MWelectroSpider

[Balu]

A while ago I mentioned that I'm going to build my second quad for a coworker. He asked for more relaxed flying with all electronical help he could get and perhaps be able to later add an action cam for some aerial footage.

So I'm trying to build for that:



Features (to be updated):

• FT Electrohub, dead cat configuration
• 12x12x250 mm wooden (beech) arms
• 10x4.5" cheap HK props
• Suppo A 2212-13, 1000kv motor
• Hobbyking F-30A UBEC ESCs
• SparkyV1 / TauLabs controller
• GPS
• OrangeRx R615X DSM2/DSMX Compatible 6Ch 2.4GHz Receiver w/CPPM

He wants to "try" flying first and see if he will get bit by the RC bug too. Since he has a DX4e, I'm hoping to get it working with the cheap OrangeRX receiver. I have to figure out if the DX4e aux switch is a fifth channel, so he can switch flight modes or have the copter "Return To Launch" - which he really wanted.

First problems I ran into today was the solder I got with my new soldering station. Looks like I have to learn soldering again with lead free solder...

That's it for now. I can hear my bed calling already. Sleep is so overrated *yawn*.

PS: MW is not for MultiWii, but the initials of my coworker

 

[Balu]

Just a quick update on what I'm trying to do different.

For some reason I don't like having my battery on the bottom of the copter. It worked okish on my Anycopter because of the included battery mount plate, but turning it over to switch the battery always felt wrong.

Since the ElectroHub doesn't have the battery mount, a lot of people use the Vibration Dampening Camera Mount Kit. It is a nice idea and the wire method is a great anti-vibration solution, but it still is located below the copter.

Most mini multirotors seem to go a different path and that is some kind of "cage" on top of the copter arms. Sometimes it uses some kind of anti-vibration method to separate the dirty (vibrating) part and the clean part of the frame.

Since the ElectroHub also doesn't allow the flight controller to be mounted "recessed" like on the Anycopter, a cage might help protecting the FC too.

I've looked into different ways of vibration damping, the fairly common vibration damping balls would've been a great solution, but they are not that easy to mount [insert link to mount method with screw and spacers as soon as I find it] on the ElectroHub plate. Also I'm a little afraid that the balls won't hold well in case of a crash. The second option - threaded bobbins - were already reported to not work very well, but I decided to give it a go anyways. Perhaps having them support more weight will have a better result even those these rubber ones feel fairly stiff - we'll see. If it doesn't work I can still try silicone bobbins or something else.



Two of the shown plate are needed to build the "cage". It is 70 mm x 200 mm x 2.5 mm aircraft plywood (birch). The holes are supposed to be spaced around 5 mm off the sides. The X in the center is the calculated CG of the ElectroHub in spider configuration. The gyro/accel sensor should be placed here.



Checking if the battery will fit, using my largest 4000 mAh battery. I'd love to have the battery mounted on the bottom plate, but I see problems with the wiring (the FC needs to go on top this way) and accessibility of the battery.



So I'm trying to mount the electronics on the bottom plate. The GPS is just there as a spacer for a possible GoPro / Mobius or even FPV camera. On the finished build I might just put the GPS on the top plate or on some kind of "mounting pole" - depening on how good the reception is.

Now I'm waiting on nylon spacers and screws to construct the "cage" because I could not find them locally. :-(

 

[Balu]

So I'm experiencing some problems with my "prototyping". First of all I forgot to order the Flexi-Port cable with my Sparky board and it seems to be impossible to find the right connector. I think it is an JST SR/SZ type, but I have not found it in any electronics store here yet.

This means I can not connect the GPS. I'm looking into using another serial port to do that, but since I don't know a lot about Sparky I have to ask around first. Is @peabody124 still around?



Other things I should've done before: Paint the booms and plywood black, think more about where to put spacers, test the electronics before mounting everything... I am having problems with the third motor while using the TauLabs GCS new model wizard.

