Thread Tools
Oct 20, 2020, 12:05 AM
Registered User
Thread OP
Quote:
Originally Posted by Bill Glover
A very short demo of the r/c seaplane 'rescue boat' I mentioned earlier, now using a pair of BLHeli_32 ESCs running in bidirectional soft mode with braking at 100%. I only have rampup power at the default 50% but that's good enough for this application ... reversing/accelerating 10" props more aggressively would surely put a lot of load on the motor mounts!

https://www.youtube.com/watch?v=0H01mZcQepo
Cool - thanks for sharing!
Sign up now
to remove ads between posts
Oct 20, 2020, 04:01 AM
Registered User
Bill Glover's Avatar
I do find it amazing that these relatively tiny ESCs heat up so little when sealed inside an airtight box. Admittedly I'm only logging the external temperature of each (I have a FrSky 100A current sensor on board, which happens to include two temperature probes), but there's clearly no issue:
Latest blog entry: Eachine QX65 FPV quad review
Oct 20, 2020, 04:34 AM
4c 6f 76 65 20 52 2f 43
Risto's Avatar
That is not true for my case. Using Holybro Tekko32 F3 metal B with Hacker A30-14L and 12x6 prop, I can see significant temperature increase, especially for throttle above 70%. With full power, I can see more than 50 degrees celsius increase of ESC temperature inside 30 seconds (from 20 degrees C to 70+ degrees C, current 35A, 4S LiPo).

Here's a real-time screenshot video of BLHeliTV software, where I throttle up slowly in 5% increments from 0 to 100% (9500 rpm). Especially from 6000 rpm upwards you can see an exponential temperature rise:
Tekko32 F3 metal B with Hacker A30-14L V4 motor, 12x6 prop and 4S LiPo characteristic (1 min 31 sec)


I decided to add a small heat sink with thermal double sided tape and stick it out of fuselage for better cooling on my foam Opterra 2m:
Last edited by Risto; Oct 20, 2020 at 05:05 AM.
Oct 25, 2020, 06:01 AM
Registered User
Quote:
Originally Posted by ib3jeff
Hey guys, maybe you can help...Im currently running

PX4 3DR Flight Controller with v4.0.3
Tekko32 35A 4 in 1 esc with V32.7
BlHeliSuite32715
MissionPlanner 1.3.73
on a windows64 based machine (Beside Windows, everything is the latest release)

I talk to the AP through Radio and talk to the ESC via the USB port on the AP

So Ive followed the dshot/passthrough tutorial posted by Ardu found here:

https://ardupilot.org/copter/docs/co...-passthru.html

With my current params, I can see live telemetry via the OSD and LOGS in MP. The issue I have is when I interface with the ESC in B LHeliSuite. I can connect to the esc, read it, flash it, configure it, read verbose messages, etc.... however when i go into the "Motors" Tab (timestamp 12.37 in the walkthrough video) and attempt to use any of the sliders the GUI closes the ESC connection. Basically I check that my props are removed, It reduces the number of motors to four ( the correct amount) and states the "Controls Enabled"... When i grab any of the slides, they only let me increase signal from 1000 to 1010, runs for about 1 second and then closes the connection in the GUI, as if i never checked off the Props removed box. I have no clue if this is an ESC error, a param I missed, a GUI error, etc.... any guidance if greatly appriciated.
Quote:
Originally Posted by 4712
Done: https://github.com/ArduPilot/ardupilot/pull/15597
Hope this is the right way for that pr...
Quote:
Originally Posted by yaapu
Cool, I asked Tridge to review your PR, thanks

Inviato dal mio SM-T825 utilizzando Tapatalk
Please check out BLHeliSuite32 32.7.1.8 or BLHeliSuite32 x 1.0.0.8.
It should fix the issue for recent and upcoming Ardupilot firmware.

Quote:
Changes in 32.7.1.8:
General:
  • Added 32.7 hex file for:
    • CLRACING_STSPINA_V2
    • Furling32_4in1_C
    • TMOTOR_L4_01
  • Added workaround in “Motor” tab for Ardupilot (FC_VERSION < 3.3.1) missing response on MSP_SET_MOTOR command.
  • Added warning message in “Motor” tab, if motor outputs are disabled.
  • Added the ability to edit and flash BLHeli_32 rev. 32.73 beta.
  • Further minor fixes andimprovements.
Oct 25, 2020, 10:47 AM
Registered User
Thread OP

Win-win with a bonus


A new test code is now published on github: https://github.com/bitdump/BLHeli/tr...ncy%20testcode
This folder contains test code that supports variable pwm frequency. You can configure motor pwm frequency to increase with throttle. The benefits of variable pwm frequency are:

- Low frequency for low throttle gives good active braking where it is most needed
- High frequency for higher throttle makes running smoother
- The variable frequency will work as dithering making running even more smooth

So this is a win for low throttle and a win for high throttle, with the effect of dithering as a bonus.

