In an interesting bit of cooperation between a representative of the “gear reviewer” community and the nerdy, techy community, Shane Miller (GP Lama, gplama here) and Keith Wakeham (initial developer of the 4iiii power meter, kwakeham here) have come up with a convincing case that the asymmetric nature of the newer Shimano 4-bolt cranksets wreaks havoc on the accuracy of crank based powermeters using those cranksets.
My takeaway: If you’re in the market for a crank based powermeter, pick an FSA or Rotor crank, or another cranks that uses the traditional 5-bolt pattern.
My take away- continue saving money with your left only crank PM.
Yeah, maybe that’s a better one maybe. As long as you know you’re not worse than say 52%-48%
Is there an accurate, readily available test for that before you throw down the cash for a dual sided power meter?
Back to the articles, seems like a great chance for some independently funded citizen science. Think “wind tunnel shootout” but with power meters. Shane sort of hints at that at the end of his article, not a lot of existing side by side comparisons.
Yeah, maybe that’s a better one maybe. As long as you know you’re not worse than say 52%-48%
Doesn’t really matter what the balance is as long as it’s stable across your range and levels of fatigue. Now if the balance changes as you increase output for instance then that’s an issue to consider.
My take away- continue saving money with your left only crank PM.
Yeah, maybe that’s a better one maybe. As long as you know you’re not worse than say 52%-48%
Is there an accurate, readily available test for that before you throw down the cash for a dual sided power meter?
Uh… I don’t know. In my case it’s about 4 years of pedal power meter data.
I don’t know if the apps that come with the PMs allow you to set your balance ratio, probably as some advanced setting? That way you can test drive a couple of pedals from your LBS and plug the result in the app.
I think it’s a well known issue that Shimano PM’s aren’t accurate even if Shimano won’t admit their PM is crap.
This is not just Shimano PMs though. This is all dual-sided crank based power meters installed on Shimano 4-bolt cranksets. Lama lists 11 different ones in his blog post, all with consistent inaccuracies (if that doesn’t make you lol, “consistent inaccuracies”, nothing will…)
I think it’s a well known issue that Shimano PM’s aren’t accurate even if Shimano won’t admit their PM is crap.
This is not just Shimano PMs though. This is all dual-sided crank based power meters installed on Shimano 4-bolt cranksets. Lama lists 11 different ones in his blog post, all with consistent inaccuracies (if that doesn’t make you lol, “consistent inaccuracies”, nothing will…)
I believe it is mainly an issue with the latest generation of Shimano cranksets - 9100 and 8000 - which are 4-bolt but, most importantly, are asymmetric designs. I don’t think it’s the 4-bolt chainring attachment that’s the issue. It’s the fact that the crank arm is not symmetric. It’s offset and sculpted. I am pretty sure GP Lama noted that he didn’t see issues with the 9000 and 6800 cranks that were used by PM makers to construct their powermeters previously.
I don’t believe the previous generations of Shimano cranks bolted on to, say, SRM’s spider containing the strain gauges have any issues.
I think it’s a well known issue that Shimano PM’s aren’t accurate even if Shimano won’t admit their PM is crap.
This is not just Shimano PMs though. This is all dual-sided crank based power meters installed on Shimano 4-bolt cranksets. Lama lists 11 different ones in his blog post, all with consistent inaccuracies (if that doesn’t make you lol, “consistent inaccuracies”, nothing will…)
I believe it is mainly an issue with the latest generation of Shimano cranksets - 9100 and 8000 - which are 4-bolt but, most importantly, are asymmetric designs. I don’t think it’s the 4-bolt chainring attachment that’s the issue. It’s the fact that the crank arm is not symmetric. It’s offset and sculpted. I am pretty sure GP Lama noted that he didn’t see issues with the 9000 and 6800 cranks that were used by PM makers to construct their powermeters previously.
I don’t believe the previous generations of Shimano cranks bolted on to, say, SRM’s spider containing the strain gauges have any issues.
Interesting. I now see what he means with “asymmetric”. I though he meant he fact that the crankarms aren’t distributed evenly, and that that caused the issues, but now I see that the 9100 (and supposedly the 8000 as well) have an “blob” off to the side of the spider centre, under the crank arm, and that’s the asymmetry he’s referring to. I thought the 9000 and 6800 were still 5-bolt.
So: you’re right, it’s not the 4-boltedness, it’s the fact that the spider looks like it developed a huge cyst.
Interesting… I put a R8000 Stages dual sided PM on the bike a few weeks ago and was puzzled by my 52L/48R balance. I’m clearly right leg biased in everything I do physically, it didn’t make sense.
Sure, if you want a power estimator, get a left side only unit.
If you want a power meter, you will need to get a dual side unit.
Obviously budgets differ, but as the price of dual side units continues to decrease, the need to settle for a power estimator also decreases.
