R9100-P & R9200-P: Are Shimano PMs still suffering L-R issues?

It is pretty well documented that 1st-gen R9100-P Shimano power meters suffer from issues with L-R accuracy. Two questions I’d love answers to:

-Has Shimano continued to try to improve the 1st-gen R9100-P powermeters with firmware updates? Have they had any success?

-Presumably, the R9200-P powermeters have a better foundation to start from, with the crankarm asymmetry removed. Can the new generation of power meters be trusted for L-R data?

I seem to recall GPLama saying that they were still suffering form the issue…

That said, I ride a Stages L-R crank and I have never had an issue. My understanding (although it has been awhile since I reviewed it) was that the issue was on higher power intervals and only for brief periods. Yes, if I was worried about absolute precise data across every second of a ride or interval, I might care…but I don’t need that level of precision and I would argue most of us don’t.

YMMV, of course.

ETA - here is the GPLama initial review. TL;DW - There are still some inaccuracies, but now it appears the PM is reading high, not low. More testing is required.

1 Like

Thanks for the link to the review. I have a Quarq Dzero that works flawlessly, however, I have some injuries that are affecting power to one side. I’ve found it helpful lately to have accurate L-R power. My interest in L-R power is the reason for the thread.

1 Like

I can’t speak for earlier version of them but the last two 9100-P cranksets I’ve owned have had zero issues with accuracy. They both read within a couple percent of what my trainer shows, which is inline with others like Stages, SRM, Quarq, Favero… and, more importantly in my book, they’ve been consistent… I have had none of the weird data abnormalities that early reviews highlighted. That’s going back to mid 2020 so I can’t say if it was a firmware update or something they changed on the cranks themselves but it seems they sorted things out.

1 Like

Miller did note that because firmware updates are expected as this was an early unit, so it’s possible the situation will improve. That said, it’s not a guarantee.

As I recall, the 9100 and its contemporaries had an asymmetrical spider. This is the fundamental cause of the inaccuracy. A power meter engineer (Keith Wakeham) responded to Shane’s initial review, and said that this was fundamentally unfixable in firmware. The 9200 and 8100 groups went back to a symmetrical spider. If this doesn’t fundamentally resolve the issue, then that’s a really big problem.