KDE Direct XF UAS ESC and Flight Controller Compatibility List

[MongranPerspective]

I tested my setup this evening and my quad flipped and crashed again. My setup incorporates 6s battery, Tiger MN3515 with Tiger motors Air 40A, 5V secondary power from a BEC to one of the Main Ports on the pixhawk and it still crashed. When I checked the logs my Roll and DisRoll were not aligned. When I also checked all the RCOut settings, Channel 1 increased while Channel 2 dropped its PWM and then crashed. I do not know if the issue is with the Pixhawk or the ESC. Does anyone have an idea of what I should do?

 

[glider]

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.

The only OPTO ESC I've had good results with on a Pixhawk is HobbyWing. Right now I'm currently having problems using KDE ESC's on a CUAV Pixhack (no, not a cheap Chinese Pixhawk knockoff). First, the PM is powering 5v to the FC and there is a HW UBEC 5v on the rail.

The copter arms. Three motors spin properly, but one motor will spin for a few seconds, then stop. Increasing the throttle a bit the motor will spin up again, but stop after a few more seconds. Rinse and repeat.

 

[R_Lefebvre]

Yeah, I've recently heard from a reliable source that this is not a closed issue. Even with power supply to the ESC from the Pixhawk, they're still losing sync. The concept never really made sense to me, if the ESC's are true Optically Isolated, then they should not power up *at all* if there is no 5V power supply. That is how real Opto helicopter ESC's work. They're completely dead without 5V supply on the signal cable.

 

[glider]

Well thanks Rob, so far you're the only one who has replied out of 4 threads I've posted. Each ESC/motor works fine separately, but once all wired, that one motor will slow down and stop spinning. I should have known better than to do all that soldering only to have this happen. I made a video of what's its doing and emailed KDE, but see no reason why it shouldn't work with the way I have it powered.

I know HW XRotor isn't true OPTO, but both that and Quattro work great on the KDE motors on my copters; really like the 515's, probably the best 4s motors made in that class IMO. And the Quattro has BEC to boot. I wanted the KDE for the good braking etc. it has; compared a friend's similar size copter with mine and his seemed more responsive with the KDE ESC on the KDE motors.

If Patrick doesn't strongly suspect a bad ESC, they are all coming out and the Quattro is going in; nice and neat, quick install, no spaghetti mess, runs smooth and stays cool. In fact, I doubt even if it is a bad ESC I'll bother replacing it; that would be a very bad hair day. The PDB, for lack of a better word, is homemade using washers with 4 JST leads (I always regret not having enough 12v plugs LOL). It's nice and neat, but to remove it all and replace one ESC will take a long time. I wanted this thing in the air by week's end. Fortunately I have no shortage of other ESC's.

I don't understand why some have no issues while others do.

P.S. It's stuff like this that almost make me want a RTF.

 

[KDE Direct]

Hey Everyone,

There's no reason the system would not be working, assuming the internal electronics are the same as the Pixhawk system (3DR uses our ESCs in their commercial-group as well), there is power being applied to the servo-rail to properly arm the OPTO-circuitry, and the programming on the TX and Mission Planner is correct.

I just received this from an alternate customer, who was having a similar problem with the motors running consistent, and it came down to proper calibration of the TX. I'm not a fan or user of the Taranis radio, purely by the build-quality and low-cost components inside, but hopefully this will help shed-light on the programming if this is the system being used:

Here's the quote from the e-mail:

"We solved the problem this way:

-on Taranis servo ch3 we put 1520 middle value -on mission planner radio calibration we moved the max and min from the Taranis until we reached 1100 min and 1900 max -then we entered calibration mode on the copter; powering it with the ch3 stick on high and esc 2 on parameter list in mission planner -now it works we have a correct the range."

View attachment 26404 View attachment 26405 View attachment 26406

Please let me know if this helps and if not, then I can help involve 3DR in this as well, to explain the proper calibration routine. Also, please test the ESC outside of the Flight Controller (via direct connection to a RX), which will rule out the ESC if possible (or potentially find a flaw in the ESC, works great to identify the root cause). We have an instructional video posted online to help with this:

 

[Old Man]

FYI, similar PWM functions impact the E.T. Vector as well. I think many forget our transmitters base intent is for RC airplanes, where servos having large throw capability are frequently used to increase flight surface deflection. Any travel volume in excess of 100% is intended for driving a servo beyond a standard travel arc of 60*, not for ESC's.

For multirotors our "travel" range needs to be between -100 and +100, -1000 and +1000, or 1000-2000 depending on your type of transmitter. Essentially a PWM range of 1 to 2.

 

[glider]

Hey Everyone,

There's no reason the system would not be working, assuming the internal electronics are the same as the Pixhawk system (3DR uses our ESCs in their commercial-group as well), there is power being applied to the servo-rail to properly arm the OPTO-circuitry, and the programming on the TX and Mission Planner is correct.