Or a PM that combines power like chainring/spider or hub based options
I think it’s a well known issue that Shimano PM’s aren’t accurate even if Shimano won’t admit their PM is crap.
This is not just Shimano PMs though. This is all dual-sided crank based power meters installed on Shimano 4-bolt cranksets. Lama lists 11 different ones in his blog post, all with consistent inaccuracies (if that doesn’t make you lol, “consistent inaccuracies”, nothing will…)
I believe it is mainly an issue with the latest generation of Shimano cranksets - 9100 and 8000 - which are 4-bolt but, most importantly, are asymmetric designs. I don’t think it’s the 4-bolt chainring attachment that’s the issue. It’s the fact that the crank arm is not symmetric. It’s offset and sculpted. I am pretty sure GP Lama noted that he didn’t see issues with the 9000 and 6800 cranks that were used by PM makers to construct their powermeters previously.
I don’t believe the previous generations of Shimano cranks bolted on to, say, SRM’s spider containing the strain gauges have any issues.
Interesting. I now see what he means with “asymmetric”. I though he meant he fact that the crankarms aren’t distributed evenly, and that that caused the issues, but now I see that the 9100 (and supposedly the 8000 as well) have an “blob” off to the side of the spider centre, under the crank arm, and that’s the asymmetry he’s referring to. I thought the 9000 and 6800 were still 5-bolt.
So: you’re right, it’s not the 4-boltedness, it’s the fact that the spider looks like it developed a huge cyst.
The 9100 and 8000 arms are asymmetric too and I am sure that is also contributing. The top half of the right crank arm is shaved/beveled as it merges into the spider.
My takeaway: If you’re in the market for a crank based powermeter, pick an FSA or Rotor crank, or another cranks that uses the traditional 5-bolt pattern.
Except that FSA has now gone to asymmetrical 4 bolt chainrings as well…
In an interesting bit of cooperation between a representative of the “gear reviewer” community and the nerdy, techy community, Shane Miller (GP Lama, gplama here) and Keith Wakeham (initial developer of the 4iiii power meter, kwakeham here) have come up with a convincing case that the asymmetric nature of the newer Shimano 4-bolt cranksets wreaks havoc on the accuracy of crank based powermeters using those cranksets.
My takeaway: If you’re in the market for a crank based powermeter, pick an FSA or Rotor crank, or another cranks that uses the traditional 5-bolt pattern.
Which is why Pioneer released an update in their V3 power meter that was released this year. They specifically adjusted due to issues found last year. I think they are the only ones who have specifically addressed this issue. Shane Miller has noted specific issues with Stages and Shimano that was also noted by DC Rainmaker. I am not sure what 4iii has done in response to this.
Which is why Pioneer released an update in their V3 power meter that was released this year. They specifically adjusted due to issues found last year. I think they are the only ones who have specifically addressed this issue. Shane Miller has noted specific issues with Stages and Shimano that was also noted by DC Rainmaker. I am not sure what 4iii has done in response to this.
The newer Pioneer was part the test pool, two of the newer ones in fact. There’s no good news for any of them at this point. See Keith Wakeham’s video on why this isn’t a simple gauge placement ‘fix’, or even a firmware updatable fix.
Which is why Pioneer released an update in their V3 power meter that was released this year. They specifically adjusted due to issues found last year. I think they are the only ones who have specifically addressed this issue. Shane Miller has noted specific issues with Stages and Shimano that was also noted by DC Rainmaker. I am not sure what 4iii has done in response to this.
The newer Pioneer was part the test pool, two of the newer ones in fact. There’s no good news for any of them at this point. See Keith Wakeham’s video on why this isn’t a simple gauge placement ‘fix’, or even a firmware updatable fix.
Interesting, Just read through the articles. I own several Quarq’s but just purchased a new DA pioneer for my primary road racing bike. In your opinion is this power discrepancy consistent? Somewhat wishing now I had gone ahead with the SRM. Or could future firmware potentially address this?
So how do we explain several rides (>60 minutes) with P2max on 4 bolt DA compared with powertap rear hub and got nearly identical curves (with a relative offset of about <3 watts (>200 watts AP)) with final AP and NP being reflective of scale offset…?
I had hopes that this analysis is accurate. It would explain a lot. Ever since my knee surgery, my left (repaired) leg has shown to be stronger than my right on my Stages dual-sided power meter. (I do my FTP testing using the Stages.) Generally, it’s between 52/48 and 54/46. Almost all the time. Makes no sense. My right leg is my dominant leg.
On another bike, I have Garmin Vector 2 pedals on an FSA direct mount crankset. Nothing assymetric about that. I just looked back at a good sampling of files from that bike. Left leg stronger, between 52/48 and 54/46. Almost all the time. Total power on back-to-back days is almost identical to what I record on the Stages.
At least in my power files, there’s no imbalance or power difference between double-sided Shimano/Stages and my double-sided FSA/Vector. But . . . is there anything to be done about the small imbalance? (Should I even care?)