SchrodingersCar
Member
Oh boy, my name is already becoming synonymous with failure!
One prop, two chunks of GPS mast so far!
- Thread starter crayfellow
- Start date
Oh boy, my name is already becoming synonymous with failure!
crayfellow
Member
Haha! Depends on one's perspective. For me it's all about the inevitable learning process and overcoming challenge! If everything was perfect all the time they'd be selling these things in Best Buy stores ... oh wait.
Sent from my iPhone using Tapatalk
Sent from my iPhone using Tapatalk
Mactadpole
Member
Sounds like your on your way to an uneventful maiden. Hopefully!
crayfellow
Member
Let's hope, fellas. I've got a new set of props (of course, just my style, only as many as I need with no spares), APC 16" since the LHS didn't have 15". I'll be fiddling with the CG and will try to find a more level launch site for a test hover.
Shawn, did you reuse some holes for your clever halved-battery-plate mounting? I unfortunately don't have a local source for standoffs, but I do have a handful of 10mm and 12mm aluminum in case that would serve the purpose.
The more I move this thing around the more it reminds me of a card table or something. A game of hearts or gin, anyone??
Shawn, did you reuse some holes for your clever halved-battery-plate mounting? I unfortunately don't have a local source for standoffs, but I do have a handful of 10mm and 12mm aluminum in case that would serve the purpose.
The more I move this thing around the more it reminds me of a card table or something. A game of hearts or gin, anyone??
Mactadpole
Member
After it was cut in half I reused the holes on the inside (left finger in pic) and drilled new holes for the outside. Used the mushroom head style screws on the outside since the battery would be on top. Second picture I just set the broken tray on top of the screws to give you an idea of what it looks like.
crayfellow
Member
Perfect - thanks much.
Sent from my iPhone using Tapatalk
Sent from my iPhone using Tapatalk
crayfellow
Member
Alright, so one more chunk of GPS mast down. Probably about time to switch that one out finally, it's had its day 
So - wow - altogether a good night in that a flight occurred at a nice big school field nearby.
View attachment 25544 View attachment 25545
@Mactadpole you will notice I used your water jug weight balancing technique, and the battery setup is as of yet unchanged. This is research, after all The extra 2+lb down there made a huge difference in the feeling of balance.
As for the throttle range, man, there is still much to be learned there. I do know I don't have it right quite yet. I tried to go through it systematically.
I first attempted to fly in Loiter with RC3_MIN 1100, RC3_MAX 1900, RC3_TRIM 1100 and existing ESC calibration: props did not spin upon arming, but throttle curve seemed smooth from there. Unfortunately I do not recall the radio config when the last ESC calibration was done, possibly the values from RC calibration (which are whatever Taranis is giving, and is not 1100-1900us).
Upon landing, it landed clean, but some of the props do not stop upon disarm.
So I went back in and tried setting it back to whatever I get when I do radio calibration, then did ESC calibration again just out of curiosity. Again, the props do not spin upon arming, and seem anemic on the throttle curve, so this is definitely not an improvement.
Somewhere along the way I changed the values back to 1100-1900, and upon landing that time, moving the throttle down and yaw left to disarm actually caused the props to speed up as if I wanted it to yaw in the air! With throttle at 0!
I tweaked it again, went back to the config I was pretty certain I had in the beginning, and tried another short flight. This time it slowly flew off and I lost orientation and tried to land it, but it tipped over and sheared the grass a bit. Then it sheared my hand when I was disconnecting the battery to get the props to stop turning since switching off the tx had no effect.
Bottom line, I need to figure out what the correct order of operations are between overriding min, max, and trim and ESC calibration. Somehow I am still not in sync somehow. diydrones seems like mostly a dead zone these days so I am unsure that is the place to ask; I may just ask Patrick at KDE in case they have some test rigs running Pixhawk.
So - wow - altogether a good night in that a flight occurred at a nice big school field nearby.
View attachment 25544 View attachment 25545
@Mactadpole you will notice I used your water jug weight balancing technique, and the battery setup is as of yet unchanged. This is research, after all
The extra 2+lb down there made a huge difference in the feeling of balance.As for the throttle range, man, there is still much to be learned there. I do know I don't have it right quite yet. I tried to go through it systematically.
I first attempted to fly in Loiter with RC3_MIN 1100, RC3_MAX 1900, RC3_TRIM 1100 and existing ESC calibration: props did not spin upon arming, but throttle curve seemed smooth from there. Unfortunately I do not recall the radio config when the last ESC calibration was done, possibly the values from RC calibration (which are whatever Taranis is giving, and is not 1100-1900us).
Upon landing, it landed clean, but some of the props do not stop upon disarm.
So I went back in and tried setting it back to whatever I get when I do radio calibration, then did ESC calibration again just out of curiosity. Again, the props do not spin upon arming, and seem anemic on the throttle curve, so this is definitely not an improvement.
Somewhere along the way I changed the values back to 1100-1900, and upon landing that time, moving the throttle down and yaw left to disarm actually caused the props to speed up as if I wanted it to yaw in the air! With throttle at 0!
I tweaked it again, went back to the config I was pretty certain I had in the beginning, and tried another short flight. This time it slowly flew off and I lost orientation and tried to land it, but it tipped over and sheared the grass a bit. Then it sheared my hand when I was disconnecting the battery to get the props to stop turning since switching off the tx had no effect.
Bottom line, I need to figure out what the correct order of operations are between overriding min, max, and trim and ESC calibration. Somehow I am still not in sync somehow. diydrones seems like mostly a dead zone these days so I am unsure that is the place to ask; I may just ask Patrick at KDE in case they have some test rigs running Pixhawk.
