TheHindmost

LOL, true 'nuff...although once you start pulling any Gs, where the object believes the "ground" is can change wildly. In an airplane during a well-executed barrel roll, an "object with mass" will believe the "ground" is the bottom of the airplane for the entire maneuver, even when the airplane is upside down.

Will

intrim

even when you are upside down and level you will have to different temps remember when about 15 years ago everyone was buying those silver ball anti gravity arches where the ball would float in the air If you dont have the two different readings to compare the servo electronics wont work The ball looks like it has stopped motionless in mid air but in reality it is being bounced up and down several times a second, riding on an infrared beam as the ball moves upward against the beam it actually starts to cut off the beam. now the servos know a couple of things like where the ball actually is and that it is getting too high. the servos cut power to magnetic field and now the ball starts downward. the lower it gets the stronger the beam of light gets as the curveature of the ball is no longer in the way. the beam stregnthens to a point where now the servo system knows again where exactly the ball is and that it is now too low.if the ball could physically come to a stop in mid air. the servos would nolonger be in controll and the ball would either fall on the ground or stick to the underside of the arch. so you will always have two temps to compare against even if level. I bet as far as computerized control systems go, there is no such thing as level. just too high or too low being broken down to such a fine degree and samples taken several hundred times a second . it just looks level Releatively speaking. This same princible sky talked about when he sets the trims so his DF will stay in front of him with constant forward stick. you dont have to worry about wiggling back and forth trying to hover. just some stick pressure and it looks like its hovering but the draganflyer thinks it is flying backwards all the time. in other words you are not letting the DF decide where it wants to go and then countering it. You are the one in charge so you decide where you want to fly and just cut your hovering workload 50% thats a load off my mind now i can concentrate on aspects of becoming a better pilot


michael

intrim

hindmost

i thought i read somewhere if you do the manuver the correct way you can put a bowl of water on the floor and when your done you havent spilled a drop

mike

TheHindmost

Yep...Another version of that is when the FAA examiner puts a cup of coffee on the glare shield during a checkride. That never happened to me, thank goodness...

Will

targit

Ok, so I set my DV camcorder up to do a little video to show my mate this thing flying, and guess what? she dropped from the sky like before, but this time I got it on film!

[link=http://media.putfile.com/DropOut-42]in flight failure[/link]

TheHindmost

Targit, I can't get your video to work on either my older Win98 box or my fully-updated WinXP box. The content type is "dsplayer, video/mpeg." Both boxes try to play it with RealPlayer, which then claims it needs a software update which isn't available...I tried to look at it directly using VideoLan Client, but that didn't work either...

Will

targit

Just says "connecting" for me, thats just super isn't it?dunno what i did wrong here, have to start again i spect :P


I left it "connecting" for a while and then it worked, so not sure what the problem is, its a "dvd quality" .mpg file

Koop

Well guys, I have had this same issue for quite some time now. Remember my thread labeled DF Wobbles? Well I have sent my board in 3 times now and I still have a glitch. The good part is that when I send it in again there might not be any possible way to fix it. I was flying in a very remote part of Florida and for some reason it did its little glitch thing a couple of times and then it decided to go completely nuts and flip itself upside down and crash on the only solid object withing 100 yards. I will post pictures of the PCB this evening for your viewing pleasure. The great thing is that the canopy and PCB took all of the impact so everything else works. It's a good thing that a CF rod didn't brake because that would have really sucked. Anyway, that's enough sarcasm for now. I'll post some pics later.

Koop

Here's the Pics.

TheHindmost

DVD-quality MPEG is a pretty high bitrate, generally at least 4 Mbs, and could be higher. Too high for the Internet, at least for most folks, if streaming. Your long "connecting" probably means it had to download most of it before it could start playing it. But at least it played for you. What program ended up playing it?

If I could grab the file itself without pissing off the folks at Putfile (who say that if I try that, they'll take away your videos for "direct linking"), I could convert it to a lower-bitrate Windows movie file easily enough...Can you convert it? If not, how big is it i? You could potentially e-mail it to me for conversion...

I'm becoming suspicious that this "power drop-out" issue is a software glitch...I'm not sure why I think that yet, given how little I know about the DF's internals, but the computer professional in me smells a bug! (I was a systems programmer/architect, UNIX environment, not Windows, for many years.)

Will

TheHindmost

Oh man, Koop, so sorry to hear about your accident and see those pics!

Since the boards are basically intact, the damage looks reparable, but still...nasty stuff. Have the folks at DFI ever commented on your wobble issue? I guess they'll just say "RF hit" in response to this "flip of death"...I'm not so sure, though.

Will

targit

Hi,
@will yeah, id did seem to d/l the whole file to play it, and its windows media player that does it, I have the xvid codec installed on all my pc's (the one I uploaded is running win2kpro - 64bitAMD media crunching machine and this little laptop is running xp-home)
so shouldn't be anything special, its 23MB, it just fitted nicely at full quality and tbh you would struggle to see what was going on if i compressed it, i had the camera sitting on a chair and the lense zoomed right out, was pure "luck" that it happened in shot at all, i included the brief flight defore the "In Flight System Failure" just to show I wasn't doing anything strange or stressfull to provoke it.

