Dual motor single battery throttle problem

IvdPeet · October 7, 2021, 11:07am

Hello,

I have a setup with a fixed-wing dual motor drone which I want to power using a single Lipo battery. When I throttle up, however, above around 60% the motors start stuttering and don’t spin smoothly anymore. A video of the problem can be found here.

The setup I am using consists of:

Believer twin motor airframe
2x PROPDRIVE v2 3536 910KV Brushless Outrunner Motor link
2x Turnigy MultiStar BLheli_32 ARM 41A Race Spec ESC 2~5S (OPTO) link
Turnigy High Capacity 16000mAh 4S 12C Lipo Pack w/XT90 link
Carbon Fiber Propeller 12x6 Black (CW / CCW) (2 stuks) link
Pixhawk PX4

It is wired up as follows:
Battery → Parallel three-way splitter (1,2,3) → (1) ESC left wing, (2) ESC right wing, (3) Power to pixhawk rail & other internal components.

A picture of the wiring can be seen here:

I have tried numerous configurations, and measured the current through the wires using a current clamp. The configurations are listed here:

Configuration 1 → Problem occurs at 60%:

One battery powering everything
Both motors attached
Both propellers attached

Configuration 2 → Problem does not occur:

One battery powering everything
Both motors attached
No propellers attached

Configuration 3 → Problem does not occur:

One battery powering everything
Only one motor attached
Propeller attached

Configuration 4 → Problem does not occur:

Three batteries powering everything separately (one for each ESC + one for the Pixhawk)
Both motors attached
Both propellers attached

Configuration 5 → Problem occurs at 60%:

Two batteries in parallel powering everything through the three-way parallel splitter
Both motors attached
Both propellers attached

While you could say that using separate batteries for each ESC (Like in Configuration 4) is a solution to this problem, it is undesirable for our application since it will weigh more, and take up more space. Configuration 4 will also introduce a possible problem when the batteries are not charged to the same voltage.

Experiment data:
Measurement 1: Use Configuration 1, measure current drawn from battery:

throttle 50% → 11.8A
engine cutoff at throttle 62% → 20.5A
while stuttering at 62% fluctuation between 10A and 15A
throttle 100% → fluctuation between 14A and 18A

Measurement 2: Use Configuration 1, measure current going to left wing ESC:

throttle 50% → 6.26A
engine cutoff at throttle 62% → 9.8A
while stuttering at 62% fluctuation between 3A and 8A

Measurement 3: Use Configuration 3, measure current drawn from battery:

throttle 50% → 6.2A
throttle 62% → 11.1A
throttle 100% → 35A

Measurement 1 and 2 seem to convey the same data, where if the current going to only one of the two ESCs is measured is about half of the current that is being drawn from the battery in the same configuration.

I am truly at a loss what could be the cause of this problem. I have debunked the following hypotheses:

Battery cannot supply enough current. Not true because firstly, the battery should theoretically be able to supply much more current. Secondly, attaching two batteries in parallel still yields the problem at the same throttle value (See configuration 5). Thirdly, it is measured that the battery can supply at least 35A (See measurement 3). The problem occurs around 20A, so this should not be the problem.
Interference to the signal wire. Don’t think this is true because the problem does not occur when only one ESC is attached. If would be weird if this configuration does not result in interference, but configuration 1 does…

I would love to hear from you what you think the problem might be!

count74 · October 7, 2021, 2:14pm

Just a wild guess, but do you have another three way splitter or another power distribution method? Is there anything fancy build into the 3 way splitter? Diodes, anti-spark circuit etc.?

IvdPeet · October 7, 2021, 3:36pm

The splitter is a diy part. It’s a simple parallel connection of XT90 connectors with nothing special inside.

I used to have a crappier version of the splitter which I thought could be the cause. That’s why I made this one, but the problem remains.

Rolf · October 7, 2021, 5:50pm

Hi Ivo,
I have just seen which BLHELI ESCs you are using. I can only warn against using these Multistar 41 and 51 BLHELI_32 ESCs for more than just disposal. Several people including me had several times total failures with these ESCs. eg: DShot and BLHeli ESC Telemetry support - #43 by Naterater

Regards Rolf

Naterater · October 7, 2021, 8:34pm

Do you have capacitors on the ESC’s? We use at least 300uF caps at each ESC. Normally it’s a 1000uF cap.

IvdPeet · October 8, 2021, 1:41pm

Thanks for your concern Rolf,

I will keep this in mind. But do you think this could be a cause for the problem that I’m having?

IvdPeet · October 8, 2021, 1:44pm

I do not have any capacitors on the ESCs. Would this be able to resolve the issue? I’m not very experienced with Electrical Engineering myself, but could it be harmful in any way if I attempt this as a solution?

hendjosh · October 8, 2021, 5:39pm

No. They are not be harmful unless you add way way too much capacitance. If you look at most higher end ESCs you will find 1000 -1200 uF total capacitance.

It is all about the voltage ripple and spikes! The purpose is to lower voltage ripple and ringing caused by long wires and motors being large inductors.

This post from RC Groups was quite good and means I don’t have to write it myself.
https://www.rcgroups.com/forums/showthread.php?3240349-Adding-Capacitors-to-ESC-for-Longer-Wire-Length

Quick Summary for capacitor selection

Electrolytic (aka check orientation before you install)
IMPORTANT low ESR type
A few 2-3 smaller capacitors are better than 1 giant capacitor.
Give yourself at least 30% voltage margin for the capacitor rating
Pay the few cents extra and get high temperature rated capacitors at least 105C
Pay extra for a good brand Panasonic, Rubicon, Chemicon

IvdPeet · October 9, 2021, 8:16pm

Thanks Josh,

From what I also read in your linked post is that long wires could also be the cause of my problem.

I think will try to shorten my wires, and additionally add capacitors if necessary. Will keep you posted on the results.

Thanks for your help everyone!

hendjosh · October 9, 2021, 8:29pm

Yes, short wires are important specifically from the ESC to battery. I still might add at least one 680uF capacitor given your application on that ESC.

But remember there are other considerations too like electrical noise and interference caused by ESC onto your radio systems. (If it gets near your wireless systems.) It can be beneficial to keep the two separated.

Naterater · October 14, 2021, 12:48am

I’m 99% sure you will solve this issue by adding a large capacitor directly to the ESC power input pads. I’ve seen the same behavior when operating BLHeli_32 ESC’s at higher voltages.

IvdPeet · November 30, 2021, 1:23pm

Hey everyone, thanks so much for your input!

I replaced the wires from the splitter to the wings with a thicker gauge, and shorter leads. After I did this, I tested again, and the throttle problem occurred after about 85% instead of 60%, which was already an improvement.

After that, I added three 470 uF capacitors in parallel to both ESCs. This COMPLETELY SOLVED THE PROBLEM! I used these with a 470 uF capacitance.

Below a picture of the capacitors near the ESC:

And a picture of the new thicker and shorter wires:

Thanks again for your help guys!