KDE Direct XF UAS ESC and Flight Controller Compatibility List

KDE Direct

KDE Direct, LLC.
Hey Guys,

With all the range of equipment in the market, making all systems compatible is near impossible as a plug-and-play offering between ESCs/Brushless Motors and Flight Controllers.

The KDE Direct XF ESCs are high-quality devices and as such, institute true bi-directional OPTO-isolation for the three-wire RX or Flight Controller control lead. Why did we add this circuitry? The OPTO-isolation provides true isolation of the ESC electronics to the flight controller (and video equipment), preventing any passage of RF Noise or Voltage-Ripples, that can cause all kinds of issues with the overall electronics. One of the key features for the smoothness and accuracy of the KDE Direct XF ESCs is the electrical isolation of the circuits, allowing the ESC and Flight Controller to remain electrically "clean" from each other and not transmit electrical noise. Other high-end systems employ the same methodology in commercial/industrial applications.

For the DJI equipment, there are no compatibility issues and the two systems work in great harmony. DJI takes care of the proper throttle range and radio calibration in the software (done during calibration of the stick endpoints), so the XF ESCs are simple plug-and-play devices.

For the X-Aircraft equipment
, no compatibility issues - only requirement is for the radio to be properly programmed to the controller. The KDE Direct XF ESCs are pre-calibrated at our factory to the Futaba standard of 1100us to 1900us, centered around a 1520us pulse width. Why do we factory-calibrate the ESCs? This allows for the ESC to have a truly linear-throttle curve from arming (1100us) all the way up to full-power (1900us), and ALL ESCs remain consistent out of the box. One of the problems with calibrating ESCs to various radios is the loss of this control and consistency between ESCs, and this can negatively affect flight performance when looking for the best setup for multi-rotors. It is important to calibrate your radio (TX) throttle end-points (for radios other than Futaba/JR standards) to this range for the best operation, and the SuperX is then an excellent match to the XF ESCs and used worldwide.

For the 3D Robotics Pixhawk equipment
, there is a hardware incompatibility of the Pixhawk to true, high-quality OPTO-isolated ESC circuitry. For OPTO-isolated ESC circuitry, the positive/negative control loop of the system must occur at each ESC port of the flight-controller, so that the circuitry is properly powered and kept separate from the full system. The Pixhawk hardware does not provide this, and rather, completes this loop through the ESC itself, grounding all ESCs to the central ground of the Power-Module. While this will work, it can introduce significant noise to the full system, as all ESCs are tied together to a central ground. For ESCs that are NOT OPTO-isolated, this will work and the ESCs will arm, but RF and electrical noise will pass through the system and can lead to instability issues or video noise.

To solve this, customers have provided great insight to the hardware and the current solution (special thanks to Rob, Freddie, and Quoc for their insight), which is to introduce a grounding-loop to the system. This allows the OPTO-isolated ESC circuitry to properly function, and the overall system will work excellent for flight. This is created by adding a negative and positive loop from the Power-Module to the Pixhawk, using one of the open ports of the "Main Out" section. In this regard, adding a jumper from the Power-Module direct to the "Main Out" section will provide the negative-ground loop - then all ESCs will properly arm. Please see the attached image to help explain the proper jumper arrangement.

Same as the X-Aircraft equipment, the radio and software needs to be calibrated to the 1100us to 1900us range for proper operation. Make sure, in the 3D Robotics Mission Planner radio calibration, that your transmitter Throttle PPM range is programmed to 1100-1900. Until Pixhawk is armed, it will output <900us to the ESCs, and the ESCs will not arm until the onboard arming button is pushed. When the button is pushed, Pixhawk will output just below 0% throttle, allowing the XF ESCS to arm and report the cell count. Finally, when armed from the radio (TX), the controller jumps up to the selectable idle speed and the system is ready for flight.

We hope this all helps to clarify compatibility with various systems and we will continue to release information for others as the technology progresses.

Thanks,
Patrick Koegler
Owner, KDE Direct
 

Attachments

  • Pixhawk Hardware Correction.jpg
    Pixhawk Hardware Correction.jpg
    67.9 KB · Views: 1,127
Last edited by a moderator:

So if I understand correctly for those of us running 6S systems with an AttoPilot battery monitor and BEC powering the Pixhawk we’d run the positive/negative control loop to the BEC?

