Does the coming Logo XXtreme really need HV servos?

[dahld]

Thought I'd pass on what I learned today.

I've stated in the past that with the Logo 600, running HV servos was unnecessary. I said this because I once pulled a fast one on Kyle, and lowered the BEC voltage on one of his Logo 600's before a practice session, and didn't tell him. After flying all afternoon, he couldn't tell the difference in how the heli with the lower voltage flew, compared to the helis with the higher voltage.

So we've been running the Logo XXtreme with HV servos on 8.4V, via a Western Robotics 14S 10amp BEC. But with the YGE ESC we've been running, in order to use its programming card, you have to lower the system voltage to 5.2 volts or lower (the card won't work for some reason at the higher voltages???), which we can do with the WR BEC, as it has several programmable voltages accomplished via micro dip switches on the unit. When done with the programming card, we bump the voltage back-up to 8.4V

So today, while doing some adjusting on the V-Bar via the laptop, I noticed the V-Bar was reading the system voltage at 5.2 volts. When I checked the BEC, sure enough, the dip switches were set for 5.2 volts.

The last time I used the YGE's programming card on the machine in question, was probably 30+ flights ago. So all this time, Kyle's been flying one XXtreme at 8.4V, and the other at 5.2 volts (using JR 8717HV servos), flying and practicing aggressive XFC competition type routines with both machines, and couldn't perceive a difference in how they flew.

So, when you get an XXtreme, use what you want, standard or HV, looks like it doesn't really matter.

That said, don't skimp on the system's power supply. Until there's data to prove otherwise, I'm still thinking a 10 amp capability is the safe bet.

(-: Dave

[ChrisH]

Impressive, Dave! I would think for sure at Kyle's level the difference between 5.2V and 8.4V would be noticeable, but I guess it is just too little of a change to observe!

I was going to try an HV setup on my Logo 500 I am building up, but it just seems pointless now :)

[kouasupra]

Very interesting. Thank you for sharing Dave.

[rcmarty]

I've often wondered the same thing myself.

Have you ever done any experiments between 22ms and 11ms frame rates ?

I'd be curious to see if kyle could tell the difference as i assume you would need to be at a pro level to be able to feel the difference.

Martin

[dahld]

Question: "Have you ever done any experiments between 22ms and 11ms frame rates ?"

No. But it's hard to imagine perceiving the difference of 11ms at the end of a very long control loop (eyes, brain, nervous system, muscles, transmitter, receiver, servos, mechanical slop, the heli finally moves, and then we're back to the eyes).

But, I've read that some guys say they can. I've got no reason to doubt them, and my hats off to them.

But I think more practice would help more than 11ms of frame rate (-:

(-: Dave

[wlfk]

I did some experiments a while back judging how accurately people could anticipate collisions. The experiments were computer based - used a computer monitor to display pictures and did all the calculations to account for the bottom of the screen being scanned later than the top etc. The observer had to press a button when they thought that two objects collided. It was quite a job - mostly had to be written in assembly language for accuracy - but never was published.

The outcome was that we could judge visual collisions to within about 10 milliseconds. There is also evidence that musicians can play in time with each other, with a consistency of about 5 milliseconds. That is, one pianist trying to play along to a prerecorded tune may lag by a few tens of milliseconds, but the lag will be consistent to about +/- 5 milliseconds.

Now, say I make a control input. If the next frame is imminent, the delay may be 0 milliseconds. However, if the frame had just been sent, with a frame rate of 50 frames per second it may be up to 20 milliseconds before the update is transmitted. That's considerably more than the 10 milliseconds inconsistency for the pianists.

The situation is slightly different for radio control, because the visual system has more lag than the auditory system; the lag depends on the brightness (it increases by a few tens of milliseconds around dusk, which is why flying around sunset can be alarming even when there's still enough light to see clearly) and because control inputs are typically of longer duration than the onset of a musical note, which was what the previous experiment noted.

That said, it might be possible to detect the difference between a 1/50 and 1/100 second delay in transmission. It's certainly true that 10 milliseconds is a small amount of time in comparison to visuomotor response times - typically a few hundred milliseconds - but we normally compensate these delays out pretty well and our physiology doesn't rule out the possibility.