I just received this from an alternate customer, who was having a similar problem with the motors running consistent, and it came down to proper calibration of the TX. I'm not a fan or user of the Taranis radio, purely by the build-quality and low-cost components inside, but hopefully this will help shed-light on the programming if this is the system being used:

Here's the quote from the e-mail:

"We solved the problem this way:

-on Taranis servo ch3 we put 1520 middle value -on mission planner radio calibration we moved the max and min from the Taranis until we reached 1100 min and 1900 max -then we entered calibration mode on the copter; powering it with the ch3 stick on high and esc 2 on parameter list in mission planner -now it works we have a correct the range."

View attachment 26404 View attachment 26405 View attachment 26406

Please let me know if this helps and if not, then I can help involve 3DR in this as well, to explain the proper calibration routine. Also, please test the ESC outside of the Flight Controller (via direct connection to a RX), which will rule out the ESC if possible (or potentially find a flaw in the ESC, works great to identify the root cause). We have an instructional video posted online to help with this:

Well by Joe that fixed the problem. How amazingly simple. Thanks Patrick.

Now we'll get things finished up and see what happens with props on.

I did notice now after exercising the throttle full up and down a few times the motors will spin up and stay there after moving the throttle back down.....takes a bit of time for them to spin down. I suspect that is not a problem but rather interaction between the FC and ESC's; we'll see in a few days. The main thing is upon initial startup, all 4 motors spin up without one petering out and will stay running at low/mid/full throttle consistently.

Again, thanks Patrick for posting this. That is one reason why I continue buying KDE stuff, not to mention the 515's are beasts.

 

[glider]

Well I spoke too soon. Today I tested again and performed Compassmot calibration. It's doing it again. I looked at the end points in MP and they are 1100/1900. The radio is set for 1520 mid throttle.

Could there be a throttle setting in MP that's contributing to this? I'm definitely not taking this thing up in the air until the issue is resolved.

BTW, it's not always the same motor.

 

[glider]

Update: I changed THR_MIN to 0 after bumping it up to 260 noticing it caused the motors to behave differently, not just the low throttle speed. When set to zero, the motors now will spin when armed instead of having to move the throttle up. After 15 seconds at min throttle, they stop, which is normal. That is a good sign. It seems there is (or was) a setting (or possibly a slider) in MP that set the speed of the motors at min throttle. I never had to mess with that so don't recall where it is, but believe it changes THR_MIN and by default is >0.

I think there is a conflict with the ESC and the parameters in Arducopter due to the KDE ESC self calibrating upon startup, and hopefully it is THR_MIN needing to be set at 0 that fixes the odd behavior of the motor(s) spinning up and down, stopping etc. I think it is possible with all the changes in AC code this past year that certain parameters must be set correctly to get these ESC's to work right. Trying to calibrate using MP like a typical ESC is futile, which makes sense. Setting the end points and mid correctly is only part of the solution.

One thing I noticed is how fast the motors respond to hard throttle movement; instant response like a well tuned drag race motor when mashing the pedal.

Tomorrow, weather permitting, I'll do some more testing, and if I get that warm fuzzy feeling, will attempt to put the bird up. If all goes well and I can narrow down the correct procedure, I will submit it to the devs at DIY to put in the Wiki.

My 2c

 

[glider]

Update #2

I found the setting for keeping the motors spinning at min throttle:

Also note:

Hmm. Maybe that has an effect on the KDE ESC as well. By default it is "Slow". As I recall when updating the ESC firmware, there was a setting for the throttle min speed? Not sure. That should be the last of the parameters I can think of that could affect the ESC.

There are these as well, but I'm not sure what they are for:

Edit: MOT_THR_MIX_MIN explained here:
http://diydrones.com/forum/topics/copter-3-3-beta-testing?commentId=705844:Comment:1998088

Edit #2: Sheesh. I should have researched this better and certainly hope they all don't potentially have a negative affect on the KDE ESC!!. There are several changes since 3.1.5 (knew that fairly well).
https://groups.google.com/forum/#!topic/drones-discuss/MZxDO-XMlQk

 

[glider]

Today is epic fail. I give up.

 

[Marco Cecala]

Today is epic fail. I give up.

Glider; I too am in the same boat with motors not responding well. Crayfellow helped me with settings and the motors do not spin at same speed while on the ground but
Upon liftoff they seem to start behaving well.

I am still uneasy with this behavior as I have had one motor always spinning slow, I replaced the Plxhawk and it got better. I sure wish there could be an answer to this issue. I'd hate to have to abandon KDE.

 

[glider]

Glider; I too am in the same boat with motors not responding well. Crayfellow helped me with settings and the motors do not spin at same speed while on the ground but
Upon liftoff they seem to start behaving well.

I am still uneasy with this behavior as I have had one motor always spinning slow, I replaced the Plxhawk and it got better. I sure wish there could be an answer to this issue. I'd hate to have to abandon KDE.