Something I've noticed is that the wizard doesn't calibrate the ESCs before finding the "motor startup threshold". That's the first thing I would do anyway?

As for the spacers: I decided to go for 8 nylon spacers (31 mm from HK - if I try to fasten them tight, they "jump over the thread") instead of the initially planned 6. This resulted in problems with the Sparky not being USB accessible anymore - and having the motor outputs a little under mechanical stress. To avoid that I simply removed the two problematic spacers. It still is sturdy this way, but looks odd - especially with the heavy battery on the part with less spacers .

I will probably move the middle spacers to the center as I had planned before to get around that.



I also made an ugly 14mm diameter hole in the center of the bottom board to be able to feed the motor control wires through - perfectly aligned with the center hole on the ElectroHub of course.

Things to do now:

• Add foam on top of the Sparky to protect the pressure sensor from wind.
• Figure out a way to use GPS without the right connector.
• Think about if it is necessary to put the GPS "on a stick".
• Add "slots" for battery straps.
• Play with TauLabs GCS more and figure out why M3 behaves irrational.
• Get the transmitter of my colleague to bind to the OrangeRX instead of my Taranis with module.

 

[Quad]

Which GPS did you go with?

 

[Balu]

I got a used Crius CN-06 v2.0 for this frame. It has a u-blox NEO-6M GPS module as far as I know.

When I ordered my Sparky from RTFQs I got two additional GPS modules, just to fiddle around with them. One bigger "uBlox 6H GPS" including a "GPS-Shield Hood" and a "Mini uBlox 6M". Both modules have a compass. But I have just started to look into the GPS stuff.

 

[Balu]

The last few days have been stupidly bad. It's raining all the time and this weekend we have storm level winds with gusts between 100-120 km/h.

I've fiddled around with the frame a little, moved the middle spacers to the middle where they belong and switched to f/f spacers that allow me to use screws on top and bottom. Those work a lot better than the smaller nylon nuts that jumped over the threads when trying to fasten them.

Of course I've test flown it before these changes in a way to small room. I have to admit that 10" props are even more scary than 8" ones when you fly in a restricted area =).

After receiving the correct wires for the GPS, I was able to connect it to the sparky and got 8-10(?) satellites while being stationary. Again there was too much wind to test it with GPS. But at least in theory I should have a GPS supported quad now. I'm looking into options to let me know how many satellites it found without having to use the GCS software, because it's probably a bad idea to start in GPS mode when you don't have a proper fix. Also I'm not sure how good the reception is with interfering signals from the electronics. So I don't know if I have to use an antenna holder for the module yet.

As for the pressure sensor there are rumors that the ones used on the RTFQ sparkys are "cheap" as in "don't work correctly". Looking at the TauLabs GCS oscilloscopes it looks like I might have received one of those.



Y axis shows the altitude in m from 45 to 85. So according to that the quad jumps 20 meters up and down. I have no comparison though, so perhaps that's how they work and you just have to calculate an average? I'll have to wait what the TauLabs community has to say, perhaps I'm just using wrong settings or displaying something different anyway.

 

[Quad]

Very Nice!

 

[Balu]

I just got a Quanton board (hooray for shipping times without leaving the country) which is the reference board for TauLabs and I'm only seeing about 1m "jumps" in the oscilloscope for altitude:



So it's not something I did wrong with the TauLabs GCS, but the pressure sensors on the RTFQ sparkys. The sensors should be identical between both, so I'm guessing Paul got a box of cheaper sensors that didn't go through quality control. I'm not really keen on bashing him though, because I feel a little bad about the problems he had lately. I'll see how he reacts to my mail.

I'll use the Quanton board (which is HUUUGE) for this build now, just so I can get it in the air and give it to my colleague who eagerly want's to fly his new aircraft.

If only the weather would clear up. It's still raining all the time. The stormy gusts would not allow testing / tuning anyway. :-/

 

[Balu]

Another quick update. After having problems with my Sparkys (still no answer from Paul@RTFQs) and switching to a different flight controller, I started playing with bluetooth modules so I could use the Taulabs app on a phone for basic information like how many GPS satellites have been found, etc.

