Active Braking and Micro Brushless

OK guys … I’d like to propose a discussion of a progressive idea. I’m thinking about turning active braking off!

Here’s why: I’m running 1102’s and compared to larger motors … it seems to be largely ineffective. I almost can’t notice it working. So I don’t think it’s contributing any MAJOR gains. I also notice after hours of reviewing flight footage on YouTube that prop wash oscillations are really hard to completely eliminate on these builds. In my own experimentation - my motors start to warm up with D values around 60 with braking on. We know this is because noise in the copter is causing D term to begin to freak out a bit and active braking on the motor is trying to keep up with every little change producing heat. I don’t feel like I have enough D and I’d like to tune it more like my 8.5 brushed. With braking off, I can tolerate a noisier copter, run higher D gains, reduce more filtering to get P more “in sync” with actual gyro measured motions - all without smoking motors since escs will only be allowed to push them in 1 direction. End result should be I can increase D till I start to experience a “fly away” condition (then back it off a little) and I theorize that I may be able to chase our prop wash more effectively with higher d gains than I can with lower d and active braking. This may only be valid for 1102/1103 if valid at all.

What do you think?

1 Like

With as new as 1102/1103s are (and their respective props), I would imagine that higher quality components would improve on the efficacy of such a dynamic feature as active braking. On top of that, I would expect a high rate of PID loop would be needed to properly manage so many wildly dynamic elements effectively in such a high stress scenario.

As you know, I’m hardly an expert, but when I think of all that’s going on, trying to be managed in a PID loop, I’m thinking we need really quality components to do that well.

This is less about quality and more about size. Small motors can’t do active braking very strongly as a physical limitation. As far as processing speed - I’ve worked wonders on many f1 boards. Configuration and tune becomes more important at slower speeds bit it is possible.

To summarize my theory - forward only application of higher D (no braking) may physically respond faster than active braking can at lower d values.

The real question is, will the old stand alone blheli suite let me turn active braking off for blheli_s? The chrome configurator will not.

2 Likes

Sounds like a great experiment! With micro quads so light, the active braking shouldn’t be a concern.

1 Like

If I can turn it off … still unconfirmed … I imagine the drag of the props should do most of the work anyway with no generation of excess heat.

Confirmed not possible with blheli_s

Slightly more complicated to test now … maybe the tarot 6a blheli esc installed in a new build… ugh

While active breaking certainly isn’t as strong on 1102s as it is on bigger motors, I’m running a bobo 95 with a softmounted Femto FC (8khz/8khz/16000 Multishot) and blheli_s. At some point the machine was almost prop wash free (D ~ 90 iirc, pretty high P too), but that changed after some serious crashes (expect-able). Still running the same settings, but with some prop wash.

Going to try 1106s next, I expect way more torque.

1 Like

I will at least compare the code between blheli and blheli_s and see if I can implement pwm fast in place of damped light. It may well he over my head to make that substitution at a code level … but worth a try before giving up.

1 Like

Hi @NotFastEnuf sorry to bump this this topic but did you find a way to disable/reduce damped light on bheli_s? Or does someone know how to do this?

Hey, to achieve reduced damping light/active braking/increased battery performance you can look into Blheli_S/M or JESC which have hex files for 24/48/96khz PWM settings and increased deadtime.

https://www.rcgroups.com/forums/showthread.php?3621257-BLHeli_M-Maverick-version

https://www.rcgroups.com/forums/showthread.php?3374061-JESC-Surgical-Precision-with-Rpm-Filter-on-BLHeli_S

Hi @yets good to see you are still here.

Found the JazzMaverick code and some interesting posts on RCG.

To make the telemetry work he implemented reversed Dshot and removed the PPM protocol.
And for my mad science project I need the PPM protocol to work.

Yesterday I found some interesting hex files on his GitHub.
From what I can find it was to prove some motor noise problem.

If this is the default blheli_s_16.7 with regenerative breaking disabled and PPM still working this could be the solution for my project.

Also he mentioned in one of his posts that it takes only one line to disable the regenerative breaking in the original firmware. That could also be a possible solution.
But Not having the SiLabs tool chain installed I hope that the first solution will work.

Do you know how it will impact the noise, voltage spikes and back EMF if I disable regenerative breaking? Positive or negative?

Is regenerative breaking, Dampened light and complementary PWM the same thing?

2 Likes

Nice video… just build a PPM to Dshot converter to test the 48/96khz PWM.
But I don’t think it does what I need… I want it freewheeling.
Maybe play with higher dead time to see if it helps in combination with the higher PWM frequency.

I found that in some blheli_32 ESC’s there is a non damped mode option.
But for that price I can buy or build a VESC.
Going to far off topic now… thanks for your input… let’s find out how far I can push my blheli_s ESC :laughing:

If you check the Blheli-S thread on RCgroups JazzMac released some custom hex files that disabled damping completely. https://www.rcgroups.com/forums/showpost.php?p=44349829&postcount=6371 The last time I checked it was on his GitHub

I think JazzMac is refering to the GitHub link of JazzMaverick I mentioned earlier.
If you ask me it’s the same person :smile:

Yup, it’s the same person. I thought that you were after a hex without damped light so that your motors could essentially freewheel, that’s why I posted a link

Yes that’s what I need.
See my 2nd post… the GitHub link is to: https://github.com/JazzMaverick/BLHeli/tree/JazzMaverick-patch-1/BLHeli_S%20SiLabs/Hex_files_16.9/Beta/Hex_files_16.7_NoDamping

1 Like