Saving a Lotus T580+P (Part 1)

LotusRC-T580P

A little background first.

Multi-Rotors are cool, after playing with the blade 350QX and my DJI based Hex I thought I wanted a larger, more stable camera platform.   I thought I wanted a Carbon Fiber frame.

Clearly I didn’t do enough research when this thing was acquired.   If you search the interwebs you’ll find some pretty glowing reviews of this flying turd.  If you search harder you’ll find some more honest reviews.

It’s a turd.

I kind of, sort of knew this going in, but I also figured worst case scenario, I’d just gut it and use the frame.  After all the price was right, it was cheap.

Well turns out, that’s what I’m going to do.   Actually that’s what I have to do because out of the box, this thing didn’t even make it off the bench before showing it’s true colors.

This is a larger Quad that swings 14” props, via some Large 400KV motors.

So it comes pre-assembled, all folded up in the box.

photo

All you really need do is slap a 4 channel or better receiver on it, a battery and go.    So that’s what I did, only if you dig around the net, you’ll find out these things have a ‘flip of death’ problem.   This thing is not going to survive a hard crash, it’s just not that stout.

Simply during testing on the bench w/o props, I could get a motor to lose sync with the ESC.   A proprietary 4 in 1 ESC.   So there’s really no point in even trying to get it off the ground.

Under the lid is a really, really poorly put together 4:1 ESC, and flight controller.

T580P UnderCover

A flight controller and ESC you have absolutely no control over.  As in none it is not user tunable.   For a copter that was originally supposed to be in the $500+ range but can be had for $300 this is still a POS.   (There are I think, 3 versions of this each claims to fix the ESC flip of death issue, but obviously that’s not the case).  There are better frames and components to be had, which probably explains why they don’t really make these or even have them on their website anymore.   They can be found on eBay and some sites, but save yourself the trouble and bypass this.

So what are we going to do?

We have a decent frame, not great but decent.   It’s billed as foldable, but you still need tools, to remove the landing gear in order to fold it.

It has motors, that seem decent, although it’s possible that it’s not the ESC and could be a motor that’s bad and if so, I doubt I can get a replacement and I’d have to replace all 4 so that they match.   The motor mounts are weak, and will likely pop in even a modest crash.  If it really needs a motor, this things will get cut into CF pieces to be used elsewhere and then at that point, even a the low cost of entry it will still be a loss.   One of those life lessons.

So we’re going to try and make this air worthy.

We’ll replace the ESC, in fact I’m going to use this one:

41JNpZwBCjL

The build has started:

T580-ReDo

I’m on the fence as to what flight controller to use.   I really like the Afro Acro Naze32 that’s mounted on my Armattan CNC 355 Quad.

But for something this size for use with a camera or photography (read a docile flying machine, non acrobatic).  I’m pretty sure I want GPS assist for loiter mode and other features.   So I’m considering a FULL Naze32, or maybe even another APM or Pixhawk setup.

The later is kind of pricey, about $400 all said and done just in flight controller, GPS, and telemetry radios.  But if the experiment fails, well I can fork lift that to something else.    I should be able to do the same with the Naze for about 1/2 of that perhaps less.

Maybe I’ll just toss a $25 afro acro naze, on it, fly it a bit and if it survives that then I’ll upgrade it.    I really don’t expect that it will take much of a crash to completely destroy this frame.

So currently the build is stalled waiting for a flight controller to free up.  A Naze is on order for now, but we’ll see.    I’ll post an update when this project progresses.

Happy multi-rotoring and remember; “Friends don’t let Friends buy LotusRC products”.

The Dead Cat Quad

Yesterday was a good multi-rotor day.  I spent the day with my buddy building quads and putting together my ”Dead Cat” or Alien Quad, while he assembled his High Dollar Team BlackSheep Quad.

Basically the same basic layout, a Quadcopter, where the front arms are spread more than the rear, to allow for better camera clearance from the props (as in getting them out of the picture).

The ultimate goal will be to make these FPV or First Person View fliers.

Two completely different approaches.  He, plopping down major cash for a TBS kit, with TBS motors and ESC’s.  The TBS board, with integrated gimbal is amazingly well done.   The onboard core will be outstanding when finished.  He set it up yesterday migrating a Naza flight controller he already had.