One problem I have with the Taulabs GCS on my Mac is that it crashes a lot. And by a lot I mean there's a ~30% chance for a crash after each action I do. For now I'm trying to ignore that since I'm using a prerelease version and I know there are problems with the QT libraries that are used. But it's making preparing and testing a little difficult.

For everything else, Taulabs seems to be exactly what I wanted for this quad even though I had problems activating the GPS mode yesterday (probably just missing a required module).

I have to figure out why it pitches forwards even though I've calibrated everything (everything but temperature calibration). CG and center of lift are close together and the controller sits at that position. Even using full trim did not allow me to get it level at center stick. It also yaws to the left more or less constantly.

Auto-tuning was an interesting experience. The copter wiggles around for a minute while still being controllable and after that it suggested some sensible parameters for tuning. Since I'm having the other problems, I didn't look closer into that, but it flew nicely after enabling those settings.

I'll have to burn a few more batteries to make it all work, but the weather forecast for the next days seems to allow that.

PS: The Hobbyking LEDs I've used on the arms don't really have any effect in daylight.

 

[HawkMan]

which HK less was that ? the strips they have are super bright, but I was planning to get their alu 120and 150 bars for front and back on my nighthawk 250 I ordered(need to model some bumpers and have shapeways print them for me to mount them on)

 

[Balu]

I used the "High Density Waterproof LED Flexible Strip" in red and blue because those were the only ones available in the European store at that time. While unpacking I immediately regretted that decision, because they smelled unbelievable bad.

Are you talking about the single 3W LEDs? I'm not sure if 3D printed bumpers will work with those. There is probably a reason for the aluminium heatsinks Also those seem to be 2.5 V and require an additional resistor to limit the current.

 

[HawkMan]

No I'm talking about these

http://www.hobbyking.com/hobbyking/store/RC_PRODUCT_SEARCH.asp?strSearch=LED+alloy

problem with the alu ones is that the existing bumpers people have made are either convex or concave... won't work with straight alu plates these should also work connected directly to 2S and 3S batteries, 3S being what's going to be used anyway.

As for ordering from europe versus international. I usually only order small stuff I don't need fast form the europe warehouse, cause even they aren't very fast, and the difference in shipping between them and UPS isn't to huge. batteries there's no point in ordering from europe, fedex shipping is like 25-30 dollars. NL registered mail is like almost or around 60 euros last I checked... it's worth it to pay the MVA on the batteries as well(Fedex won't let you write down the price..)


The water proof ones, as far as I know are the same as the non water proof except for the smelly rubber

 

[Balu]

I've ordered a few props in one order and different sizes of nylon spacers and screws in another. Those are cheap enough to not require tracking, so I payed $3.50 and had them delivered 3 or 4 days later. That's a lot better than the international warehouse.

Can you shorten the alloy ones?

 

[HawkMan]

I've ordered a few props in one order and different sizes of nylon spacers and screws in another. Those are cheap enough to not require tracking, so I payed and had them delivered 3 or 4 days later. That's a lot better than the international warehouse.

Can you shorten the alloy ones?

Props and other smaller stuff is about all I order from the euro warehouse or if I want it reasonably fast and they have it and the shipping is decent. sometimes if it's bigger order I might as well order from the international warehouse anyway since it's cheaper and chances are big the euro warehouse don't have the stuff, they're terrible at small parts and components.

I doubt you can shorten the alu ones, it's basically one very ling led component ish.

 

[Balu]

Aaaand it's gone. :-/

My coworker really got into flying it. He re-built it into a standard quad though because we were having some issues with stability on the Spiderquad. He actually flew a lot more then I can because he has a big garden and apparently more time than me.

But now he got too brave and started flying in higher winds and further away. And by further I mean too far...

He also didn't follow one rule I explicitely told him: If you lose control of your copter / orientation and your first try to recover doesn't work - let it drop. A broken quad is better than a missing one.


