JUMP TO:

Shimano’s Dura-Ace & Ultegra Power Meter (2nd Gen): Technical Deep-Dive

While much of the attention of this week’s new Dura-Ace R9200 & Ultegra R8100 Di2 product launches have been focused on the wireless and 12-speed access, far less has been written on the 2nd generation power meter. Likely because unlike the rest of the group set, the power meter isn’t available yet.

Still, it remained the part of the Shimano next-gen group set story I was most interested in, given the accuracy challenges the first generation Shimano Dura-Ace power meter had (the R9100P). Even more, if Shimano could rectify those issues while concurrently expanding it to Ultegra (as they are doing now with the R8100P), then that could make for a very compelling integrated power meter option on bikes.

Thus, I spent a chunk of the morning talking power meter tech with Shimano and their power meter team. In doing so we walked through their explanation of why the 1st generation units had the accuracy issues they did, what they’ve done to fix it in the 2nd generation, and then a slew of other technology features going forward leveraging Pioneer’s power meter technologies that they acquired nearly a year and a half ago.

Now, unfortunately Shimano did not have a rideable unit at Eurobike. In fact, if you look at all reviews of the new system, you’ll notice nobody has a power-meter equipped unit. Those will likely start making their way to reviewers in October, with widescale availability in December. Thus, while I don’t have any accuracy comparison data at this point, I am going to dive into what they did different with this unit.

What Stays The Same:

P1033325

First up, some quick basics and related things about what stays the same with the R9200P from the R9100P. Oh, and a very brief primer on naming:

– Shimano R9xxxx series = Dura-Ace crankset without power meter
– Shimano R9xxxxP series = Dura-Ace crankset with power meter
– Shimano R8xxxx series = Ultegra crankset without power meter
– Shimano R8xxxxP series = Ultegra crankset with power meter

And then from a generational decoder standpoint:

– Shimano R9200 = This new Dura-Ace stuff
– Shimano R9100 = The last generation Dura-Ace stuff
– Shimano R8100 = This new Ultegra stuff
– Shimano R8000 = The last generation Ultegra stuff

Got all that? Probably not, it’s OK.

Here’s what’s stayed the same:

– 300 hour battery life same as before.
– Internal rechargeable battery is identical as before
– Weight remains of the same (though, the new crank arms are about ~60g heavier than before, depending on exact chainrings on them)
– Claimed accuracy of +/- 2% remains
– ANT+ & Bluetooth Smart power broadcasting (total power/power balance/cadence)
– Magnet still required on frame for determining cadence/position of crankarm
– Active temp compensation same as before
– The strain gauges are placed in the same spot as previous as well.
– Still a dual-sided system (both left and right crank arms have sensors on them)
– Still a fingerprint magnet and impossible to take non-shimmery photos of

In terms of claimed accuracy, while officially it’s +/- 2%, Shimano says in testing they’re seeing it closer to +/- 1% to 1.5%. Obviously, that’s something I’ll dig into in testing on the road.

What’s Changed:

IMG_0106 2

Next up are the most important things – what’s changed. This is the area where Shimano focused the most on one key thing: Materials.

In my discussions with them, Shimano noted that with the previous generation unit (Gen1 on the R9100P), the power meter had to basically be ‘retrofitted’ onto the crank arms that the crank arm team had made. As a result, the materials and manufacturing process of both the R9100 and R8000 crank arms meant that power meter strain gauges had a very hard time being consistent on that material. This wasn’t just an issue for Shimano – but really everyone else. Be it Stages, 4iiii, or all companies adding power pods to the side of those two specific Shimano units (interestingly, this wasn’t an issue in previous generation Shimano crank arm sets).

Looking at the R9200/R8100 series however, Shimano says they worked as a cohesive team such that the materials and manufacturing process were designed to support power meter inclusion/addition, rather than be just ‘best effort’ as was the case on the previous design.

At one point during the discussion, one of Shimano’s leads noted that when it came to changes, they specifically “redeveloped the material of the crank arm itself to have better predictability of the strain gauges”. That’s notable, because the core challenge of the previous gen units was that the material wasn’t predictable, and thus, you saw variations from those cranksets (be it Shimano’s power meter or others).

