Gimbal Not Centered

[User Name]

For what it's worth, I just received my replacement gimbal and, unlike the original, this one faces near dead ahead. Maybe 1-2 degrees to the right at most.

If anyone else is encountering this issue, the best solution is to contact 3DR for a repair or replacement. It's clearly not normal and impacts framing during orbits. Can't hurt to try calibrations, but they never worked for mine. The new gimbal is dead silent too, where the original made a humming sound at certain points while I pitched the camera up/down.

I had placed my order for the Solo and gimbal immediately upon the gimbal's release, so this was probably just a case of early teething problems mass producing a new product, which is understandable. In the end, 3DR stands behind it's product, and that matters much more than the minor hassle of having to RMA a component.

 

[pyrate]

or just fly it
Mine does that from time to time, and I never have legs in my shots
one time it initializes not straight next time it does

Never an issue
Unless you just enjoy returning things
I had a roommate that could not stand paintings tilting

 

[User Name]

or just fly it
Mine does that from time to time, and I never have legs in my shots
one time it initializes not straight next time it does

Never an issue
Unless you just enjoy returning things
I had a roommate that could not stand paintings tilting

You an Ian make a hell of a team!

I'm telling you, honestly, my gimbal faced 10-13 degrees to the right since purchase. It *never* faced directly ahead. And this makes it *impossible* to center a subject in an orbit.

So despite your denial, it was, most certainly, an issue. Given that the replacement gimbal seems not to suffer from this issue, and given that 3DR techs took my information and determined a replacement was warranted, I'm at a loss to explain why you seem to suggest it was unnecessary.

And no, I don't like returning things unless absolutely necessary. I waited 4 months to seek a replacement in the hopes that a simple firmware update might fix it. Your presumptuousness regarding my propensity to seek an unnecessary replacement is apparent.

If you happen to break your leg... please don't be a baby and seek medical attention. After all, my leg has never given me any problems.

 

[pyrate]

OCD much?
Ian knows his stuff, a lot more than the feelings you keep bringing up
So if that was some sort of Dig, I would chose that team any day

 

[User Name]

OCD much?
Ian knows his stuff, a lot more than the feelings you keep bringing up
So if that was some sort of Dig, I would chose that team any day

Too funny.

OCD, questions the person who insists there isn't a problem, despite visual proof, the determination from 3DR techs that the component was faulty, and the fact that a replacement doesn't suffer from the issue.

You just can't let it go... that, right there, is the definition of OCD.

 

[Ian [P13]]

It's software, the EKF updates from your own Solo in flight, if you'd have been flying your Solo instead of looking at it on the ground whilst posting on the internet, the gimbal would have tuned itself and pointed straight.

The gimbal works out its own offsets. You could have simply done a system restore, updated the software, recalibrated and flown in the worse case scenario. If you'd have done all that, then still had a problem then fair enough.

Could someone pass me some crayons as I don't know how many times this needs repeating.

 

[Joe Horn]

It's software, the EKF updates from your own Solo in flight, if you'd have been flying your Solo instead of looking at it on the ground whilst posting on the internet, the gimbal would have tuned itself and pointed straight.

The gimbal works out its own offsets. You could have simply done a system restore, updated the software, recalibrated and flown in the worse case scenario.

Could someone pass me some crayons as I don't know how many times this needs repeating.

Although what you say has apparently worked perfectly well for you and for Pyrate, it did NOT work (believe me, we tried) for me or for Ian. I'm still waiting for you to prove your point by uploading an off-center Solo orbit video which you claim is easy to make, and which I claim can only be made by a defective gimbal. By the way, I just received an RMA number from 3DR to return my gimbal, since they agreed that my gimbal is defective. They apparently disagree with you and Pyrate, and agree with Ian and me.

 

[Ian [P13]]

I've spoken to 3DR about this in the last few days, I've also posted Jason Short's post on the working of the software, another 3DR staff member has posted the same on the Facebook group.

Each gimbal tunes itself in flight, and uses the GPS and compass to set its direction so that it points straight. So if you have a gimbal off in flight do a compass calibration and retry.

