When switching from a standard chainring to a solid TT chainring, does the power reading from a crank based power meter tend to read high, probably due to the increased stiffness of the chainring?
When switching from a standard chainring to a solid TT chainring, does the power reading from a crank based power meter tend to read high, probably due to the increased stiffness of the chainring?
No, I fail to follow your logic? Crank based power meters just measure the bending of the cranks via strain gauges. The dont know if your chain rings are solid or not, The only difference is if the chainrings are oval or not.
No, I fail to follow your logic? Crank based power meters just measure the bending of the cranks via strain gauges. The dont know if your chain rings are solid or not, The only difference is if the chainrings are oval or not.
Then explain this from SRM:
Link: http://www.srm.de/support/faq/powermeter/
The calibration is usually done by hanging a known weight a fixed distance from the spindle. I suppose there would be less deflection with a solid chain ring and this would lead to a higher power reading, but I would be surprised if you could see/tell the difference associated with a stiffer ring. Maybe it could be a problem if you were trying to aero test a chain ring on a track?
It sort of reminds me of the problem of determining if a stiffer frame really lead to more efficient power transfer. Nobody that I know of has ever proven that more compliant frames affect power transfer and there have been some real noodles and buck-boards over the years.
No, I fail to follow your logic? Crank based power meters just measure the bending of the cranks via strain gauges. The dont know if your chain rings are solid or not, The only difference is if the chainrings are oval or not.
Then explain this from SRM:
Link: http://www.srm.de/support/faq/powermeter/
Your original post was about crank based powermeters, the link you posted to was for a spider based powermeter.
The spider is part of a Crank(set).
So is the spindle and crank arms.
I suppose there would be less deflection with a solid chain ring and this would lead to a higher power reading, but I would be surprised if you could see/tell the difference associated with a stiffer ring?
So it’s correct to assume that a stiffer ring would lead to a higher and not lower power reading for a spider based power meter? My gut and common sense tell me that this is the case, but I’m not 100% sure.
My hunch is the opposite. The spider meter measures the strain in the spider. If it is firmly attached to a rigid body, then there is a larger body to to participate in the distribution of the force and stain. Therefore the PM would detect lower strain.
This would have almost no effect on a crank-based PM, however.
It’s definitely not safe to assume that a “solid” chain ring is stiffer than a conventional one, especially if comparing to something like a hollow core Shimano chainring.
It is safe to assume that changing chainrings can effect the readings of a spider based power meter, and one should definitely recalibrate (if possible) when doing so.
It’s definitely not safe to assume that a “solid” chain ring is stiffer than a conventional one, especially if comparing to something like a hollow core Shimano chainring.
It is safe to assume that changing chainrings can effect the readings of a spider based power meter, and one should definitely recalibrate (if possible) when doing so.
Why would it affect the readings of a current generation spider based power meter? (P2M, Quarq, SRM) I don’t see any reason for it.
Hang a weight on a spider based power meter with and without a solid chainring on it and you’ll see no difference.
Why would it affect the readings of a current generation spider based power meter? (P2M, Quarq, SRM) I don’t see any reason for it.
Hang a weight on a spider based power meter with and without a solid chainring on it and you’ll see no difference.
I have done so, and there was a difference. YMMV.
It’s definitely not safe to assume that a “solid” chain ring is stiffer than a conventional one, especially if comparing to something like a hollow core Shimano chainring.
It is safe to assume that changing chainrings can effect the readings of a spider based power meter, and one should definitely recalibrate (if possible) when doing so.
Why would it affect the readings of a current generation spider based power meter? (P2M, Quarq, SRM) I don’t see any reason for it.
Hang a weight on a spider based power meter with and without a solid chainring on it and you’ll see no difference.
It’s a best practice. The reason srm notes is partially bc they offer service that includes calibration and partially since anyone who buys a 3k power meter probably also cares about having it in top condition.
On my old 7900 srm swapping from a solid fsa ring to a 7900 ring changed the slope a couple percent.
With all due respect, all of the PM’s I checked didn’t show any difference. P2M, Quarq and SRM
Edit: I work in calibration so I have the equipment to do it.
Edit on 10/5 i said SRM here, but according to my records i didn’t do any SRM’s in the runs doing dynamic calibration check. Only static.
Like I said, YMMV. I’d imagine my sample size is a fair bit larger than yours, given the hundreds of PMs I’ve dealt with over the years.
Having said all that, yes “current generation” PMs are much more robust in this regard, but there’s nothing in the OP limiting the line of question to current generation PMs.
It certainly is larger than the 19 or 20 that I’ve done.
How have the calibration checks gone on PM’s that didn’t change rings. Is the error statistically significant?
Sorry, not sure I understand the question… are you asking if “normal” calibration checks on cranks I have done typically display variance from “correct” setting?
If so, it’s been all over the map. Enough that I wouldn’t try to draw general trends/conclusions, other than that current stuff is much, much better than anything but Power Taps were a short while back.
These days, I mostly look at data from client PMs, and I don’t find myself having to chase down obvious calibration errors as much as I used to. Oddly enough, I have actually had a client change rings on a PM without calibration and had it blip their power numbers enough that I asked them if they had changed PMs. I also once made the mistake of not performing a static calibration when I changed the rings on a PM for a pro at Kona, and even though they were the same size/brand, the calibration was off enough that they had to wing it with RPE for the event. Ouch.
FWIW, I have had to look at PM data in several different contexts over the years, including as part of a PM product development project, and as a coach who has to look at a lot of data from many different PMs. The product dev/manufacturing role provided an interesting look at the objective accuracy of the products on the market at the time, combined with the subjective data presentation issues that are go into serving up data to users in a familiar/actionable format. The coaching perspective, though, has proven particularly illuminating in terms of general robustness of PM data. It’s, well… generally not so robust. Certainly not when you consider that people are pretty commonly trying to do things like perform intervals with output ranges that fall within the margin for error of their PMs.
You pretty much answered my question
.
It certainly is larger than the 19 or 20 that I’ve done.
How have the calibration checks gone on PM’s that didn’t change rings. Is the error statistically significant?
Wherever I’ve done calibrations I setup the bike with a little side to side roll as possible in a trainer or stand, then hang weights from pedal spindles. The problem I run into is finding and maintaining the correct crank position is difficult.
Have you built a static cal rig? That would go a long way to minimizing calibration error. I’ve heard legends of call rigs that better simulate the range and dynamics of pedaling forces but I’m pretty sure most companies are just doing two or three point fixed cals.
I built a dynamic one years ago. That’s some of the perks of working where I do. It’s shortcoming was that it produced even power throughout each rotation which is really what a static calibration assumes but is not what the PM sees. I used the PM’s I owned and from friends that loaned them to me
Doing a static calibration is not difficult as long as you make your length correct. It requires measuring the length and leveling the arm. I did a jig that was not part of a bicycle. All of the items used in calibration need to have a uncertainty study.
All of my calibration checks would have passed a 16949 and17025 audit. I simply don’t have the data driven from a few to several hundred PM’s to make a better claim than Fredly and the solid vs the non-solid chainring was not the intention of my study.
Edit: I just didn’t see any statically significant difference between the two types of chainrings on anything other than the Quarq Cinqo