Thus, in summary, here’s what’ changed on this unit from a hardware standpoint:

– Changed crank arm materials
– Changed crank arm manufacturing process
– Battery cap is now replaceable
– Battery is on side of spider rather than top
– Battery connector cable changed to match Di2 rear derailleur charger cable (so you only need one cable)

Also, while not changed, there’s a few quick things that are notable per se about the newer units:

– Both the Dura-Ace & Ultegra power meter are identical. Only the underlying crank arm materials differ.
– Ultegra pricing is $1,160 for a dual-sided crankset with chainrings and power meter ($315 without power meter)
– Dura-Ace pricing is $1,470 for a dual-sided crankset with chainrings and power meter ($625 without power meter)

Of course, very few consumers will be paying those prices directly. Most will buy it via bikes that are pre-built, and thus, the actual costs are far lower.

What’ll arguably be just as interesting is whether or not 3rd party power meters (e.g. Stages/4iiii/etc…) have better luck on the newer cranksets than older ones. That’s ironically something we might know sooner than we know from Shimano’s own power meter. Merely because people can go out and buy a new R9200/R8100 equipped bike today. And thus, so too can companies like Stages. Therefore it won’t be long till they start testing their pods on those crank arms.

In any case, interesting times ahead for sure.

Oh, and for lack of anywhere else to put it, here’s a quick front and back shot of both units, starting with the Dura-Ace:

IMG_0106 2 P1033324

Ultimately, there’s really no visible differences here from the previous generation, aside from the connector and pod placement. And then, the Ultegra:

IMG_0113

Virtually all the meaningful changes are under the hood…err..crank arm.

What’s Upcoming:

IMG_0120

The last bit is what’s upcoming. And for that, we have to go back to the past. In this case, what’s old is new again. Shimano will be leveraging Pioneer’s pedaling efficiency metrics going forward in their power meters. However, the exact branding and precise features aren’t fully finalized yet. Still, Shimano’s goal is to have this out at the end of the year.

There’s a few notable bits within that. Right now they’re aiming to support some aspect of the ANT+ Cycling Dynamics standard (that’s part of the ANT+ power meter profile). While the standard won’t support everything that Pioneer had, Shimano said numerous times they want to support the standards here, so they don’t end up in the same situation as Pioneer with only a handful of people using the metrics since it relied on a Pioneer bike computer.

PioneerPowerMeter

All of this will also upload to the new Shimano site. Yes, that one. Since we’re talking about that, they seemed to sorta sheepishly imply that what went down back in June was pretty much their worst nightmare. However, also a contractually obligated nightmare with Pioneer. In effect, that date was decided in a legal contract, and hell or high water, that was the end date. Both hell and high water came, and the site went down for…well…till the next month or so.

Nonetheless, it sounds like by ensuring open standards, they’ll be able to work with other players so that there’s not only no dependency on a given Shimano platform, but no lock-in with any given head unit manufacturer either. Also, that’ll probably give good reason for companies like Wahoo, Stages, and Hammerhead to fully support the cycling dynamics standard before Shimano users arrive later this year.

Wrap-Up:

I’m cautiously optimistic about accuracy potential on the new 2nd gen Shimano power meter. At no point during our conversation did Shimano shy away from the challenges and limitations they had trying to make the 1st gen unit ‘work’ on a crank arm that wasn’t designed for it. While the rest of the power meter industry had long come to the realization of this conundrum, Shimano hadn’t done so publicly prior. So for them to both explain the situation as well as how it changed going forward, that seems promising.

Also promising is their repeated insistence on following an ANT+ & Bluetooth Smart specs for advanced pedaling metrics as broadly as possible (within the limitations of the specs), which is also a good step forward. Shimano has a long history of beating to their own drum in the tech field, so trying to color within the lines would be much appreciated by both consumers and industry alike.

Ultimately, the next step is simply getting the R9200P & R8100P units into the DCR Cave and out onto the roads for more detailed testing. It’s really as simple as that. Hopefully that’ll start in October, and then be ready to release whenever Shimano confirms the product is final. As always, that’ll include both indoors and outdoor testing, in what is the typical craptastically rainy fall Netherlands weather. Ain’t no warm sunny locales here.

With that – thanks for reading!

Post a Comment

Your email address will not be published. Required fields are marked.
If you would like a profile picture, simply register at Gravatar, which works here on DCR and across the web.

You can click here to Subscribe without commenting

Add a picture

*

44 Comments

  1. seattlejp

    Cadence magnet means a hard no for me. Looking forward to see how well quarq works with the new 12 speed drivetrain.

    • While I don’t disagree with the slight annoyance, I’d love to understand the hard no part (truly curious).

      Cheers!

    • seattlejp

      Just seems strange that all other crank based power meters use accelerometers. In 2021 it feels like table stakes – requiring a magnet feels like a shortcut. Makes me wonder what other shortcuts they took. And in all honesty from an ascetics standpoint – it will just be one of those things that would always bother me. Given all of the other great options, I can’t see a reason why to use Shimanos.