I am also begining to conclude a microcontroller type problem, I've been programming PICs for a little while now and its kind of familliar the way it just stops. I'm gonna connect up my oscilloscope tonight and take a look at the 5volt rail, the same supply feeds the camera as the microcontroller, and the cmos camera doesn't like running off the onboard power at all, soon as the motors start running the signal goes to crap, but its fine when i add a seperate battery , so I suspect a power filter problem.

Koop, my sympathies. If the IFSF (makes it sound like I know what I'm talking about) that I caught on video had happened when I was going higher or faster I could be looking at a wreck like yours. I got off light, hope it's not too expensive!

not gonna fly mine again untill I find some answer to this.

TheHindmost

Thanks, Targit...I got RealPlayer to stop trying to play .mpg files and Windows Media Player worked, after waiting for the download. Did you drop the throttle as soon as you saw it glitch? Or did it just fall all the way to the ground on its own? The one and only time I've seen a power drop-out, I was only five feet or so off the ground and it was so short that the DF recovered before it hit.

Interesting for sure...I ran my CMOS camera off the on-board power and it worked fine. Are all your motor capacitors installed/intact? Let us know what you find with the scope! Noisy power could definitely be a cause of this...

I've been reluctant to fly mine as well...But I'm going to get back to it. I've been laid up a bit (first the flu, then a dislocated shoulder--ouch!), but I'm on the mend.

Will

targit

Got side trackedin the workshop last night, no news on the noise yet, but no I didn't touch the throttle, it just fell down like the power went off , i was hoping to see if i could see if my flight lights went out, but it's too far away. I edited the swearing off the end, all you hear is the stunned silence!

TheHindmost

That seems like a different failure mode than the "power glitch" which has been discussed previously. Those have lasted significantly less than a second (like, maybe a motor update cycle or two?), whereas yours lasted long enough to hit the ground from over 10 feet (looked like). Did your fuse blow? Have you even powered it up since then?

One thing I've always wondered about the power glitch: If it's due to an actual power failure, why is the DF staying armed? Even a brief glitch should cause a reset, I'd think. So that argues against a total power failure. The problem's on all four motors, so that argues against FET problems. It really looks like the microcontroller is commanding all motors to zero thrust and then back to wherever they were, as if it mis-decoded a packet from the receiver.

I'd love to have schematics for the DF, or at least a good block diagram, but I don't think DFI will release those...

Will

targit

I'm just in my workshop, I looked at the 5v supply and its clean as. no sign of anything to frighten the horses.

However, it did need to be reset several times on the bench. I investigated further and found that if the signal from the Tx is interupted
at all, my DF resets and needs to be re - armed.
Not sure if this is how they are supposed to behave, but its not a very robust control system. My XUFO does better than this.
I'm gonna have a word with the guys at spectrolutions and see if they can tell me if I got a faulty unit , or if this is how they are supposed to behave.


OOPS!, please reasd my correction below.

TheHindmost

Oh, well that explains that, then...One RF hit and you're on the ground. As you said, not very robust, so it sounds like something's busted!

One possibility, mentioned (I think) by Sky High in some other thread: Perhaps the microcontroller is interpreting even a brief glitch as LOS and is trying to auto-land...But the saved trim settings are so strange that it just behaves badly, either flipping itself or just dropping from the sky. If that's the case, though, it means that once auto-land is initiated, it happens even if the signal is regained, which doesn't seem like good behavior, either...But it could be that bad trim settings are leading to a fault in the program and a programmed reset. At any rate, it doesn't sound right!

Will

targit

OOPS!!!!
A correction, I just tried to re-create this, and have spotted the flaw in my test, I was running the DF from a current limiting supply for my own safety. when I replaced this with the battery and secured the DF on another bench some distance away i get slighlty different results, when I hit it with a signal on the remote frequency ( blank carrier ) it goes CRAZY ape**** and tries to tear itelf from the bench
until I kill the other signal. The reason it was resetting on my current limited PSU was because it tried to pull full power, this cut the psu off vbriefly and reset the unit.
This explains the wild behavior reported by koop, but not my "drop like a stone" incident.

Phodel

Koop:

Did you have the fail safe recovery set (auto Land) before this happened? Just wondering...........I had the same issue and have had my PCB replaced twice now! Does yours go from flying fine to wobbling out of control or did it begin at take off?

One thing that Spectrolutions says is be careful not to bump the throttle stick to the bottom right while carrying the DF...........I am anal about that and the fail safe recovery being set and it still has done it to me. Go figure.....must be a defect!