I on the other hand was going the cheaper route.  A simple frame.   I was still waiting for some parts, so I robbed/borrowed ESC’s and motors from a failed DJI F550 hex build that he had on hand.   I did however use my shiny new PixHawk flight controller, 3DR Ublox GPS and 3DR telemetry radios.   Basically the same gear I have on my Hex, with the exception of the newer PixHawk over APM 2.6 I’ll have some pretty significant comments about the PixHawk towards the end.

This is a super cheap frame ($25.00 on amazon, complete with knock off arms), I’ll replace those with legit arms after the first crash, and when I really go FPV, I’ll probably upgrade to the TBS plates minus the gimbal.

It turned out pretty nice and flies pretty well, still needs to be tuned but it’s too windy today.

DeadCatFrame001

DeadCatFrame002DeadCatFrame003

I tossed some DJI Landing gear on it so that I could mount the battery underneath.  Without the camera gear up front, putting the batter in the back where it’s supposed to go simply wrecked the center of gravity.   In FPV mode that won’t be the case.

PixHawk Critiques and Observations

I so want to like APM, and I so want to support 3DR but they are making it harder every day.

APM, on my APM 2.6 Flight controller works pretty well.  I know it’s basically out of resources so they are moving to PixHawk.   I know BeagleBone will be 100x better and that’s coming.   I know they are trying to make this ‘consumer friendly’. 

It’s not plug and play, at least with 3.1.2 copter code, there are too many hoops to jump through, to many places you have to rub your head and your belly in opposite directions at the same time.   Maybe these are mission planner issues, but they don’t exist with my APM and they do with PixHawk.   Connecting, downloading, fetching parameters and logs.  It works, oh about half the time.   Try to do something more than once and you’ll probably have to reboot the PixHawk.   I know it’s “Open Source”, but some of this is pretty stinking buggy.   No other way to say it.

The PixHawk is noisy, too noisy and by that I mean too many sounds, and beeps.   My ESC’s hate life until you push and hold the button.   Maybe this is their way to make ‘drones’ more conspicuous.   It’s just an annoyance.  We should be able to turn the sounds down or off.   We should also be able to make my ESC’s happy, pre button push.

Speaking of the button.  The button is dumb.  I have to pre-arm the PixHawk with a button push and hold before I can arm it?  What the hell?   A lawyer probably made them add this.

Ok fine, put the button in/on the PixHawk, allow the remote button if and only if you need a remote button.  I don’t, so it’s one more thing to mount, one more set of wires to route and tend too.   But let me, as a non-novice, disable this silly preflight, preflight check.

I don’t hate the case, others do, in fact someone’s already making square non-sexy case for it.   I actually like it.
side_back

I do like the LED’s, tons of them.  The status LED is really nice.   But if I can see that, I can push it, that remote button should have a multi-color LED like the top.   So when you remote mount it you get the light where you can see it.

OK, I think my batteries should be charged…

Update: 03/11/2014

OK, so on assembly day, it got dark, all we could do was hover it, and it was clear there that the battery w/o camera in the battery slot was bad.  It was woefully out of balance.

So I put landing gear on it to put the battery underneath.  It flew, in fact in Stability mode it flew just fine.  It loitered OK too, and looking the logs, vibrations on the PixHawk were all OK too (under +/- 3).  The darn thing would NOT auto-tune.   I  chalked it up to being windy Sunday, too windy.

Today was a different story.   Nice and calm.   It still would NOT complete an Auto-tune.  It would ‘finish’ but if I tried to switch back and forth between the pre and auto-tuned values, it would drop like a rock, get hung up in it’s own prop wash and crash.   Luckily after 3 attempts the only thing I broke was a single landing leg.   After looking at logs, I just couldn’t decide if this frame sucked that bad, or if PixHawk was just crap.   It’s less than ideal for a product that was supposed to ship in Sept/Oct of 2013.  It’s now March and it’s still less than polished.  So on a whim, I decided I had spare arms, let’s replace this HJ DJI knock off arms with legit DJI arms.   So that’s what I did.

WOW, that’s all I can say is WOW, with stock pids, it’s amazing.   Now I’ll try an auto-tune.

