VTail frame

[teslahed]

Ive noticed this phenomenon before and a couple of other member s have mentioned it, but Im interested in why it works. Is it a chemical reaction, or are you just creating a larger surface area with the solder on the tip.  I would be interested to find out.

My guess - once you coat the soldering iron tip in molten solder it makes better thermal contact with the solder you are trying to melt and so more heat is transferred and the unleaded stuff starts to go.

I don';t think it';s a chemical reaction - just better heat transfer. But leaded solder will melt onto your soldering iron tip better than non leaded does which is why this is the stuff you want to use for this process.

Once i get the lead free stuff to start melting i wipe as much as i can off the pads i';m soldering to with the iron and then replace it with leaded solder. 60/40 probably ends up as 40/40/whatever the old stuff is but that';s still better than no lead.

[kilby]

We where taught that the molten solder gives a better transfer of heat

[barneyg]

I wonder how the Vtail would fly with the motors angle 15 degree foward

as in put a small spacer under the side of the motor that is closest to the back of the frame so in effect when the model is level to the ground it would actually be moving forwards ?

[kilby]

You also need to mount the flight controller at the same angle (according to the guy who did the real design work on the tbs Gemini)

[teslahed]

I';ve heard other people say it doesn';t matter. Worth experimenting for yourself if you want to be sure.

[kilby]

I';ve heard other people say it doesn';t matter. Worth experimenting for yourself if you want to be sure.

Thinking about it, it shouldn';t actually matter as all you need to do is calibrate the flight controller when it';s level.

But then such seemingly logical things have defeated me in the past

[barneyg]

I';ve started up a blog about my making exploits including the latest episode of the development of this quad.

http://blog.h4ckl4b.co.uk/

but in short it is now flying without any major oscillations but it wants to pitch forward quite a lot

[attach=3]
[attach=2]
[attach=1]

[kilby]

Are you using the built in vtail mix or using your own?

[barneyg]

Are you using the built in vtail mix or using your own?

I used my own based on some info from the multiwii forums and rcgroups but I';ve reset the Naze to defaults and then redone the config from scratch but with a looptime of 2500us.  I just need some time to do some more test flights

[barneyg]

hmmm ... quick update ... tried with stock pids using both the quadx and vtail4 mixes the results :

vtail4 wobbles about all over the place.
quadx lifts smoothly and has no wobble but spins like a top on speed.

[barneyg]

Ah ha ah ha ah ha ...


While helping the wife with some TCL coding (of which I have exactly zero prior experience) I footered with the mixing.

The basic VTAIL4 mix is :

Code Select Expand


Motor Thr Roll Pitch Yaw
#1: 1.000 0.000 1.000 1.000
#2: 1.000 -1.000 -1.000 0.000
#3: 1.000 0.000 1.000 -1.000
#4: 1.000 1.000 -1.000 0.000

(cw)4      2(ccw)
(ccw)1   3(cw)


While the QUADX mix is  :

Code Select Expand


Motor Thr Roll Pitch Yaw
#1: 1.000 -1.000 1.000 -1.000
#2: 1.000 -1.000 -1.000 1.000
#3: 1.000 1.000 1.000 1.000
#4: 1.000 1.000 -1.000 -1.000

(cw)4     2(ccw)
(ccw)3     1(cw)


Part to the problem with both of these is they expect a motor rotation of :

[code]
cw     ccw
ccw     cw


while I';ve wired mine up so that the torque and thrust from the rear motors work together which is the reverse.

Code Select Expand


ccw     cw
cw     ccw


Also the motor order on my quad lines up with that of the QUADX rather than the VTAIL4 mix.

so what I';ve come up with is the following :

Code Select Expand


CUSTOM
Motor Thr Roll Pitch Yaw
#1: 1.000 -1.000 1.000 0.500
#2: 1.000 -0.866 -0.866 -1.000
#3: 1.000 1.000 1.000 -0.500
#4: 1.000 0.866 -0.866 1.000
Sanity check: OK OK OK

(ccw)4      2(cw)
   (cw)3   1(ccw)



with the fractional values based on some basic trig calcs :

[attach=1]

where the Black line is a thrust factor of 1 (ie 100%) and theta is 60 degrees.

Does this seem right ?  Testing will be performed when I get a spare minute or 2 of daylight.

[kilby]

I think I would have been very lazy and just rewired the motors

[barneyg]

Given the way this thing is put together rewiring the motors is not the lazy way 

[barneyg]

Ok I';ve been thinking about this and I think I';m coming at this wrong ...

from the code  :

Code Select Expand


motor[i] = rcCommand[THROTTLE] * currentMixer[i].throttle + axisPID[PITCH] * currentMixer[i].pitch + axisPID[ROLL] * currentMixer[i].roll + -cfg.yaw_direction * axisPID[YAW] * currentMixer[i].yaw;


so essentially the mix is a combined PID modifier for each axis.  In the case of a vtail with the arms at 30 degrees from horizontal you could reduce the yaw component by 50% since they produce 50% of their thrust horizontally.   So that bit makes sense though leaving them at +-1 may be "better" if you was some major yaw action.

The Pitch component I think should be left at +-1 aside from anything else it doesn';t actually pass the sanity check - I made a cut and paste error - as the mix values for all the motors should be a zero sum.

The Roll component is interesting since as the motors are producing 50% of their thrust diametrically opposed to each other you the Official Vtail mix uses 0 for the rear motors which I think is what causes the oscillations as you';ve essentially dropped the Roll PID to 0 ... I';m not sure what to do here.

I think in general what is needed to compensate for the reduced thrust from the rear motors is to actually modify the throttle value of the front motors to sin (theta), which in my case is 0.866, and not modify the calculated axis pids.

[barneyg]

I';d got the PID';s tweaked, the buzzer attached and vbat set up on my Naze32.  It was a bit windy but I thought what the heck.

Did a couple of laps round the field ... the felt a bit sluggish but it was steady in the sky.  Brought her in for a quick checkover and to slightly lower the level P term and realised the rates were set to 0.5 so I banged them up to 1.0 like I had on MultiWii on my f450.  Now that was better ... lets see how Horizon mode handles a flip.  Bang over on the aileron and it had done 2 flips before I had a chance to let go.  hmmmmm ... ok lets try that again.  This time I lost count at 5 when all hell broke loose ... one of the front arms looks like it worked loose and it plummeted to the ground with bits of props and motor mounts and centre plates being spewed out behind it.  I cut the motors and could do nothing but watch as it dropped from on high ...

[attach=1]