Thanks for all the help! It's refreshing to see a company investing in great customer service.
 

KDE Direct

KDE Direct, LLC.
So if I understand correctly for those of us running 6S systems with an AttoPilot battery monitor and BEC powering the Pixhawk we’d run the positive/negative control loop to the BEC?

Thanks for all the help! It's refreshing to see a company investing in great customer service.


Yes, you are correct and you'll be running the loop off of the BEC to complete the circuits.
 

glider

Member
As I'm reading this, my wiring is incorrect and should splice into the PM and "loop" it through the power rail? Why does it need to go to the Main Out? Suppose you have an 8 motor copter? I was under the impression any of the remaining 'Main Out' or 'Aux Out' is ok. Here is my wiring. I need to get this right as this weekend will be the maiden voyage for my new Y6. Thanks.

2dNMG5Z.png
 

R_Lefebvre

Arducopter Developer
Hello Patrick:

"special thanks to Rob", is that me?

I'm really curious about this. I had wondered if the problem was to do with poor ground connectivity in the system, as I've seen quite a lot of noise on the ground line when using the Pixhawk with servos. But Lorenz, who designed the Pixhawk, says that the ground is a single plane on the Pixhawk. So I don't know why this would help at all? Maybe the problem is not so much that an additional line to the PWM ports is required, but rather that the ground wires from the Pixhawk to the PM are simply too small, and additional wire gauge helps by reducing the return path resistance?

I've actually been using Pixhawk with Opto-isolated Helicopter ESC's with no problems for over a year.
 

R_Lefebvre

Arducopter Developer
Also, in your diagram, you are showing a dedicated 2S Lipo for avionics. Is this something you're recommending? Just to eliminate motor noise?

Note that, while this is generally good practice and I do it on my larger helis, you do have to be careful. The regulator in the Power Module has a 1.1V dropout voltage, and it is set to output 5.3V to account for the diode drop on the APM port diode. So the minimum input level is 6.4V. Below that, things get sketchy. A 2S LiPo can be drawn down to 6.0V, or worse. If using a 2S lipo for dedicated avionics supply, you must never let it go down below 6.5V. This would be good practice anyway, but it's extra important in this case. It would be better to use a small 3S LiPo for avionics supply to solve this problem.

On helicopters, I do not use the PM for Pixhawk supply. I have engineered my own regulator which allows use of 2S Lipo for the Pixhawk and the HV servos. The regulator I use has a dropout voltage of only 0.1V, so it will be good down to 5.1V. This is more critical for helicopters and airplanes, as the servos can draw the battery down below 6.5V easily.
 

min0nim

Member
Hi guys, any tests with the HoverflyPro?

I have the boards, but haven't pulled apart my current working bird to test myself (currently running with DJI).

When I originally put it together I couldn't get the ESC's and Hoverfly to play, but I didn't try changing the throttle end points.
 

glider

Member
Hello Patrick:

"special thanks to Rob", is that me?

I'm really curious about this. I had wondered if the problem was to do with poor ground connectivity in the system, as I've seen quite a lot of noise on the ground line when using the Pixhawk with servos. But Lorenz, who designed the Pixhawk, says that the ground is a single plane on the Pixhawk. So I don't know why this would help at all? Maybe the problem is not so much that an additional line to the PWM ports is required, but rather that the ground wires from the Pixhawk to the PM are simply too small, and additional wire gauge helps by reducing the return path resistance?

I've actually been using Pixhawk with Opto-isolated Helicopter ESC's with no problems for over a year.

Rob,
Should I worry about my setup using OPTO ESC's? I've posted on DIY and ******** with no reply. As this is my first multicopter using OPTO's, I'd like to prevent any electrical issues.

Why wouldn't the UBEC accomplish the same thing as it is grounded to the battery?

To the OP, what if the copter is using all 8 connections on the 'Main Out' section for motors?
 
Last edited by a moderator:

R_Lefebvre

Arducopter Developer
Glider, I would think that simply using a UBEC would accomplish exactly the same thing. This is how my helis are done, and I have no problem with them.

But I do not understand what's happening with the KDE ESC's. I thought the symptom was that they kinda sorta worked with Pixhawk, but didn't always arm, or maybe didn't have a consistent throttle response. Is that correct?

