USB Sim + OS X El Capitan: Missing switches

Hi,

when connecting my VBC to my Mac in USB Sim mode I can't seem to use most of the switches in e.g. neXt :-( Note that I'm running OS X El Capitan. This is how I assigned my switches in VBC:

Bankswitch 3
Motor 4
Buddy 6
Par Lock 1
Safety On
Option 1 2 Rescue
Option 2 5
Option 3 5

It's probably not important but I physically exchanged switches 1 and 2 so that 2 is the spring-loaded one. I didn't notice any effect on the joystick mode anyway.

In USB Sim mode VBC advertises the gimbal controls as well as a total of 16 (!) 2-position buttons to OS X. However, only buttons 1 through 6 will ever change their values! Buttons 1 and 2 together emulate the 3 states of the motor switch. Similarly buttons 3 and 4 reflect the bank switch and buttons 5 and 6 emulate the buddy switch. That's it, no more live switch states on my Mac

I checked this with neXt directly but also with Mathematica, which allows you to dynamically probe the raw states of all sorts of input controllers connected to the Mac. I also checked the raw joystick signals using Ubuntu Linux running as a Parallels virtual machine. Same picture every time.

I'd love to use the option 1 switch in neXt, too, in order to train my reflexes for pulling rescue. Is there any way to get more switches to work?

Cheers,
Oliver

Hi Oliver,

please be a little patient, we are working on the Sierra issue anyway, and will also try and expand the number of channels/switches.

Seems we are close to a solution

Cheers

Eddi

Born to fly ...
forced to work.

Sure, I'll be patient! It's great to hear you're about to kill two pigs with one bird

Thanks!

Just wanted to report that with the latest software update I can successfully use all assigned switches in neXt… including rescue

One thing I noticed after the update: the tail now has a slight, but noticeable drift in certain pitch ranges. I'm not sure whether this has anything to do with the update but I wanted to ask anyway. I tried to calibrate the left gimbal a couple of times already and it turns out that the raw value for zero tail depends quite a bit on pitch. At full negative it reads -310, at zero pitch -330 and at full positive -350, give or take. When setting the deadband to 0 in neXt, I'm observing a slight tail drift in inverted hover, while upright hover is fine. Increasing the deadband to 1 kills the tail drift and naturally a bit of overall precision, too. The problem is that I can feel the change on the cyclic stick which would be perfect as is with zero deadband Should I be worried about the left gimbal or is this normal wear and tear? I only bought the VControl some two months ago…

Cheers,
Oliver

Hi Oliver,

sorry for the late reply, holiday season ...

From my experience in Service, I would not be worried about the gimbals, we have had no complaints about them since the release of VBar Control.

We have a resolution of about +/– 12,000 units, both directions. So an offset of 40 overall (between -310 and -350) means an offset of 0.33 %

On the two VBCs I have on my desk it's like 30 offset and like 60 offset when I operate the corresponding collective stick, and similar but of course different values on the other gimbal.

In the VBar, there is a slight deadband to compensate for this, and in a sim, I would also set it, if only to the minimum.

Kind regards

Eddi

Born to fly ...
forced to work.

Hi Eddi,

thanks for expanding on this!

Interestingly enough, I never observed any tail drift when flying the real thing, and neither did I miss any cyclic precision when hovering it. The VBar must be hitting a sweetspot with its deadband Is there any way to know how large it is? Presumably, the minimum deadband of 1 (out of 100) in neXt is way too coarse…

Out of curiosity: what's the advantage of applying the deadband at the FBL (or sim), compared to the VControl?

Cheers,
Oliver

Hi Oliver,

afaik the default dead band in the VBar NEO is 6 (out of 1,600 means .375 %, or even out of 3,200 (1,600 both directions), which would mean 0,18 % ... not sure here), which about just covers what the gimbals require.

The dead band can only be set in the device which operates with the input signals (e.g. the VBar), which means it is set to just not react on given inherent deviations. Usually you would set it as low as possible without noticing the drift effect.

If only servos are connected to e.g. a normal receiver, what you would notice is servo noise (the servo jittering with the slight fluctuations).
With a rate control device like the VBar, this noise can lead to drift, because even these slight inputs are interpreted as control inputs, so you need a suitable dead band.

Kind regards

Eddi

Born to fly ...
forced to work.

Hi Eddi,

> afaik the default dead band in the VBar NEO is 6
> (out of 1,600 means .375 %, or even out of 3,200
> (1,600 both directions), which would mean 0,18 %
> ... not sure here), which about just covers what
> the gimbals require.

Good to know, thanks!


> The dead band can only be set in the device which
> operates with the input signals (e.g. the VBar),
> which means it is set to just not react on given
> inherent deviations. Usually you would set it as
> low as possible without noticing the drift
> effect.
>
> If only servos are connected to e.g. a normal
> receiver, what you would notice is servo noise
> (the servo jittering with the slight
> fluctuations).
> With a rate control device like the VBar, this
> noise can lead to drift, because even these slight
> inputs are interpreted as control inputs, so you
> need a suitable dead band.

Sure thing. For some reason I thought the jitter would be eliminated directly at the source. That is, the VControl itself would supply a "cleaned" output around center stick and the amount of deadband required would be part of the radio's calibration procedure. Would there be any disadvantage with that?

Cheers,
Oliver