52/36 with a Power Meter

based on the picture i think changing chainrings won't affect the power meter, at least the SRM, why don't you contact them directly to ask ? might be a matter of configuration ? (p.s. you need to deal with 130mm or 110mm crankset sizes) [IMG ALT=""]http://www.cyclingforums.com/content/type/61/id/282652/width/350/height/700[/IMG]

With a 36T small chainring, you have to use a compact crank. You cannot put a 36T chainring on a standard (130mm BCD for everything but Campy, 135mm BCD for Campy) crank.

Quoted post said:

Originally Posted by alienator [IMG]/img/forum/go_quote.gif[/IMG]

With a 36T small chainring, you have to use a compact crank. You cannot put a 36T chainring on a standard (130mm BCD for everything but Campy, 135mm BCD for Campy) crank.

Yep, as Alienator posted you'll need a 110 BCD bolt circle for at least the small chainring. That either means a compact Quarq, SRM or similar if you want a crank based set or a hybrid like the FSA SRMs that have a 130 BCD outer ring drilling and a 110 BCD inner ring drilling. I run one of those on one bike and it's one way to get a crank based PM that supports the mid compact gearing.

-Dave

P.S. To answer the implied part of your question, you can buy and mount 36 or 52 tooth chainrings on a 110BCD bolt circle crankset. IOW, you can buy a normal compact power meter crankset with 50x34 chainrings and swap the chainrings to 52x36. IOW, there really isn't a 'mid compact crank' just different bolt circle diameters and then chainrings that match those bolt circles. The only gotcha is what Alienator mentioned above, you can only go down to a 38 tooth inner ring with 130 BCD cranksets.

Last year I purchased a slightly used Quarq Cinqo S975 SRAM 110 w/ Quarq installed 52/36 SRAM Red rings using steel bolts instead of the alloy. I am using this PM on a TT bike and with a 11 cog available I have no problem hitting my top end speed even on slight downhill gradients. I really like the gear ratios as well and that with me being a big gear masher/lower cadence type of guy.

I use 50/34 on my road bikes with a Powertap wheel, but find it just as easy to climb with the 52/36.

The newest Quarq PM's, from my understanding, do not need factory installation when changing chain rings. I do like the Quarq a bit better than the Powertap wheel and I certainly like having the flexibility to use different wheels.

Thanks for the responses. I ended up purchasing a Quarq Elsa. I kept the 110 so I can have the option to run a compact or the 52/36. Though after a few days with the 50/34 I think I might stick with the compact for a while - It just feels better and I get a little extra to work with on the climbs.

I might try out a set of compact q rings. They should work fine with the Elsa - I am curious to see if I feel a difference, especially when climbing. I need all the help I can get on the climbs.

Anyone have experience (good or bad) with the q rings?

Quoted post said:

Originally Posted by bronj [IMG]/img/forum/go_quote.gif[/IMG]
...Anyone have experience (good or bad) with the q rings?

No direct experience with the Q rings. Just know that they'll inflate your power numbers using a crank based PM.

The crank based PMs assume that crank angular velocity is constant for each full cycle of the cranks. With round rings that's a reasonable assumption. With Q rings your angular velocity is higher during the low torque parts of the crank cycle and lower during the high torque power phase. The net result is your PM will report higher power than you're actually sustaining and higher than what you'd see on a hub based PM (even considering the drive train power losses).

That may or may not change your view on Q rings but seeing as you just spent a lot of cash on an accurate PM it seems a shame to degrade that accuracy with ovalized rings unless there's a good reason. And from everything I've seen on Q rings and variations on that theme the performance improvement ranges from non existent to perceived (but not measured once you remove the known crank based PM measurement errors) and marginal.

-Dave

Hey Dave,
Thanks for the response, but I have to respectfully disagree with your assumption that q rings will inflate power numbers. The velocity of the PM wouldn't change because of the oval ring, the velocity will be whatever speed you are spinning the crank - the torque will be higher than usual during your down stroke (assuming everything else is equal, rpms & gear ratios aka a 50/14 round ring vs 50/14 oval) when compared to a circular ring. However, that's not an inflated number, the torque will actually be higher because you have 'more teeth' during the down stroke. This is then 'smoothed out' during the upstroke because you have less teeth to pull up and around. The resulting math of the changing diameter would be the same as if it were circular since it is spinning.