This test code also has many fixes, some of them are:
- Signal detection is now much more simultaneous between ESCs
- Telemetry no longer increases rpm info error rate for Dshot600

Note that for the latest functionality you will also need the latest version of BLHeliSuite32.
Oct 25, 2020, 03:08 PM
Registered User
nice, benefits of both worlds. thx
Oct 26, 2020, 10:18 AM
Registered User
Quote:
Originally Posted by 4712
Please check out BLHeliSuite32 32.7.1.8 or BLHeliSuite32 x 1.0.0.8.
It should fix the issue for recent and upcoming Ardupilot firmware.
Thanks I'll test it asap, will have test hardware soon.
Oct 26, 2020, 02:08 PM
Paul
ZzyzxTek's Avatar
Does the variable PWM frequency vary it linearly from low throttle/low range value, to high throttle/high range value? Or is some sort of curve/formula applied?

Thanks!


Quote:
Originally Posted by sskaug
A new test code is now published on github: https://github.com/bitdump/BLHeli/tr...ncy%20testcode
This folder contains test code that supports variable pwm frequency. You can configure motor pwm frequency to increase with throttle. The benefits of variable pwm frequency are:

- Low frequency for low throttle gives good active braking where it is most needed
- High frequency for higher throttle makes running smoother
- The variable frequency will work as dithering making running even more smooth

So this is a win for low throttle and a win for high throttle, with the effect of dithering as a bonus.

This test code also has many fixes, some of them are:
- Signal detection is now much more simultaneous between ESCs
- Telemetry no longer increases rpm info error rate for Dshot600

Note that for the latest functionality you will also need the latest version of BLHeliSuite32.
Oct 26, 2020, 02:24 PM
Registered User
Thread OP
Quote:
Originally Posted by ZzyzxTek
Does the variable PWM frequency vary it linearly from low throttle/low range value, to high throttle/high range value? Or is some sort of curve/formula applied?

Thanks!
Pwm frequency is varied linearly with throttle. So any nonlinear thrust compensation will have to be done in the FC.
But the range of pwm frequency is programmable (independent low and high frequency setting).
Oct 26, 2020, 03:51 PM
Registered User
tested this afternoon, must say flys really nice on my 5 inch quad. better propwash control at low throttle and smoothness at high. this is really a nice feature added.
Oct 27, 2020, 04:24 PM
Registered User
why is sine mode not enabled at default, as it makes your motors running more smooth? are there any downsides?
Oct 28, 2020, 01:58 AM
Registered User
Thread OP
Quote:
Originally Posted by jlandegent
why is sine mode not enabled at default, as it makes your motors running more smooth? are there any downsides?
Sine mode is only suited below the very high erpm range for two reasons.
First of all, sine mode requires that you have a number of pwm periods per commutation period, as it uses the pwm period to modulate the sine.
Secondly modulating the sine requires a significant amount of processing power, so the max erpm will be lower for sine mode.

For a 48MHz MCU doing 24kHz pwm, already at 50k erpm there is only about 5 pwm periods per commutation period.
While the MCU will probably not give up until more like 200k erpm (depending upon input signal rate).

Of course sine mode can make more sense for a higher speed MCU (e.g. F3 running at 108MHz) and higher pwm frequency (e.g. 96kHz).
But anyway, the effects/benefits from sine mode are quite subtle.

I'd think that the effects from the improved braking for low pwm frequency at low throttle is much more noticable for most systems,
Oct 28, 2020, 02:06 AM
Registered User
Is there any plans for a FOC esc/firmware?

Inviato dal mio HTC U12+ utilizzando Tapatalk
Oct 28, 2020, 02:21 AM
Registered User
Thread OP
Quote:
Originally Posted by riktor
Is there any plans for a FOC esc/firmware?

Inviato dal mio HTC U12+ utilizzando Tapatalk
FOC is a whole different beast.
Again, as for sine mode, max erpm is limited by pwm frequency. But high pwm frequency is possible, so that's not the real limitation.
There are other limitations...
FOC needs to estimate rotor position based upon motor currents (block commutation uses BEMF voltage), and estimation using currents is much more complex.
There are many brushless systems using FOC, but are there any that are capable of handling the very crisp acceleration of BLHeli?
Estimating rotor position in a system that runs at a stable rpm may is not that hard, but doing so accuratly during rapid throttle changes is very challenging.
Then there is the fact that FOC needs accurate info about motor parameters (resistance, inductance I believe and maybe more). So ESC parameters will have to be matched to your motor. Which maybe can be done by an automated learning process, but still...

The primary benefit of FOC is improved efficiency. Apart from that I think potential benefits are small. So it is very well suited for endurance systems, where the prolonged endurance can outweigh the increased complexity and cost.
Oct 28, 2020, 02:40 AM
Registered User
[QUOTE=riktor;45751415]Is there any plans for a FOC esc/firmware?

There is already a very good Open Source FOC Project.

But as sskaug said: It's a complete different thing.

Look at the VESC Project and the VESC Tool to get a short view.

Here is a FOC ESC in BLHeli Size based on VESC Project and more or less open source:
https://teamtriforceuk.com/a50s/


Quick Reply
Message:

Thread Tools