 

[Joe Horn]

I've spoken to 3DR about this in the last few days, I've also posted Jason Short's post on the working of the software, another 3DR staff member has posted the same on the Facebook group.

Each gimbal tunes itself in flight, and uses the GPS and compass to set its direction so that it points straight. So if you have a gimbal off in flight do a compass calibration and retry.

Excellent advice. HOWEVER, if that doesn't work (as happened with both "User Name" and me), and if you call 3DR Customer Service and they determine that your gimbal is defectively and permanently off-center (as happened with both "User Name" and me), then they will replace it if it's under warranty (as they did for "User Name" and are currently doing for me).

I *still* await a counterexample video (of an off-center Solo Orbit Smart Shot video). I don't think you are able to make one (without cropping it in an editor).

 

[User Name]

Yup, it goes without saying that all possibilities should be ruled out, starting with firmware updates, calibrations and time spent flying, before seeking a replacement. But, as has happened with several of us, doing exactly these things, many times, over several months, has failed to "work the problem out".

In these cases, your best course of action is to contact support and provide them the relevant details. In my case, all it took was the above video showing the issue. They didn't even ask me to do firmware updates or re-calibrate (although I assured them I had done exactly that numerous times). They sent a shipping label, I returned the gimbal and they sent a new one that, so far, appears to function perfectly.

As for the functioning of the gimbal, this is the information I received from Aram, an engineer at 3D Robotics:

Q - Is there any function to compensate for gimbal rotation when performing orbits?
A - The Solo orients its airframe to the target point, and the Gimbal maintains the frame in place using the tilt control. I believe there's no additional compensation

Q - It has been suggested that in smart shots, Solo becomes the slave to the gimbal and updates it's orientation based on that of the gimbal.
A - it is the flight controller that tells the gimbal what to do

As for certain people's expertise, it speaks loudly to one's credibility when they repeatedly refuse to answer direct questions and just keep repeating the same non-answer like a broken record, totally ignoring the repeated assurance that their suggestions have proven ineffective.

 

[Ian [P13]]

Support don't write the code or design the hardware. Jason Short actually writes smart moves on the Solo and is one of the admins of the Solo Mod Club.

I already said support don't know the Solo as well as the actual tech guys as much as anything as the software is always changing.

Jason's information overrides anything you can post from a support guy.

As I said 3DR will RMA as the gear isn't numpty proof so it's often easy to swap things out.

 

[User Name]

As I said, a broken record.

 

[Joe Horn]

View attachment 2367 Support don't write the code or design the hardware. Jason Short actually writes smart moves on the Solo and is one of the admins of the Solo Mod Club. ... Jason's information overrides anything you can post from a support guy.

That's certainly correct; that's how the gimbal is SUPPOSED to work when it's working correctly. But are you saying that gimbals CANNOT EVER have a manufacturing defect that DEFEATS Jason's code? Either that IS what you're saying (strange, since NOTHING is immune to manufacturing defects) OR that's NOT what you're saying (in which case, you must admit that Ian and I *might* have had gimbals that were permanently off-center due to a manufacturing defect).

AN EASY WAY TO PROVE ME WRONG: You claimed that a Solo Orbit smart shot can have its target off-center in the video if not framed correctly before the orbit is begun. Oh yeah? DO IT. Post a Solo Orbit video (not cropped in post!) that has its target obviously off-center (like my balloon video above). I say you can't do it with a good gimbal. Are you up for the challenge? Or will we just see more bloviating that doesn't address the topic of DEFECTIVE gimbals?

It WOULD be interesting to ask Jason Short whether he believes his code to be so awesome that it will always compensate for every possible manufacturing defect. Wanna guess his reply? But arguments from authority are never as strong as direct evidence, so I eagerly await your video.

 

[Ian [P13]]

I came into this thread on the comment "And just for the record, the gimbal shouldn't point right at all." that is false and it's explained why above, it's not a fault to have the gimbal pointing in a direction on the ground, or if the calibrations haven't been run.