    • BrianAZ

      I had the frame magnet come off twice on the old Quarq Riken. 2nd time almost crashed as my drivetrain came to an immediate halt, while I did not.

      I prefer not to have a frame magnet.

    • For all my sins I quite like the R9100p that I regularly use.

      re: the annoyance of a magnet

      My original magnet was white…which looked bad. even though only I noticed it.

      i lost the chain a couple of weeks ago…di2 went into crash mode and wouldn’t shift and then, when i got that working, there was no power. took me an hour to figure out the magnet had come off as well. cadence feeds into the power calcs

      later, it’s then a bit tricky (but entirely possible) to affix a new and fairly expensive magnet even tho the instructions seemed to say to remove the cranks.

      finally, because of the magnet, there’s no sensible way to easily swap it temporarily to another bike.

      I’m disappointed the magnet is on the new version as well, that’s one thing I hoped would go.

    • tadaka

      any chance you an add a picture of where the magnet sits on the bike? I don’t think i’ve seen any around with the riders i do see as most of them have moved to the quarq/axs groupsets.

    • this was mine when new. see if you can spot it

      link to the5krunner.com

      i have a black cover now 😉

    • tadaka

      is it the white circular one behind the crank right next to the bottom bracket?

    • Correct. You can also see it (and all it’s ugliness) in this older review picture of mine too: link to media.dcrainmaker.com

    • tadaka

      thanks! doesn’t really look nice unless it can blend in with the bike. and as indicated in the other comments, if the magnet falls out then that’ll suck if it stops recording info at that time.

  2. Hummm… looks like all those material and manufacturing issues may have something to do with the bonding process. The hollow arms are actually two pieces epoxy-bonded together. You may have heard about some pieces failing under (supposedly) high humidity conditions.
    All of this makes sense, since the bonding area may be less predictable in terms of deformation under load. It’s a guess, I have no insights that support this. But I honestly hope that they have improved this for better durability and better power meter suitability.

    • “Materials” is just code for they made it symmetric with more defined arms to the chain rings to deal with asymmetry to axial / lateral forces and coupling issues due to gages on the inner surface and correction gages. It’s why the crank gained weight too because they couldn’t rely on asymmetric geometry absorbing the weight of a defined chainring arm by attaching it to the main crank arm.

      Aluminum from a stiffness perspective from one series to another isn’t really different — not like yield / ultimate tensile strength is, fracture resistance, fatigure, etc, Stiffness and mechanical response is the same. Actually you can swap ANY metal in and it’s response will be the same scaling only by Elastic/Youngs Modulus. The bond line is very similar with the same improved consistency from the picture so from a mechanical engineering background “materials” is being used to mean “where material be at” which means Geometry.

      When you’ve studied and invented tech used on these things you get a bit peaved with marketing wank.

    • Col W

      Absolutely agree – spot on post!!

      This is all about symmetry, and nothing to do with ‘materials’

    • Dave Lusty

      In their defence, it’s a much simpler conversation to say “materials innit” than a lengthy explanation of how they changed the shape and why that might work. The general public will hear “materials” and accept it’s probably fixed, whereas design is a bit wooly and leaves it open for Internet debate and armchair engineers to paint a bad picture. It’s bad enough with all the nutters suggesting these cranks fall apart all the time, despite evidence that of the huge number of units sold only a few have actually failed. They’re darned if they do and darned if they don’t, so may as well keep the conversation short and snappy.

    • Gabriel Vargas

      Hey Keith, nice insight. Thanks.

  3. Tyler

    I’m also a bit disappointed with the magnets. Did they give any indication if it will be similar to the old Pioneer units where a magnet is required for advanced metrics, but you can run it without the magnets for basic power? If so that would make the magnets optional and everyone can choose advanced metrics or aesthetics. Curious if you know more?

  4. Felipe

    Could Shimano just open the Pioneer standards and just used PM’s that worked really well?

  5. Stuart

    “As a result, the materials and manufacturing process of both the R9100 and R8100 crank arms”

    R8000, surely? (ie: possible typo alert.)

    Looking forward to the final review, though given that I already have Quarq cranks on my two main bikes (not my commuter, I can’t justify over a thousand bucks Australian on a power meter for the commuter) I’m not likely to buy any time soon.

  6. Alan

    I hope that Shimano corrected the issue with 9000/8000 crank arms separating at the seem and cracking. There were enough cases just in my little group that it probably should have been a recall.

    I’m also surprised and disappointed that Shimano didn’t develop a pedal-based power meter. Their platform is so popular that Garmin developed the Rally version.