There's no reason the PM would assume that there is more torque or more velocity then there really is. There should be no difference in 'real' power reading from a round ring vs an oval ring.

I may be explaining some of the specifics wrong, but I verified this with a highly trusted mechanic. The bonus with the Elsa is that there doesn't need to be a recalibration when switching rings.

I am interested to try them out. I have heard that they help in climbing and long steady state rides - especially for riders whose pedaling mechanics are less than perfectly circular... aka mine haha

Would definitely be interested to hear any real world results - especially when climbing.

Quoted post said:

Originally Posted by daveryanwyoming [IMG]/img/forum/go_quote.gif[/IMG]

No direct experience with the Q rings. Just know that they'll inflate your power numbers using a crank based PM.

The crank based PMs assume that crank angular velocity is constant for each full cycle of the cranks. With round rings that's a reasonable assumption. With Q rings your angular velocity is higher during the low torque parts of the crank cycle and lower during the high torque power phase. The net result is your PM will report higher power than you're actually sustaining and higher than what you'd see on a hub based PM (even considering the drive train power losses).

That may or may not change your view on Q rings but seeing as you just spent a lot of cash on an accurate PM it seems a shame to degrade that accuracy with ovalized rings unless there's a good reason. And from everything I've seen on Q rings and variations on that theme the performance improvement ranges from non existent to perceived (but not measured once you remove the known crank based PM measurement errors) and marginal.

-Dave

Quoted post said:

Originally Posted by bronj [IMG]/img/forum/go_quote.gif[/IMG]

There's no reason the PM would assume that there is more torque or more velocity then there really is. There should be no difference in 'real' power reading from a round ring vs an oval ring.

brony, just reading this exchange. Dave explained the reason that the PM would give higher readings, and your reply was to state "there's no reason...". Apparently you reject his explanation, which your free to do of course, but he did offer a reason. Is his reason and explanation wrong, and if so, why?

Note, I know nothing about how the PM works, but his explanation sounds very reasonable to me.

Quoted post said:

Originally Posted by dhk2 [IMG]/img/forum/go_quote.gif[/IMG]

brony, just reading this exchange. Dave explained the reason that the PM would give higher readings, and your reply was to state "there's no reason...". Apparently you reject his explanation, which your free to do of course, but he did offer a reason. Is his reason and explanation wrong, and if so, why?

Note, I know nothing about how the PM works, but his explanation sounds very reasonable to me.

The velocity would not change due to the oval rings. Velocity is how quickly the crank is spinning - the crank would not spin any faster due to an oval ring. Your legs will still spin the crank at whatever speed you choose to spin them.

Quoted post said:

Originally Posted by bronj [IMG]/img/forum/go_quote.gif[/IMG]

Hey Dave,
Thanks for the response, but I have to respectfully disagree with your assumption that q rings will inflate power numbers. The velocity of the PM wouldn't change because of the oval ring, the velocity will be whatever speed you are spinning the crank - ..

The angular velocity issue related to non circular chainrings attached to crank based PMs is well documented.

Let me break it down as simply as I can.

The crank based PMs such as the Quarq and SRM use a reed switch to measure each complete cycle of the crank. They then calculate angular velocity as 2*pi radians divided by the time interval between two successive pulses of that reed switch. If intra-cycle angular velocity is constant or nearly constant as it is with circular chainrings then that is a good approach.

The problem comes with ovalized rings where by design the chainring radius is smaller across the top and bottom of the pedal stroke and larger through the power stroke. This is exactly the purpose of ovalized rings to make it easier to get over the top and bottom of the pedal stroke but apply more torque through the power phase of the downstroke. That ease of getting over the top of the pedal stroke results in your foot and the crank speeding up as it passes through those portions of the pedal stroke, That higher torque during the power phase results in your foot and crank speed slowing down during those portions of the stroke.

Sure alternatively you could guess that foot speed and crank angular velocity remains constant and the bike actually speeds and slows during those respective portions of the pedal stroke. That does not happen as a result of the linear and angular inertia of the system. Your bike speed does not pulse up and down to any great extent during portions of each pedaling stroke. Those are your only two alternatives, the angular velocity of the pedal stroke is non uniform and the bike speed is roughly constant or the crank angular velocity is constant and the bike speed is non uniform.