Thanks. Have you flown a number of flights? I've never seen this type of behavior with any other ESC, so watching a motor slow down and stop is a very uneasy feeling. Have you tried Auto Tune? I've read someone trying AT and their copter flipped.

Ok, ignore my last post and see below

With MOT_SPIN_ARMED at 130 (so motors spin when armed) and THR_MIN at 0 and your encouragement, I decided to take it up no holds barred; if it crashes so be it.

Well what do you know! It flies! Albeit a bit wobbly, but it is a brand new build and first takeoff and I don't think it is a sync issue, at least that I could hear. I'm using a Pixhack on this quad, so it will be interesting to see what the vibration levels are as it is hard mounted.

Thank you for chiming in here. I was THAT close to tearing down.

glider

 

[Marco Cecala]

Thanks. Have you flown a number of flights? I've never seen this type of behavior with any other ESC, so watching a motor slow down and stop is a very uneasy feeling. Have you tried Auto Tune? I've read someone trying AT and their copter flipped.

Ok, ignore my last post and see below

With MOT_SPIN_ARMED at 130 (so motors spin when armed) and THR_MIN at 0 and your encouragement, I decided to take it up no holds barred; if it crashes so be it.

Well what do you know! It flies! Albeit a bit wobbly, but it is a brand new build and first takeoff and I don't think it is a sync issue, at least that I could hear. I'm using a Pixhack on this quad, so it will be interesting to see what the vibration levels are as it is hard mounted.

Thank you for chiming in here. I was THAT close to tearing down.

glider

We did a flight with stock PID's and it was twitchy. Went to some lower settings and did OK. Flying Wednesday with larger props, and hopefully that would be it.

Don't think I'll ever try auto tune. This big quad just needs stable, not snappy.

I know everybody says that PID's will vary, but I would like to see some from other heavy lifters. I bet they are not that far apart.

My specs.
Vulcan quad
KDE 5215
KDE 18.5 x 6.3 triple blade
Maxamps 16000 or 22000
AUW around 20 pounds, hopefully.

 

[glider]

I've used AT dozens of times. I was just referring to someone with KDE ESC's, but don't take that as fact.

So much has changed in AC 3.3 for tuning, I need to learn all over again. I use ch6 for tuning enough for Auto Tune, then from there I'll need to dig into the 3.3 thread at DIYDrones.

This build is an Alien 680 dedicated for GoPro and good gimbal. BTW, here are the vibration levels from the first test flight. I'm using a CUAV Pixhack FC hard mounted (no foam). Quite impressive.

I deleted the pics because they are the wrong ones. Doh!

 

[Mactadpole]

Don't be worried about autotune with AC 3.3 and the big copters. I'm not flying KDE ESC's but I just finished tuning a hex at AUW ~15 pounds. No problem and autotune of yaw is incredible. Tame the tune down with AUTOTUNE_AGGR parameter (I used 0.08) and then smooth out flight characteristics with adjustments to these parameters on the Pixhawk and maybe some expo:

1. RC_FEEL_RP
2. ATC_ACCEL_RP_MAX
3. ATC_RATE_RP_MAX

See this post for some details: http://www.multi-rotor.co.uk/index.php?topic=7625.0

 

[glider]

I'm happy to report the KDE 35A ESC's are working fine so far. In addition to what Patrick from KDE posted, I set THR_MIN to 0 and set MOT_SPIN_ARMED to 130.

 

[glider]

Nope, things are still quirky. I can't quite pin it down, but there seems to be issues with how the ESC's calibrate and react to the settings in the parameters above. Yesterday one motor refused to spin throughout the entire throttle range. I had to set those parameters both to 0. It's not consistent.

After landing, the motors would not disarm at one point. I had to use the "kill" option in Tower. The ESC's are hyper sensitive to AC parameters it appears. Once in the air it flew ok, but while doing Auto Tune a few days back, it nearly flipped. After reviewing the logs, the VIBE levels were high, however this weird "near flipping" and Z axis full throttling has never happened on any copter I've built to date during AT. I remounted the FC to better isolate vibration. Excellent.

I then completed Auto Tune and PID's look good. However, it sure seems like the motors are trying to desync after arming on occasion; I then shut down, reboot and things seemed ok, but if I don't get the motors spun up, it's like the signals to the motors go wonky.

Baffled.

 

[Marco Cecala]

A couple things to try. I solved my problem with a new Pixhawk. Crayfellow felt a full delete of the EPROMs on the old controller would have worked. I have yet to try it.

The unit is flying well and the motors are behaving. I tried everything on the other controller and never got it in the air.

 

[glider]

A couple things to try. I solved my problem with a new Pixhawk. Crayfellow felt a full delete of the EPROMs on the old controller would have worked. I have yet to try it.

The unit is flying well and the motors are behaving. I tried everything on the other controller and never got it in the air.

This is a brand new FC, motors and ESC's. I finally out of necessity to keep my sanity, replaced them with a HW Quattro which works fine. I may try KDE's again on another project.