The more wobbly ride of A5, and the observed discolouration around the welds by another user got me thinking - is it a bit of a dud perchance, that lead to the reported more recent improvements in quality control?
Not that I know of. The "smoothness" of the ride is down to the airbags (one per bogie), which are supplied by Hitachi as part of the traction system I believe. I'm not sure why one set would feel different than another, other than perhaps the number of passengers on board or the train running on a different line.
Could be the passenger thing, but was the same line so unless they have done trackwork it rules that out. Perception is also very subjective to the point where my two journeys being weeks apart could have altered my perception - the seats also seemed a tad more comfortable for example, where they would be stock also.
Had a ride on A5 earlier today and yesterday, can't say I noticed any discomfort or unusual motion with the ride. What I did notice were the weld marks along the body, which seem new as I don't remember seeing any of them when A5 first cam out and they are quite obvious. The other thing, today I was riding in N5505 (in the vestibule at the closer to the middle of the train), what I suppose were the motors sounded positively sick. On both acceleration and braking the sound they were making was, well the only think I can think of that is similar is the sound made when you extend or retract the flaps on an airliner, a quite loud almost drill like sound. Didn't seem to be affecting performance any, as we were accelerating like a bat out of hell and braking was all quite smooth also. It did seem like we were going a lot harder and faster out of stations, which seems rather distinct from my previous experience with the A-sets, where I would have said the acceleration performance displayed was fairly average. What I saw today was certainly not average and was amongst some of the better acceleration I have seen around the world (including many single deck systems).
Has A5 been tinkered with at all, perhaps playing with some different set-ups to test performance?