 at 4:54am

Sorry, need it to fly properly. :)

Comment by Olivier on November 1, 2016 at 6:48pm

Yeah yeah ...

Ok, nice! ;)

Comment by Thomas Butler on November 3, 2016 at 3:30pm

How about posting a PDF of the schematic somewhere? Thanks!

Comment by Guy McCaldin on November 3, 2016 at 11:39pm

@Thomas, seems the schematic and board files (Eagle format) are available from the PX4 github page:

https://github.com/PX4/Hardware/tree/master/sapog_reference_hardware

Comment

You need to be a member of DIY Drones to add comments!

Join DIY Drones

© 2017   Created by Chris Anderson.   Powered by

Badges  |  Report an Issue  |  Terms of Service