There was an argument being made pointing in a direction at power up is a fault for RMA, it is not. It was then argued the Solo was incapable of doing an orbit correctly, then when I posted an example of it doing so, it was made out that Solo was the exception not the rule.

User Name is still arguing on this very page because he argued from the outset the gimbal was dumb, was simply keeping itself stabilized based off the frame and that the Pixhawk is controlling the Smart Shots when it is the SoloLink.

My argument has been:

1) On powering on the Solo Gimbal is in low torque mode and is getting no inertial information so will likely not point straight forward. It will point forward once in flight.
2) The Solo software and hardware is capable of an orbit and there is no inherent fault in the platform preventing this.

 

[User Name]

I came into this thread on the comment "And just for the record, the gimbal shouldn't point right at all."

I suppose I have to admit an enjoyment of arguing.

On the point about the gimbal not pointing directly ahead, I'll readily admit I was mistaken. It is, apparently, intentional to point the gimbal *slightly* to the right to avoid the left leg entering the field of view.

But we've seen many examples of the gimbal facing significantly more than slightly right. My replacement gimbal faces about 1.5 degrees to the right. I'm inclined to believe that this is normal. And even though my former gimbal's 13 degree yaw was certainly excessive, even a 5 or 6 degree yaw will effect framing when orbiting.

Torque at rest vs. flying is irrelevant, since the examples we've seen show no difference - a gimbal facing more than a degree or two on the ground, still faces more than a degree or two in the air, as evidenced by the footage of orbits showing that the gimbal is NOT facing the orbit point.

And anyone who cares to read my posts can clearly see that I've never referred to the gimbal as dumb, despite the repeated claims to the contrary from Ian. I simply don't think it's as intelligent as Ian suspects, in the way he suspects. By my, Joe and other's experience, it does not, nor does the Solo's flight controller, automatically compensate for a moderate to severe yaw.

Secondly, there is no need for it to be intelligent to the point where it's specifically relied upon in determining how to the Solo is controlled. The Pixhawk platform has all the smarts needed to fly the airframe and interactively yaw while following a flight path. Any Pixhawk based multirotor can maintain a pre-programmed flight path while yawng in arbitrary directions. It's called Region of Interest. The only thing the gimbal need do is stabalize the camera (through the use of EKF algorithms common to many motion controlled systems) and keep the legs out of the field of view, likely through simply limits on yawing.

There is simply no need for it to do anything beyond that. That doesn't mean it's dumb. Yet I guarantee all Ian will get out of the above two paragraphs is "The Solo Gimbal is dumb". Unfortunately, I have no inherent ability to penetrate such stubbornness.

And just to, once again, demonstrate the unwillingness to validate a position, I'll again ask Ian to explain how the orbit point he chose in Joe's balloon video could possibly be correct given that it objectively and without question targets a moving point. I have little doubt this challenge will, again, be utterly ignored.

 

[Ian [P13]]

Let's draw a line under the conversation here.

My concern when I posted was people could be mistaken into thinking any offset at power on was a fault, when in fact it is the norm.

I have to add, apparently the gimbal with the current version of software in flight doesn't add an extra tilt due to the left leg being nearer, that's a misconception, the reason this misconception is commonplace is that code was written to do that but it has not been committed yet, it's possibly not needed. So the gimbal will go straight in flight, however that is of course reliant on the sensors being correct such as the compass.

I wasn't saying the gimbal tilts to frame a subject, I was trying to say the gimbal uses its own logic to work out where forward is, rather than simply assuming the airframe is pointing in the correct direction. From a logic point of view the gimbal is flying for the smart moves and the body is yawing as gimbal doesn't have a 360 degree axis. So we're probably arguing the same point to a large degree without realising it.

I would add the only gimbals that work well with the Pixhawk region of interest are those that are intelligent such as the SToRM32 which is MAVLink too, so it's an intelligent gimbal like the one in the Solo.

The point I was making about framing was even if you had a gimbal pointing forward, if the framing was wrong at the start it would be wrong throughout, so the only way to demonstrate the point well if a gimbal is misaligned is to photograph it from below in flight or ensure the subject if framed central at the start of the move, then it will be obvious when viewed that there is an issue.