Phodel

Koop

I am amazed at how many of us are experiencing similar problems. As for the auto land adjustment I always calibrate before every flight. Once my DF has its momentary loss in motor power it wants to pull in one direction when I have the ti enabled. Once I turn of the Ti it flies alright and then I have to land and recalibrate the Ti. Once I do this it flies alright once again. I still have issues with the DF wanting to wobble when maintaining a hover. I am beginning to become very upset with the DF. The ONLY time I crash is when it decides to fall from the sky. I thought that it was mostly my fault in the past due to hard landings and the like. But now I have noticed that the DF likes to develop its own personality from time to time. I really hope that this issue is corrected because I'm getting tired of sending my PCB into Spectrolutions and I'm sure they are just as tired repairing it.

Has anyone tried looking at the test points with an O Scope yet? I'm thinking about giving it a try when I get my board back. I'm very curious as to how this piece of machinery functions. I asked Mike for schematics back in the day and needless to say I was shot down. Anyway, keep up the good work guys and once I get my board back I'll try to give my two cents as well.

TheHindmost

Koop, I think you speak for a lot of us! I wish I could say the only time I crash is due to the DF falling out of the sky, but I'm still learning to fly the darn thing. And I've been reluctant to practice, reading about all these problems (especially the "flip of death" which I thought the DF was essentially immune to). I've seen the "power glitch" once, though...And of course, my one big crash was due to in-flight rotor failure. I am going to get back to flying it, though (a combination of weather, flu and a dislocated shoulder has significantly limited my outdoor flying time), and of course I'll report on problems encountered, as well as successes achieved!

I, too, would love to have schematics as well as the source code for whatever microcontrollers are involved. I can see why DFI and/or Spectrolutions (I'm still not clear on the precise relationship between those two companies) wouldn't want to release them, though. But it's really too bad...A knowledgeable DF owner could help them with field debugging. Gremlins ALWAYS crop up in the field which don't crop up in the lab.

It's clear the DF is a worthy flying machine. Sky High has had a lot of success with it, and the videos on DFI's web site speak for themselves. But a few fatal glitches is all it takes to ruin the fun, especially on something this expensive.

Will

laceycopter

Ok, after flying my SAVS for about 2 weeks now I was feeling very confident. Most of my flying is without TI, I only use it if I'm going high and shooting video. Well yesterday I had it on and was up about 150 feet, all of a sudden the DF got a Bad case of the wobbles. It got so bad I was afraid it was going to do the Flip Of Death....I turned my TI off, I don't know if that was a good idea or not it seemed like all my trims were way off, it was Very hard to control and I was lucky to get it down without a crash.
I checked everything and it seemed fine, I did some lower level flights (about 50-75 feet) with and without TI...it seemed fine.
I decided to try the same thing with one of my other heli's....well in almost the same spot it started getting a Bad tail twitch. The difference is I can adjust the gyro sensitivity from the radio and it cured the problem. Now I don't know what causes the problem in that spot but it seems that it's Not a DF problem, but some outside source causing it.The gyro's in the the DF are VERY sensitive and and may just be effected more then other heli's. I have been flying at that field for the last 2 weeks without a problem. I'm thinking that all the computer stuff or whatever makes this thing fly just feels an outside interference more then other heli's...maybe there is nothing wrong with the board. I've never had a problem with the TI off.(just won't shut it off in mid flight again)
Oh yeah I had pink shorts on....but I have ruled that out as a possible problem.
Lacey

TheHindmost

ROFLMAO!

Will

Sky High

I said it before and I'll say it again. I have got to be the luckiest DF owner around considering the vertical and Ti board on my DF that crashed has been bent twice now and this thing continues to fly just fine. Now it hasn't been bent nearly as bad as the others I have seen so maybe no connections have been broken. I think it's time that DFI or someone seriously considered creating a very lightweight spherical or halo roll cage or at least something to protect just the vertical and Ti sensor board during an impact. It wouldn't protect against shock but sure would prevent the boards from being bent. The only other option would be to use a ballistic shoot or those ballistic air bag cushions like on the Mars lander.

ericlawrence

I have built the HMX4 and own two Draganflyer VTi's. I have experienced the wobbles I am reading about. I live in the desert so I calibrate it on sand. My newer Draganflyer wobbles but not as bad as my earlier Draganflyer. I have fashioned a roll cage over the board by bending .91 wire in a hoop from one cross member to another. I connect the hoops at center over the board. This has saved me a lot of grief when the wobbles become so pronounced a crash is inevitable. Both fly great without TI. Only when I engage TI do I get the mysterius wobbles. No shade, no concrete, only sand. Alas, no grass in my part of the desert. If I did not have a new Draganflyer to compare, I would have thought is a bad board. Not so, and the folks at Draganflyer are not very helpful, suggesting I am not calibrating them properly. I fly single rotor birds without any problem. Something is wrong, confirmed by reading these posts. Perhaps they should post a "buyer beware" notation if you cannot calibrate over grass. Until I get the problem solved, or buy a patch of grass, I will continue to work towards solving the problem. Again, a roll bar not only protects the bird, it also allows for easy calibration without taking the plastic canopy on and off. Who needs it unless you are a purist who must have the canopy look - - Eric