But I have discussed Patrick's comments here with 3DR, it seems 3DR believe the problem is actually because the Pixhawk does not supply power out to the servo rail. So Patrick's solution is not so much about improving the ground, it's actually that he is supplying power to the servo rail. Now, that makes sense, because if his ESC's are true Opto, they NEED a 5V supply on the power rail to function at all. But that's where it doesn't make sense. If they kinda sorta work with no power, that would seem to indicate they are not true opto.

My Opto heli ESC's, if I connect them to a battery, and connect ground and signal to the Pixhawk, but without a 5V supply, they are dead as a doorknob. Not only do they not arm, they don't even beep. The processors for the ESC's are all on the isolated side, and they are unpowered until you provide them with 5V. So what is the story with KDE ESCs?

Unfortunately Patrick doesn't answer my questions, so I can't figure out what's going on.
 

KDE Direct

KDE Direct, LLC.
Hey Guys,

This is all a great discussion and I just received an answer from 3DRobotics in terms of the hardware and the issue at hand in arming true OPTO-isolated ESCs. Rob, you are correct and for a true OPTO-isolated ESC, such as the KDE Direct XF UAS series (many competitors state OPTO, but are truly not in the hardware...), there has to be a power-loop to run the opto-circuitry. There has to be a completed positive/ground loop, since that's the purpose of the OPTO-isolation, to prevent this loop for occurring past the ESC and other electronics (and hence, introducing noise to the full system).

3DRobotics did the analysis and turns out, the +5V positive lead is what is missing on their "Servo rail", or the connection lines for the ESC. In this regard, there is no power supplied to the ESC to arm the OPTO-isolation circuitry and as such, the ESC will not arm (it cannot receive a control signal). So, in order to supply power to the Pixhawk board, the option is to supply the power from an external source (such as an external BEC), or from the Power Module or additional port of the Pixhawk itself. I believe they'll look to correct this in future product-lines or revisions of the Pixhawk, but for now the power has to be brought in with a jumper or external source.

3DRobotics updated the diagram to help explain this, as well as their Wiki to help with setup: http://copter.ardupilot.com/wiki/co...l_note_about_KDE_and_other_Opto_Isolated_ESCs.

I've asked 3DRobotics if they can start including the proper jumper to supply power for OPTO-isolated controllers and/or offer this for sale to make it easier for customers. There are true benefits in preventing noise to the electronics by using OPTO-isolated ESCs, so it's a good thing to have their hardware properly support it.
 

Attachments

  • Pixhawk.jpg
    Pixhawk.jpg
    57.5 KB · Views: 480
Last edited by a moderator:

jrlederer

Member
Glider, I would think that simply using a UBEC would accomplish exactly the same thing. This is how my helis are done, and I have no problem with them.

But I do not understand what's happening with the KDE ESC's. I thought the symptom was that they kinda sorta worked with Pixhawk, but didn't always arm, or maybe didn't have a consistent throttle response. Is that correct?

But I have discussed Patrick's comments here with 3DR, it seems 3DR believe the problem is actually because the Pixhawk does not supply power out to the servo rail. So Patrick's solution is not so much about improving the ground, it's actually that he is supplying power to the servo rail. Now, that makes sense, because if his ESC's are true Opto, they NEED a 5V supply on the power rail to function at all. But that's where it doesn't make sense. If they kinda sorta work with no power, that would seem to indicate they are not true opto.

My Opto heli ESC's, if I connect them to a battery, and connect ground and signal to the Pixhawk, but without a 5V supply, they are dead as a doorknob. Not only do they not arm, they don't even beep. The processors for the ESC's are all on the isolated side, and they are unpowered until you provide them with 5V. So what is the story with KDE ESCs?

Unfortunately Patrick doesn't answer my questions, so I can't figure out what's going on.

Rob,
As usual, I'm very impressed with your response. Don't know how else to phrase it. You are the man!

jonathan
 

R_Lefebvre

Arducopter Developer
Thanks. I try to be accurate.

So Patrick's response confirms a few things. Still odd that we saw a few reports of the KDE ESC's kinda sorta working. There must be another piece to the puzzle still.

Patrick, are your ground planes separated? Is there any continuity between the signal side ground and the battery ground? Hopefully not, then we know that there's no way we should be getting noise on the ESC signal, and hopefully no sync issues.

