Can anyone think why my mech is asymmetrically unstable in 1.05

[allmhuran]

I have some mechs. 

I have just updated KSP to 1.05 and moved the mechs across, hopefully cutting down the work I need to do for 1.1. Happily, most of them came across just fine, just a few part references needed changing.

However, on driving one I noticed that the stability of the mech is now different. I worked for a very long time in 1.04 to get the mechs as stable as possible so that it was possible to aim BD armory weapons. But now, in 1.05, this mech  (and perhaps others) has a weird asymmetrical instability. You can see a comparison between 1.04 and 1.05 in the gfy below. Note how, when the right foot is planted down, the reticle jumps vertically upwards.

The geometry of the mech is perfectly symmetrical, and it is running a perfectly symmetrical set of co-ordinates for the IR servos. I have already spent several passes trying to figure this out, which is why the mech is stripped down to the bare essentials. I simply have no idea why this would be happening. It could be a difference in infernal robotics, it could be something in KOS, it could be something about underlying KSP itself. For whatever reason, when the right foot lands the hips are jerked into a brief pitch up movement.

I'm fishing for ideas here, I have totally run out.

 

 

[Geschosskopf]

Well, I have never made a walking machine and haven't messed with IR in a long time (and it's changed a lot since then).  However, I see your problem..

The reticle jumps up on the right foot because the central box pitches upwards and back down quickly at that point in the stride.  At that instant, the left foot is also still on the ground.  So you seem to have an extra step in the servo cycle that's causing it to do a little pelvic thrust.  The issue seems to be with the hip joints, as the knees and ankles don't look to be moving.