  7. Tom

    Will these be compatible with R9100/R8000? Obviously there’s a limit to how interchangeable parts can be from a 12-speed and an 11-speed groupset, but if I’m running Ultegra R8000 and these power meters work well, is there any reason I couldn’t swap out the R8000 for R8100 cranks?

  8. Joseph

    Thanks Ray! Great deep dive without the unit. Love your approach to this. Can’t wait unit you test the unit.

  9. Davie

    All this talk of availability in October is hilarious. I needed a good laugh so many thanks Shimano. I don’t own any SRAM but every time a new drivetrain is announced, I get emails same day from online shops saying it is in their warehouses for immediate delivery. It’s just too hard for Shimano.

    • Pavel Vishnyakov

      To be fair – Ray was talking about power-equipped sets when he mentioned October availability. I’ve checked a local bike shop and they have regular (non-power) Dura-Ace and Ultegra groupsets “available next week” but there’s no mention of the ones with integrated power meter.

    • Correct. I specifically noted December availability for power meters to consumers, and October is likely for when I’ll have hands-on with one (power meters).

      As Pavel noted, non-power meter units are out there basically immediately.

      As for SRAM, it’s honestly not much different. One only need to look at the SRAM RIVAL power meter-equipped crankset earlier this year to see availability was identical to how Shimano is doing this.

  10. Matthias Wolf

    “materials and manufacturing process of both the R9100 and R8100 crank arms meant that power meter strain gauges had a very hard time being consistent on that material.”

    Hi Ray, it should say “R8000” here, because previous version

    :O)

  11. Oliver

    That pricing is absolutely brutal. Obviously very few are going to buy it at that price and that weight for custom builds. But it’s a real kick in the balls for those buying a complete bike with non-P and wanting to then swap out to P later – keeping it Shimano … that doesn’t seem like an option for any except those with the deepest pockets.

    On another note, any news on if new Ultegra or DA will work with 11s cassettes? Those 12s cassettes are going to be like hen’s teeth for a while.

  12. chris benten

    I thought the whole “problem” with 9100P/8100P is/was the asymmetric spider arms. Has that question been resolved?

  13. Don Rhummy

    Ray, when you get it can you please test the following types of scenarios?

    1. Changing temperatures (indoor to outdoor) and heading off riding quickly

    2. riding in cobbles or bumpy roads

    And other common but not smooth situations

  14. Ian S

    Thanks Ray, good review. So if its the crank materials and manufacturing process that caused gen 1 to be flaky, shouldn’t that have also been seen as an issue on 4iii and stages?

    What am I missing?

  15. ROBERT STEINBERG

    I WANT TO HELP. MAKE THE DONATION SITE MORE USER FRIENDLY

  16. GLT

    Shimano can be inscrutable at times but all the recent information about the new groups seems promising. I do still hope they have some quirky & unexpected things going on. Conventional counter attacks against SRAM work to.

  17. Gescheidle Chris

    Great article!! Looking forward to your review from trying it on the road!

  18. Martin Winson

    From the video it looks like Shimano have developed a “STAIN GUAGE.” Is this to measure your skid marks or perhaps it is just one of those predictive text mishaps?

  19. Gergedan

    I think killing one of the best bike computer was the biggest mistake by Shimano. I was so happy with my CA600 performance especially with its easy configurable screens.

    I transferred my data to new Shimano web site but it is still not clear if I think upload my ride data automatically. Does anyoine have insights on that?

  20. Kirk Volk

    Ray, are you still linked up with Clever Training? I am making a big purchase and wanted you to benefit.

    • Hi Kirk-

      No, we parted ways about a year ago now. These days using the links on the side for CompetitiveCyclist.com, Backcountry.com, or REI.com are the best bets to support the site.

      I appreciate thinking of me!

      Cheers!

  21. Dave Mohr

    So will the K-Edge chain catcher with magnet work for this and be a ‘fix all” ??

  22. “the materials and manufacturing process of both the R9100 and [b]R8900[/b] crank arms”

    Ray, I think you’ve fixed a typo there but replaced it with another typo. (Done it myself; every journo’s nightmare)

  23. Good morning!

    Can you post a closeup of the open battery charge point cover and how it hinges?
    Also, is it still possible please, to fasten the magnet to the frame without removing the crankset? it used to be possible to do that through the gap now occupied by the battery charge point.

  24. Pavel

    I just bought 4iiii PM dual sided Dura ace 9100. I was not aware of accuracy issue. What exactly is the issue? are the measurements inaccurate all the time or only in specific situations? thanks.