So yeah, when you read in the internets that the HK knock off arms are bad.  Believe that player.  While they don’t look or feel horrible, they just are.

More Multi-Rotor Fun.

First, the weather was mild Thursday Night so I slapped my legacy ContourHD camera onto my Armattan “Sport Quad” for some sunset flipping and rolling.   Working on some acro skills. 

Armattan CNC 355 Quad.

Then after getting home Saturday.  I went outside with Matthew and let him get some stick time on the Franken Quad.   Once he got bored, I noticed the Hex was calling me.   So I fired up Mission Planner and decided to pick up where I left off when I hit the trees.

Plotted out a new autonomous route, verified the feet/meters settings, and let her rip.

No music in this vid, only prop and wind noise. 

DJI F550 Hex, APM 2.6 running Arducopter 3.1.2

Certainly open to suggestions for the Jello video.  It seems EFI related and once you get flying is almost gone, but when the Hex is just sitting there, it’s bad real bad.

Acro Quad Flight

Flying a quad or multi-rotor is a lot of fun.   Flying isn’t really the proper term though.  Like a  helicopter, they don’t ‘fly’ traditionally.   They simply beat the air into submission.

Collective pitch helicopters are hard, multi-rotors when tuned well are not.  But getting this proper tune and balance is hard.

Doing acrobatic maneuvers isn’t simple.   By design, multi-rotors have a flight controller that’s working really hard to keep the thing level and they often limit your ability to Roll or Pitch too far.   You have to turn that stuff off if you want to get jiggly with it.

An upside down multi rotor that’s no longer fighting gravity, and is instead assisting it, usually ends up badly as it plummets into the ground and breaks things.   It also can be really easy to lose your orientation.  If you’re upside down and your right is your left and/or your forward is aft.   Well, that’s generally not good.   The results usually suck.

We had a break in the weather and I finally got to spend some time really tossing my Armattan qaud around.    This thing is awesome.

Chris really makes an outstanding product.

My CNC 355 is using the Afro Acro Naze32 flight controller.   This is just a quad for quad’s sake.  No GPS, No compass, no other smarts other than accelerometer and gyro stabilization which can be turned off and/or tuned.   And fine tuning we will do.

Getting Jiggly with my Armattan CNC 355 Quad

 

It’s tough as nails.  I’ve crashed it multiple times, usually only losing props, which are cheap.  I’ve bent arms, and you just bend them back.   Might need to take them off to really straighten them but you can just bend them back as you’ll see at the end of the video.

If you have a Blade or a Phantom and you want more performance and don’t want to stop buying plastic parts.   Buy an Armattan Quad.   You won’t regret it.

Building flight skills apparently just takes time.  There is no magic pill.   I think, or at least hope I’m over the hump cause crashing sucks, and crashing you will do when you start doing stuff like this.   Also, trees are still the enemy.

Not even 24 hours later and we have snow again.

But I just had to test the PID and Rate changes I made after yesterday’s fight above.

Freshly balanced props too.   Wow, so much more responsive, pretty easy to flip and roll without losing altitude.

On quad camera, front porch launch cause I’m not dressed for the weather 🙂

More Hex Gimbal Footage

Last weekend was a good multi-rotor weekend.  Got to fly my high-performance quad a few times without incident.  Also flew my Franken-QX350 a bunch.  Each of the kids got a turn too.   All without incident.

Finally got spend some time getting the hex to fly right.  It would fly great in stabilized mode but loiter was broken, and if loiter is broken, none of the autonomous flight modes are available.  Well, they are available but you’ll just crash so it was in my best interest to get that fixed.    Turns out, somewhere along the way my compass orientation flag had changed form 8 which is roll-180, to 10, which is roll-180 yaw 90.   Honestly the way things are mounted I think they should be roll 180 yaw 270, but since it works just rolled, I’m leaving it that way.

For what it’s worth, my APM is mounted like this.

APMOrientation

After a handful of manual flights and confident that loiter was fixed, I launched it a few times and flew it using droid planner.  Guiding it to fly here, then over there, and then over there.

Still have some Video Jello-issues to work out.   Looks/feels like electrical interference from something, likely with the Gimbal.

Super long video for my own amusement.

More DJI Hex, Gimbal enabled footage. Still working out some bugs.