[X] My Assistant

[Burmese]

So it looks like they may work to get that joint extended at any cost, then leave it at a suitable position and just work the other motors to position the instruments. Obviously, they will disable the code that would otherwise not allow the rover to move if it senses that the arm is extended.

[dot.dk]

Just to be sure. It is the joint that moves the arm in the azimut direction (left/right)?

[RNeuhaus]

Just to be sure. It is the joint that moves the arm in the azimut direction (left/right)?

The troubled joint is of No. 1, the ones which extend and retract the arm. The joint 2 handles the azimut which you mentioned is still workable.

Rodolfo

[Bill Harris]

I think that RNeuhaus has it wrong.

Ths problem joint is Joint 1, which is an azimuth joint.

From mars_armer's Post #92:

"The first move (apparently completed) raises the elbow toward the roof above it, in order to release a hook. The second move (not completed) is a joint 1 (azimuth) move that results in the elbow end moving forward, while the turret disengages its stow feature."

I interpret "raises the elbow" as an altitude movement, and "a joint 1 (azimuth) move" is clearly an azimuth movement.

--Bill

[mars_armer]

What do you mean when you say it is "open"?

And why would the heater be responsible for this?

I meant "open" in the electrical circuit sense. The motor has two redundant windings, and one of them now does not allow any current flow. This could happen if a wire was broken or if the brushes failed to make contact. The stuck heater causes the shoulder joint to get extra hot overnight, and the increased thermal stresses could possibly have accelerated the process that led to this failure.

Bill has it right. The shoulder is azimuth/elevation. Joint 1, which has the problem, is the azimuth, which moves the arm left/right.

One of the design rules for the arm was that if any one joint fails at any time, the arm can be posed (using the remaining joints) so that all parts are above the rover's belly pan. In principle, you could drive this way. In fact, if the arm was extended and the azimuth joint failed, you still could tolerate the failure of a second joint and still be able to raise the arm for driving. So the prospect of a deployed arm with a dead azimuth joint isn't too bad.

[odave]

...Trying to put a picture to these words. From this image on the MER Technical Data website, I marked what I think are the joints we're talking about, with Joint1 having the problem.



Doug: Not sure if this image is verboten on the board, if so, please delete to keep us out of trouble.

[djellison]

I've seen that image all over the place - it's certainly not from the should-have-that-pdf-cad-document. No problems from my end.

Doug

[mars_armer]

...Trying to put a picture to these words.  From this image on the MER Technical Data website, I marked what I think are the joints we're talking about, with Joint1 having the problem.

Well, not quite. There are some good diagrams and pictures in this document:
Actuator Development for IDD

[RNeuhaus]

I think that RNeuhaus has it wrong.

Ths problem joint is Joint 1, which is an azimuth joint.

From mars_armer's Post #92:

"The first move (apparently completed) raises the elbow toward the roof above it, in order to release a hook. The second move (not completed) is a joint 1 (azimuth) move that results in the elbow end moving forward, while the turret disengages its stow feature."

I interpret "raises the elbow" as an altitude movement, and "a joint 1 (azimuth) move" is clearly an azimuth movement.

--Bill

Oppss...the joint 1, -the shoulder- has two functions instead of one, elevation and azimuth. Sorry for the bad interpretation.

Rodolfo

[Bill Harris]

>There are some good diagrams and pictures in this document:
>Actuator Development for IDD

Good reference, mars_armer. That was good reading, there was a great deal of R&D that went into the IDD. At the end of that document was a very telling on the working life of the IDD, and this is probably the key to this problem:

===
This was deemed acceptable because the mission
duration is only 90 Martian days and the lifetime
number of output revolutions for any actuator, including
testing and ground operations, will be no more than 500
at the close of the mission. Much of the design
philosophy behind the IDD and its actuators takes this
into account: harmonic gears are often operated at
torques approaching their momentary peak torque
rating; ball bearing stresses during operations are
reasonably high (up to 1.79 MPa, or 260 ksi), possibly
subjecting the Bray lubricants to pressure cycle
degradation.
===

My first thought was that the IDD could be unstowed, the arm placed in a useful azimuth position and left alone, and the IDD operated with the elevation (elbow) actuators without latching the arm in the stowed position. But it seems that although the static loads on the IDD are such that the "detents" on the motors are adequate to lock it in use, the live (dynamic) loads are too much and the arm will shift in elevation as the Rover moves and the arm must be locked down.

We'll see though.

--Bill

[Guest_BruceMoomaw_*]

The problem is indeed with the azimuth joint -- at the moment, the arm isn't swung out far enough for the instrument head to clear the little shelf on which it sits in the stowage position. So, to contact the suface with it again, they MUST swing it out at least a little bit in azimuth -- but the MER press conference by Squyres and about four other people that I attended yesterday at the AGU Meeting made it clear that, if they can do so, they will resume using it in what might be called the "gimpy" mode, and that includes driving with the arm that way.

As for the nature of the problem, they consider it unlikely that it's a mechanical jam -- both because of the really high gear ratio (8000 to 1, albeit in two stages), and the fact that the azimuth motor isn't drawing as much power as it should when it stalls. It looks like some kind of electrical problem -- maybe with the motor itself (windings or brushes), or maybe with the power controller circuitry. No comments from them, unfortunately, on what corrective measures they may now try to take, although they made it clear that there are some.

Before we get too persnickety about this, however, we might keep in mind that the damn thing has worked perfectly for 7 times beyond its design life.

[Burmese]

Press conference yesterday, eh? Any other tidbits?

I'm guessing they will try to deploy the arm by applying a lot of power to the winding that is still fully functional. I wonder what are the consequences if the second winding suddenly starts to draw power and add in its' own torque? Like maybe spinning your tires and suddently getting full traction and overshooting?

[mars_armer]

Yes, the fix would involve providing more current and voltage to the motor, by changing parameters in the motor control electronics. They do need to consider possible ramifications if the open winding "heals". You definitely don't want to harm any electronics that are common to the other joints.

I think I misspoke when I talked about two windings. I actually now understand there are 7 windings. They are seeing about 2x normal resistance in the motor, and I don't completely understand how this is consistent with a single open winding. But I think it makes sense to the motor experts.

[Bill Harris]

Per the .pdf file on the IDD, the motors used are the Maxon RE020 brushed motor. The Maxon motors website is http://www.maxonmotor.com . There was no listing for the RE020, but I'm sure that the RE family of motors share common characteristics.

Maxon motors are also used in the driving wheels.

They appear to be serious hardware*.

--Bill




*translation= "serious hardware" means a well-designed and well-built motor.

[Guest_BruceMoomaw_*]

"Any other tidbits?"

Yep; I got pretty clear descriptions of what they intend to do with both rovers next -- and a couple of very intriguing debates among scientists which I'm trying to get a little more background on right now. (Alas, I really can't tell you guys about any of this yet.)

Also, a little more on Mars Express than has been announced in the press releases yet.