Can anybody confirm no sync issues with the ESC's/Pixhawk set up properly?
 

R_Lefebvre

Arducopter Developer
3DR is stating that they have tested KDE ESCs with Pixhawk, and the wiring done in the advised manner, and the system works fine. No problems with calibration, and no problems with sync.
 

KDE Direct

KDE Direct, LLC.
Hey Guys,

Just a quick update - Rob Andrews sent me some additional information to the Pixhawk wiring diagrams, and these are two solutions that work excellent with the 3DR system and no problems. He states over 200 hours of flight-time and no issues with the KDE Direct XF ESCs, so as long as proper power is applied to the "Main Out" rail, there's no problem arming OPTO-isolated ESCs. These two diagrams are excellent in showing the proper setups to provide power, and are simply-modifications to the system to achieve proven compatibility.

3DRobotics has also added a note to their systems in response to this work - http://copter.ardupilot.com/wiki/common-pixhawk-wiring-and-quick-start/, so please contact them if you have any further questions on their systems and options for using high-quality components.

View attachment 21843 View attachment 21844
 

Attachments

  • KDE WIRING.png
    KDE WIRING.png
    189.9 KB · Views: 711
  • KDE WIRING 2.png
    KDE WIRING 2.png
    160.6 KB · Views: 594

crcr

Member
Thanks. I try to be accurate.

So Patrick's response confirms a few things. Still odd that we saw a few reports of the KDE ESC's kinda sorta working. There must be another piece to the puzzle still.

I have KDE ESC's, getting no power from the Pixhawk yet I can arm and even fly the machine no problem. But did run into issues when doing an auto tune - it just flipped as soon as the flicked the switch. Is this caused because there is no power running the opto coupler??

Thanks
 

Thank you very much f
Hey Guys,

Just a quick update - Rob Andrews sent me some additional information to the Pixhawk wiring diagrams, and these are two solutions that work excellent with the 3DR system and no problems. He states over 200 hours of flight-time and no issues with the KDE Direct XF ESCs, so as long as proper power is applied to the "Main Out" rail, there's no problem arming OPTO-isolated ESCs. These two diagrams are excellent in showing the proper setups to provide power, and are simply-modifications to the system to achieve proven compatibility.

3DRobotics has also added a note to their systems in response to this work - http://copter.ardupilot.com/wiki/common-pixhawk-wiring-and-quick-start/, so please contact them if you have any further questions on their systems and options for using high-quality components.

View attachment 21843 View attachment 21844
Thank you very much. I will test my setup today and confirm. I am not using the KDE ESCs, I am using the new Tiger Motors Air T40A NO BEC with 6s lipo batteries. I think and believe I have been having sync issues which causes the drone to just flip in mid air within a minute after airbrone.
 

crcr

Member
Are you using Tiger motors? I was having very similar sync issues using Tiger motors and KDE esc's. Swapped out the esc's for Tiger Air's and was then all good.
Tiger ESC's seem to run the KDE motors better but still not ideal.
I think the best is to keep the motor esc combos of the same brand.
 

glider

Member
I use HobbyWing XRotor on the Spy 750 Y6 with KDE 2814XF-515 motors; work great. A quad with same motors uses HobbyWing Quattro 4-in1 ESC and that works well too.

I think you'll find at least some Tiger ESC's are rebranded HobbyWings.
 

Are you using Tiger motors? I was having very similar sync issues using Tiger motors and KDE esc's. Swapped out the esc's for Tiger Air's and was then all good.
Tiger ESC's seem to run the KDE motors better but still not ideal.
I think the best is to keep the motor esc combos of the same brand.
Yes. I am using Tiger motors MN3515. Would you says that once you added the 5v to the main ports on the pixhawk, you stopped experiencing sync issues? I am really hoping this solves my problem or else I would have to go back to Esc with Bec.
 

I use HobbyWing XRotor on the Spy 750 Y6 with KDE 2814XF-515 motors; work great. A quad with same motors uses HobbyWing Quattro 4-in1 ESC and that works well too.

I think you'll find at least some Tiger ESC's are rebranded HobbyWings.
Did you experience sync issues as well and did the wiring specifications provided in this tread help solve the issue?
 

Top