I guess I now know why he wanted GPS with return-to-home so badly.

 

[jhitesma]

Bummer.

Double bummer about giving up on the spider with what Peabody did last night:
http://buildandcrash.blogspot.com/2015/05/tau-labs-and-cyphy-level-up.html

Short version - New kickstarter project comes up with a mention that they're using Tau with tilted motors to achieve level flight for smooth video without a gimbal. So of course Peabody goes right to work experimenting with the idea

But the offshoot of that is he came up with some generalized code for figuring out mixers to deal with oddball frame layouts. So soon figuring out a mixer for your Spider will be as easy as plugging in a few measurements and hitting go

 

[narcolepticltd]

Bummer.

Double bummer about giving up on the spider with what Peabody did last night:
http://buildandcrash.blogspot.com/2015/05/tau-labs-and-cyphy-level-up.html

Short version - New kickstarter project comes up with a mention that they're using Tau with tilted motors to achieve level flight for smooth video without a gimbal. So of course Peabody goes right to work experimenting with the idea

But the offshoot of that is he came up with some generalized code for figuring out mixers to deal with oddball frame layouts. So soon figuring out a mixer for your Spider will be as easy as plugging in a few measurements and hitting go

mixers for tau? I just got a bit more interested

 

[zenguerilla]

Aaaand it's gone. :-/

My coworker really got into flying it. He re-built it into a standard quad though because we were having some issues with stability on the Spiderquad. He actually flew a lot more then I can because he has a big garden and apparently more time than me.

But now he got too brave and started flying in higher winds and further away. And by further I mean too far...

He also didn't follow one rule I explicitely told him: If you lose control of your copter / orientation and your first try to recover doesn't work - let it drop. A broken quad is better than a missing one.


I guess I now know why he wanted GPS with return-to-home so badly.

I don't know, it seems like GPS is more fly-away likely.

That must have been some dense property lines or terrain unless the person just kept flying further and further away until it looked like a little dot and then said... "what was I supposed to do at this point... ?"

The trick probably should have been...

"have a carefully range tested FPV or equally tested GPS RTH... or it turns out flying very far away in LOS can become very very tricky very very quickly, requiring the equivalent of high expertise and/or very good luck" ...even worse in 'populated' locations.


"I guess he didn't know..."

Just give him this ringtone to make him feel better: http://ringtones.mob.org/mp3/busy_child-4224/

 

[Balu]

@jhitesma: Yes, I just flew over the article, very interesting stuff, but too much at this time of night for me . Gotta have to take a closer look on the weekend. He also mentioned that Sparky2 is not that far out anymore. He is testing some manufactured samples at the moment.

@zenguerilla: It was an old army airport where he had a lot of space to fly without having to worry. As far as he told me he lost orientation and had the quad get pushed away by the strong winds further and further while trying to get it back... He said he had used half the size of battery than he was used to and thought it was losing altitude, but that might just have been the visual effect of it moving further away while he was giving it too much throttle to actually go down because of the less weight.

A lot faster than he expected it was only a small dot far away.

He wanted the RTH feature just as failsafe, not for the regular flying.

 

[zenguerilla]

@zenguerilla: It was an old army airport where he had a lot of space to fly without having to worry. As far as he told me he lost orientation and had the quad get pushed away by the strong winds further and further while trying to get it back... He said he had used half the size of battery than he was used to and thought it was losing altitude, but that might just have been the visual effect of it moving further away while he was giving it too much throttle to actually go down because of the less weight.

A lot faster than he expected it was only a small dot far away.

He wanted the RTH feature just as failsafe, not for the regular flying.

"Knowing is half the battle" ...then you have to act.

That should have been precautionarily preventable. But that's the trick, LOS becomes a wolf in sheep's clothing very quickly.

The pilot thinks... "safe, beginner" ...but quickly becomes triple black diamond+ difficulty, once the last couple of hundred feet or so have been ignored and still increasing, plus unaccounted wind, etc.