Replies to This Discussion

Permalink Reply by Jake Stew on May 24, 2012 at 11:55pm

Has nobody bothered to actually test the APM unit out properly?  This should have been caught in beta testing, well before the 1.0 release.  Seems kind of strange that I find several bugs in the first few days of playing with it.

The failure to properly manage the throttle is a major bug and safety issue.  You really need to add a warning to the pseudo-wiki and to the documentation.

"I turned off the signal, but the 3DR product continued to accelerate towards the daycare playground on it's own accord," is exactly the sort of thing that will end up in a multi-million dollar lawsuit being won against you.

It's one thing to tell people who want features or enhancements or changes to jump in and do it themselves, but it's really difficult when the existing code was written with major design flaws and nobody seems too concerned.

• ▶ Reply

3D RoboticsPermalink Reply by Chris Anderson on May 25, 2012 at 7:47am

Jake: I will simply repeat: it was a design decision, not a bug. You may not have made the same decision. (Indeed, your contributions on this site seem to consist almost entirely of complaining that you disagree with our decisions on everything, from the wiki to forum software to PPM encoder design. Fortunately there are lots of other communities and autopilots, so you can choose another one that makes you happier.)

For the others, let me explain a little more more how the PPM encoder works. First, it runs on the 32U2 chip, not the 2560, so it's loaded at the factory and can only be changed with an AVR programmer. Second, because it's running on a different chip the same PPM encoder code applies to all platform software, from ArduCopter to ArduPlane to ArduRover, etc. 

What is the correct throttle response with the signal is lost on Ch 3 for all of these platforms? We thought that holding the last signal was the best choice, given the range of vehicles that are supported. Obviously that's the best choice for planes, so you can bring it home. It's arguably the better choice for copters, given the alternative: falling out of the sky wherever it happened to be when this happened. And it might not be the best choice for rovers, but that's the least common usage. 

 

Again, it's perfectly reasonable to run through the logic above and come up with a different strategy. That's the beauty of open source code. You can change it for your own taste. 

But it's not a bug, it was tested, and it's not stupid: it's simply a design decision made by a team who thought hard about all the options.

BTW, switching into auto or RTL gives throttle control back to the autopilot and should bring it back safely, even if the RC control is lost, assuming it has been set up properly.

• ▶ Reply

Permalink Reply by Jake Stew on May 25, 2012 at 4:28pm

When the APM gets NO THROTTLE SIGNAL it continues to put out a throttle signal.  That is a BUG.  You can't spin that any other way.

Why not just admit you have a bug and fix it?  Sounds to me like a very minor code change.  Certainly not worth trying to argue that it's supposed to be that way.  Nobody is buying it.

Sure, I complain a good deal... but I don't see anyone even trying to argue that any of my complaints/suggestions aren't valid.  I'm amazed how great the APM is in so many ways, but equally amazed that some of the most easily fixed and basic issues seem to remain.  Seems like most of the hard stuff has been done well, but the easy stuff is ignored.

• ▶ Reply

Permalink Reply by Brad Smith on May 25, 2012 at 7:49am

its a balance safety for you platform or safety for the people its all depends on what and where you fly but with the APMs ability to RTL and autoland it should be setup to bring her home and set her down , could be all we need is to set throttle high and have a range setting if it goes out of range RTL if we don't overide AUTO LAND anybody se anything wrong with that ?

• ▶ Reply

Permalink Reply by Dave on May 25, 2012 at 12:48am

Chris, the problem is not quite as rare/obscure as just a Throttle wire falling off. 

The Turnigy 9x is probably one of the most popular budget Radios.  When you fly out of range, the Turnigy Rx response is to drop all channels except 4 & 5 - equivalent to no Throttle and more.  The APM response to that is to continue flying on its merry way!

It seems to me that if you lose any other channel, OK hold the previous setting and you have a good chance of flying home.  However, to hold the Throttle, particularly if climbing, can only lead to bad things.

• ▶ Reply

Permalink Reply by Brad Smith on May 25, 2012 at 7:54am

my turnigy switches too circle then RTL if i turn off TX and its stock firmware did you "upgrade" yours this is important to know !

• ▶ Reply

Permalink Reply by Dave on May 25, 2012 at 8:10am

Through the APM?  Has yours been modified with FrSky modules?

The standard 9X does not have failsafe, so everyone says.  As I said earlier, the Rx responds to loss of Tx by freezing channels 4 & 5 and dropping the other channels to no output.

My Tx is running ER9X firmware, but I don't think that's relevant - we're talking about the Rx response to loss of Tx.

• ▶ Reply

Permalink Reply by Brad Smith on May 25, 2012 at 8:19am

my turnigy 9x is stock and has a failsafe but i,m not using the APM failsafe .i set my rx fs to switch to RTL and hold 80% throttle but im flying a plane and i only did this because i wasn't keeping up with with the the forum and didn't want a software change to do unexpected behavior 

DeveloperPermalink Reply by Andreas M. Antonopoulos on May 25, 2012 at 8:32am

The firmware does not matter - the RX is the device that does failsafe, not the TX. 

The problem with the Turnigy9x and other 9x variants are the crappy RX, not the very nice TX. 