The problem really comes back to the single reed switch for measuring cadence and then deriving angular velocity for complete pedal strokes on the assumption that there is no intra-cycle variation in angular velocity. If you had something like optical encoders that sampled crank position every few degrees and did an on the fly angular velocity estimate this problem would not occur.

This is not my theory or an idea I originated. It has been heavily discussed over at the Google Wattage Group, Weight Weenies, Slowtwitch Forums and other places. Come on over to the Wattage Group and search the archives for more on this subject or perhaps just Google 'power meter error Q rings' I'm sure you'll get many hits and they'll all map back to the angular velocity issue. Search for Tom Anhalt's thoughts on this as he did quite a bit of early measurement and experimentation comparing Q ring setups using both crank and hub based power measurements to quantify the power estimation error that results from non round rings.

Here are some of Tom's thoughts and tests on the subject: http://bikeblather.blogspot.com/2013/01/whats-up-with-those-funky-rings.html

Use them, don't use them, makes no difference to me but they will inflate the power reported from your new Quarq which may be fine with you but don't stick your head in the sand and deny that there's a power accuracy issue with current crank based PMs.

-Dave

But Dave's point is that angular velocity (spin rate) does change with the oval chainrings during each revolution of the crank. That's the reason they are claimed to work after all. The average cadence is whatever you pedal, but the angular velocity still changes throughout the revolution. So, because the crank velocity and loads are varying throughout the revolution of the oval chainring, just using the average cadence combined with the varying torque readings throughout the revolution isn't good enough for an accurate PM reading any more.

If the PM had a high sampling rate for both the velocity and torque, and could integrate those, you'd have an accurate power result, But if the PM only measures the average cadence of the crank, like a standard bike computer does (with a single magnet pickup), that's just not good enough for the oval chainrings. Since both the velocity and pedal torque are constantly varying around the pedal circle, you'll need to have instantaneous readings of both to multiply together to get the instantaneous power values (force x velocity). Then those power samples can be averaged out, but not before. If you've ever studied integral calculus and wondered how in the world it could be useful in the real world, this is what it's about. The result needed would be an integral of the product of velocity and torque over the complete 360 degree revolution.

Hey Dave,
I have to apologize. I hadn't seen Tom's post about it, but after reading what you said and Tom's in depth real world tests it makes sense...
Sorry I doubted what you were saying. I guess it's one of those things that makes total sense in your head, but on paper is a different story.

So it seems that unless there becomes a way to 'calibrate' the cranks based PM's for oval use you will always end up with a 3-4% inflation on the data (or if they change the way it's calculated).

With that said... It seems like it's a roughly consistent inflation number. So let's say for example that it inflates the data by 4% - couldn't you just increase your FTP by 4% over what it actually is to be able to train within your proper zones? Or would that be inconsistent?

Lastly, have you read anything about real world tests (where that inflation is taken into consideration) that shows any improvement while using oval rings? Even if it's just improvement on mechanics? After reading all of that I am now questioning if it's even worth trying the q rings...

Quoted post said:

Originally Posted by bronj [IMG]/img/forum/go_quote.gif[/IMG]

Hey Dave,
I have to apologize. I hadn't seen Tom's post about it, but after reading what you said and Tom's in depth real world tests it makes sense...
Sorry I doubted what you were saying. I guess it's one of those things that makes total sense in your head, but on paper is a different story.

So it seems that unless there becomes a way to 'calibrate' the cranks based PM's for oval use you will always end up with a 3-4% inflation on the data (or if they change the way it's calculated).

With that said... It seems like it's a roughly consistent inflation number. So let's say for example that it inflates the data by 4% - couldn't you just increase your FTP by 4% over what it actually is to be able to train within your proper zones? Or would that be inconsistent?

Lastly, have you read anything about real world tests (where that inflation is taken into consideration) that shows any improvement while using oval rings? Even if it's just improvement on mechanics? After reading all of that I am now questioning if it's even worth trying the q rings...

Yeah, it takes some digging to get your head around the power inflation thing. I didn't buy it at first glance either and had to convince myself by looking at how the angular velocity and torque profiles were out of phase during the pedal stroke with Q rings to get my head around it.

