Admin

Team CDC Slash Sparkfun 2014 AVC results

Posted by Thomas J Coyle III on July 2, 2014 at 3:39pm

Hi All,

I was only able to complete two of the three heats and took 7th place with a score of 696.

If I would have been able to complete the 2nd heat, for the same number of points that I received in the 3rd heat, which I won, I would have had 1042 points for an easy 2nd place finish.

When I pushed the Start/Stop switch to enable the throttle on my rover at the start of the 2nd heat, the Pixhawk went to disarmed due to a power brownout.

At first I thought that it was due to a marginal battery that had excessive voltage sag when the Pixhawk throttle output was applied to the ESC to get the motor going.

However analysis of the Pixhawk tlog at the moment the Start/Stop switch was pushed revealed the following:

A review of the Pixhawk CH2 throttle output at the time the Start/Stop switch was pressed indicates that the Pixhawk was not ramping the throttle to the Cruise_Throttle value, as it does when using the R/C control to go from the Manual Mode to the Auto Mode, but was jumping the throttle right to the Cruise_Throttle value (~1700us).

This jump in the throttle PWM output put a step function demand on the battery which caused the battery voltage to sag to the point of causing the Pixhawk to brownout and disarm resulting in the no-run 2nd heat performance of my Slash rover.

Had the throttle been ramped, as when using the R/C control to go from the Manual to Auto Mode, there would have been very minimal battery voltage sag experienced during the 2nd heat and my Slash rover would have made the full course run and would have come in 2nd in the Peloton Class.

The code is ramping the throttle, but by the time the Start/Stop switch has been pushed, it has completed the throttle ramping and is driving the throttle output with the full Cruise_Speed PWM value (~1700us).

What I think is going on is that when the Pixhawk is placed in the Auto Mode and it is Armed, it proceeds to ramp the throttle output to the Cruise_Throttle PWM output value (~1700us), but there is no throttle output on CH2 until the Start/Stop switch is pushed. 
When the Start/Stop switch is pushed the Pixhawk applies the full Cruise Throttle PWM value (1700us) to the ESC on CH2 without the ramping experienced in the R/C control Manual to Auto Mode change.
 
The step function jump to 1700us puts an extremely large load on the battery which results in voltage sag and a subsequent brownout and disarming of the Pixhawk. Had the throttle ramped to the Cruise_Throttle PWM value there would have been no step function load on the battery and no brownout.
 
I have used the NMHi batteries in my Slash during testing on my Home Depot test course and have never experienced a Pixhawk brownout and disarming due to battery voltage sag even when the battery voltage was down to 8.6vdc. This was because the R/C control Manual to Auto Mode change code ramps the throttle output unlike the Start/Stop switch code step function jump in the throttle output.
The present way to correct this issue, if you are competing where a Start/Stop function is required, is to use a high C (30 - 40) 2S LiPo battery which will not have the voltage sag experienced by the NMHi battery, when subjected to a step function throttle input ESC current demand.
 
Regards,
TCIII ArduRover2 Developer

You need to be a member of diydrones to add comments!

Join diydrones

Email me when people reply –
Follow

Replies

  • T3
    Thorsten July 3, 2014 at 1:06am

    Hi Thomas,

    congratulations (at least for the third heat)!

    I already thought about this problem when playing with the Start/stop switch. Since I'm using it frequently during a mission and not only once at the beginning the step function results in some permanent and unnecessary stress for the system.

    My suggestion/feature request is to implement a more smooth (linear?) fade in of throttle instead of the step function. As a developer do you think this is possible/easy to implement? 

    Kind regards,

    Thorsten

    • Admin
      Thomas J Coyle III > Thorsten July 3, 2014 at 8:17am

      @Thorsten,

      Thanks for the support:-) You win some and you lose some:-)

      The step function in the throttle output, when using the Start/Stop switch, was unintended and the code just needs to be reworked to make it function like the R/C radio start version of the Auto Mode.

      Regards,

      TCIII ArduRover2 Developer

  • T3
    healthyfatboy July 2, 2014 at 7:13pm

    I think you would be quite surprised how much these road brushless systems can pull. There's a reason Traxxas rates their brushless ESC's at 320A. Realistically, you won't get that much on a typical system but that initial punch really is high.

  • Kirby G July 2, 2014 at 4:34pm
    Im using a 5000mah turnigy 40c 2s lipo and get brownouts at high throttle starts on my brushless slash when the battery isn't fully charged. I got around this by powering the pixhawk on a smaller 1700mah 2s lipo.
    • Admin
      Thomas J Coyle III > Kirby G July 2, 2014 at 6:01pm

      Hi Kirby,

      Interesting that you are getting the brownouts even with a high C LiPo.

      How high a Cruise_Throttle value are you using that is causing the brownouts?

      I have never used a Cruise_Throttle value over 45% and have never had a brownout using my 25C NMHi 7 cell packs when using the R/C transmitter to change the mode from Manual to Auto.

      Though I do suspect that you could cause a brownout with Cruise_Throttle values in the 50 to 100% range:-)

      My rover would always start out smoothly and ramp up to the Cruise_Speed value without a brownout.

      I have graphed the target speed and the gps speed and the target speed always had a medium ramp to it when shifting to the Auto Mode when I kept the Cruise_Throttle value at 45% or less.

      Regards,

      TCIII ArduRover2 Developer

This reply was deleted.