I just confirmed on my brand new Th-9x, arrived this week: Fail-UNSAFE - loss of signal or power on TX cause all channels to retain prior settings, which in my case meant keeping FULL THROTTLE on the bench. Good thing it wasnt a real flight. Until I figure ut how to replace the radio module so I can replace the RX, this controller is GROUNDED!!

My Spectrum 5X works as expected: all channels to 0.

Permalink Reply by Brad Smith on May 25, 2012 at 10:15am

Permalink Reply by Zen on May 25, 2012 at 5:58am

I'll have to pipe in and apologize for all my previous posts as being a little hot headed, but my expectation when purchasing from diydrones is that I could build a quadcopter by following the directions provided, learn to fly it well enough to attach a camera to it, and then offer flight photography/videography services for my customers (pending future law changes) as well as personal use and fun.

Now, I realize being completely new to the world of RC means that I have a lot to learn - lot's of "pro-tips" as it were that simply do not exist in the documentation at all.  Things that only an RC expert would know.  Following the instructions led me to a perfectly flyable quad the first time - no hiccups, no adjustments.  However, I did have quite a bit of help from Mr. Finisterre who has been building these exact quads for at least a couple years that I know of.  So in that sense I was able to do exactly as I set out to do and began the "learning to fly" portion of my expectations.

Like the ~10,000+ users you mentioned I ended up with a standard Turnigy 9x, borrowed from Kevin who to this day has yet to experience an issue with them, and on my 3rd battery the quad locks throttle and flys so far away that recovery was not possible...

Now, was I pissed at the time, yes, and did we know exactly what happened? No.  Through considerable discussions both in this thread and mine I think the issue has been fleshed out enough to consider it a strong possibility that this "bug," "feature enhancement," or whatever you want to call it was a direct cause.  It could have been a Turnigy issue, it could have been a shady solder-job on my part (which wasn't really THAT bad), or it could have been a vibration that disconnected something mid-flight.  Whatever the case may be, the APM blasted off at full-throttle with no hope of recovery or shut down. (believe me, I tried everything as I was running full-tilt across the park chasing my hardware and getting laughed at by all the bystanders)

My initial gripe was that I lost 600$ in parts because of an unknown issue, now it looks like I've lost 600$ in parts because of a "bug" in the APM that will launch your shit into the stratosphere. I paid close attention to the news for about a week afterward hoping that my quad didn't crash land on some kids head, or a car, or a house - a 5lb object falling from >2000 ft could conceivably do some damage.  My next thought was what if this throttle "bug" happened while I was banking left or right?  I was in a MASSIVE field, mostly uninhabited, but there were a scattered few people, kids, parents, etc playing soccer, baseball, football, hanging out, in other parts of this field.  If that specific scenario played out my craft could have been mere feet off the ground and accelerated non-stop horizontally and taken out any number of people.

When I posted my experience people on this board jumped my shit because I didn't know a thing about RC, wasn't aware of any specific safety features as I didn't purchase my rx/tx gear, and failed to follow some undocumented pre-flight testing procedures.  As it turns out they may have been partially correct, and now I know, but that doesn't absolve whomever built and/or programmed this hardware from acknowledging the fact that there is a serious, and very repeatable, "bug" that could get someone very injured, or killed.

From my perspective as an uninitiated newb your website leads me to believe that I can build or buy, then fly.  That's what I did.  Then my quad accelerates uncontrolled and is lost with the potential to have hurt people regardless of whatever safety features I could have used.  This, to me, seems to be a pretty major concern, and it's even more disconcerting that you seem to gloss over it as though it's no big deal and should be considered a "feature enhancement."

In my very humble opinion I believe this hardware to be geared specifically to the RC electronics hobbyists and code monkeys out there, NOT to any "normal" person that happens to stumble onto your site and decides to build a quad.  Proper, and complete, documentation is very lacking, proper warnings are very lacking, and it would seem that despite following everything to the letter the hardware does things that are completely unexpected!  (even for folks that have been working with these for years)  I guess luck would have it that some major event hasn't occurred before now, and I'm glad that my incident and subsequent rant pulled all of this out of the woodwork because apparently it IS a big deal, IS repeatable, and has happened to other less vocal customers.

The bitch of this situation is now that I have another nearly assembled unit do I try to fly it or not?  I can say with 100% certainty that I am very leery of doing so.  DEFINITELY not with a stock 9x, maybe with FrSky, but more than likely I'll need to double my investment and get the Spektrum gear.  It also looks like I'll need to do a lot of extra work in researching geo-fencing, get some telemetry gear, and ensure that all the failsafes are in place.  Oh wait, even then my quad could STILL fly away uncontrolled.  Blerg.

Thanks for listening.

• ▶ Reply

DeveloperPermalink Reply by Andreas M. Antonopoulos on May 25, 2012 at 8:33am

Do not fly a copter with a 9X. The failsafe is dangerous.

I'm in the same boat with you: My brand new TH9-x is now on a shelf and will stay there until I figure out a way to fix the RX failsafe. 

• ▶ Reply

• ‹ Previous
• 1
• …
• 4
• 5
• 6
• 7
• 8
• …
• 12
• Next ›
• Page