I'm not aware of any real world tests that take into account the power inflation (or tests using a PT hub which avoids the inflation) that show gains with the Q rings. I do know many folks like them but I haven't seen any good evidence that they do what they claim to do.

I was thinking the same as you. It's super easy to field calibrate the slope of an SRM and I don't believe the Quarq is much tougher. The nice thing about this power inflation is that it manifests itself as a percentage so it can be modeled into the PM slope. It's easy enough for instance to increase my SRM CPU slope by 3%(which decreases the reported torque by 3%) or so if I wanted to run Q rings. As long as you're fairly confident about that inflation value you can offset it in the PM slope. BTW, the error value should vary a bit depending on what setting you use for your Q rings as I believe they have multiple settings that alter when the large radius teeth are relative to the cranks. I'm not aware of anyone that has tried that or published any results using a PT hub, Computrainer or other dyno device for reference.

-Dave

Interesting.

In lieu of adjusting the power curve, couldn't you just raise your ftp, and just be conscious that you're actual ftp is 4% lower?

I am surprised there isn't more testing out there to look at or read. I head good things from some racers who use them. A few Cat1s here swear by them. I would be interested to try them and see if they feel different - no reason you couldn't go back to circular rings after. Though I hear it takes 2 weeks to get used to the difference in mechanics.

BTW, I just found this.
Really in depth and interesting.

http://www.noncircularchainring.be/pdf/Appropriate%20non-circular%20chainrings.pdf

I ran some non-circular rings for 3 months, then switched back. One thing I did like about the rings is that I found them to be a bit easier on the knees, and I felt as though power was applied in one specific spot of the rings allowing a higher percentage of the rotation to be used for recovery. Riding on the circular rings again (first time back) it feels like somewhat of a chore to keep the pressure up on the pedals the whole way around. That said, after 2-3 rides I'm totally used to them again, and riding out of the saddle feels better.

Quoted post said:

Originally Posted by joroshiba [IMG]/img/forum/go_quote.gif[/IMG]

I ran some non-circular rings for 3 months, then switched back. One thing I did like about the rings is that I found them to be a bit easier on the knees, and I felt as though power was applied in one specific spot of the rings allowing a higher percentage of the rotation to be used for recovery. Riding on the circular rings again (first time back) it feels like somewhat of a chore to keep the pressure up on the pedals the whole way around. That said, after 2-3 rides I'm totally used to them again, and riding out of the saddle feels better.

Thanks for the input. I struggle with IT band issues which sometimes causes me knee pain. I'd be interested to see if it helps.

Quoted post said:

Originally Posted by bronj [IMG]/img/forum/go_quote.gif[/IMG]

...In lieu of adjusting the power curve, couldn't you just raise your ftp, and just be conscious that you're actual ftp is 4% lower?...

Well that gets into a discussion of power training philosophies and how you intend to use your PM.

If all you're after is target training zones then sure you can inflate your FTP or realistically just perform an FTP test with the new rings installed and use whatever you get from that test as your Q ring FTP.

But if you go on to do aero testing or want to estimate times for things like hill climbs or time trials based on your power curves then accuracy in the actual measurement matters. Personally I'd rather have a power meter that is accurate in absolute terms same way I'd want a voltmeter, oscilloscope, spectrum analyzer or other test instrument that's accurate in an absolute sense as it's a more versatile tool.

If you can reliably determine how much PM error is introduced with the round rings (sounds like 3% to 4% or so) then you can either calibrate it out or you can adjust your targets appropriately for the purpose of guiding training. I guess it doesn't really matter in the end but actual accuracy to known standards is a very useful feature in a meter and helps a lot if for instance you go back to circular rings or a hub based PM in the future.

-Dave

I emailed Quarq about the Elsa with q rings and this is what they said:

"The ELSA features Omnical and will not require a re calibration regardless of chainring selection." -Quarq

I then responded and got more specific I reiterated my specific concerns about oval rings: "But does having an oval chain ring distort the calculation for power on the Elsa? Or does having an oval chain ring not effect the power calculation at all?" - Me

"an oval chain ring does not effect the power calculation at all." -Quarq