Attachments
Mactadpole
Member
Man I give you kudos for perseverance. Glad you got it going but please be careful with that spinning prop situation, maybe take some leather gloves next time. I just made a mental note - don't use kde esc's. Definitely would ask Patrick at this point.
crayfellow
Member
I think the KDE ESC's are good; what is lacking is the APM documentation. I can't imagine there aren't others running this combination, but maybe they just know what to tweak. It's such a simple thing, the PWM range from the radio and how it is ultimately delivered to the ESC's, but it's definitely not sufficient to simply set the min/max values in the Mission Planner advanced parameters.
One way or another, I'll figure it out then we will have them update the docs and you'll be able to give KDE ESC's a try!
Now that I know how the condition manifests itself, I can test by simply arming, then disarming, to see if it behaves. Pretty weird though, huh?
crayfellow
Member
the more I read, the more I think this would be a nonissue if I did just override the min/max values on the Taranis to be 1100-1900. Then RC calibration and ESC calibration would get the proper values. I believe based on my studies that ESC calibration on truly opto-isolated ESC's actually has no real impact, since the range is indeed fixed.
So tomorrow if the weather is good I will give that a try.
EDIT: I've got the Taranis set up. For anyone having a similar issue and needing to adjust the PWM range for the throttle channel for opto-isolated ESC's with a fixed range, this is done in the channel or "mix" settings. You can see the values in the "servos" screen, just scroll down to the channel in question and it will show its pulse width. I adjusted the "weight" of the throttle channel in the mix settings from 100 down to 78. Now this channel behaves like a Futaba/Spektrum throttle channel would, giving a range from 1100-1900.
Now, when I do an APM RC calibration, I can expect to see 1100-1900 for the throttle channel range, and will not have to override anything in the advanced configuration. If this works I will ask the APM folks to update the "Connect ESC's..." documentation, which specifies KDE but is really going to be crucial for any true opto-isolated ESC.
So tomorrow if the weather is good I will give that a try.
EDIT: I've got the Taranis set up. For anyone having a similar issue and needing to adjust the PWM range for the throttle channel for opto-isolated ESC's with a fixed range, this is done in the channel or "mix" settings. You can see the values in the "servos" screen, just scroll down to the channel in question and it will show its pulse width. I adjusted the "weight" of the throttle channel in the mix settings from 100 down to 78. Now this channel behaves like a Futaba/Spektrum throttle channel would, giving a range from 1100-1900.
Now, when I do an APM RC calibration, I can expect to see 1100-1900 for the throttle channel range, and will not have to override anything in the advanced configuration. If this works I will ask the APM folks to update the "Connect ESC's..." documentation, which specifies KDE but is really going to be crucial for any true opto-isolated ESC.
Last edited by a moderator:
fltundra
Member
Motopreserve
Drone Enthusiast
Wow. Glad you didn't get more seriously injured while reaching in with spinning props!
Definitely a good call to hang weight. Especially considering the design was based on having the gimbal down there. I would imagine if the power system is designed to lift (x) grams at 50% throttle, lifting off with (y) AUW would make the craft/throttle VERY twitchy.
@jfro uses old hard drives for strapping underneath to approximate gimbal weight. Good use of old drives, and they don't swing around.
Also, KDE escs do not need calibration. They are pre-configured from the factory. Have you spoken to Patrick yet?
Definitely a good call to hang weight. Especially considering the design was based on having the gimbal down there. I would imagine if the power system is designed to lift (x) grams at 50% throttle, lifting off with (y) AUW would make the craft/throttle VERY twitchy.
@jfro uses old hard drives for strapping underneath to approximate gimbal weight. Good use of old drives, and they don't swing around.
Also, KDE escs do not need calibration. They are pre-configured from the factory. Have you spoken to Patrick yet?
crayfellow
Member
Yeah.. in retrospect I could have triggered failsafe differently via the laptop, but it was at an angle against the ground so I had to make the call right then. I was extremely careful, but then again these are big props.
I am trying again today with the tx reconfigured to 1100-1900, and RC calibration showing those same numbers, and with ESC calibration redone. So we shall see. All I need to do is arm, see if the motors spin, then disarm, and see if the motors stop.
I do want to reiterate the challenge has to do with the opto+pixhawk combination, where the PWM range is inherently fixed. KDE happens to be opto-isolated, and I happen to be using an OpenTX radio (Taranis) that has a wider range by default. So I cut my hand due to challenges in integration, not because KDE ESC's are trouble (or Pixhawk, or Taranis for that matter).
crayfellow
Member
You are right, it is an unnecessary formality due to the opto. I will talk with him if my simple test does not yield the results I expect it to.
Motopreserve
Drone Enthusiast
Agreed, I did not think the KDE would be the issue. I have used them successfully with the Taranis for a while (but never Pixhawk). And i have used the Taranis with Pixhawk (but never KDE) so it seems the specific combo you have is the issue.
crayfellow
Member
OK I tried it. I've got the Taranis locked down to 1100-1900us range on throttle, redid RC calibration, and tried arming. The motors do not spin upon arming, however I was able to add a few clicks of throttle just to get the motors to spin, then disarm to confirm they stopped spinning. I think I am close.
Maybe I just need to adjust the PWM preset for the motor's neutral position so they spin upon arming, now?
Motopreserve
Drone Enthusiast
When you say few clicks of throttle, you mean with the sub-trim?
crayfellow
Member
just a few detents on the throttle stick (same as a few clicks of trim, essentially)
Motopreserve
Drone Enthusiast
You had to move it up to get it to engage the arming?
Typically its the other way around - requiring lowering the trim to get it low enough to kick in.
You said that the throttle shows the correct reading in mission planner? And the taranis stock cal has been done?
Typically its the other way around - requiring lowering the trim to get it low enough to kick in.
You said that the throttle shows the correct reading in mission planner? And the taranis stock cal has been done?