Mountain Bike Reviews Forum banner

Power2Max MTB Power Meter

55K views 229 replies 28 participants last post by  WR304 
#1 ·
As my Powertap hub has died I've been trying to consider what may be a good replacement. The Power2Max MTB power meter looks like a possibility. One nice thing about it is that it uses an accelerometer for cadence so you don't need a cadence magnet attached to the frame for it to work.



https://www.power2max.de/europe/en/Produkt/power-meters-en/rotor-3d-mtb/

DCRainmaker review of road version:

https://www.dcrainmaker.com/2013/01/power2max-power-meter-in-depth-review.html

Has anyone tried one and how does it hold up in bad weather? Riding in the UK getting soaked is unavoidable.

For my bike I'd probably go for a Rotor 3D crank (alloy rather than carbon crank arms so can be welded and modified) in a triple chainset version with a 169mm q factor, allowing me to keep my current setup of 44/33/23 front chainrings.

I have to use a modified left hand crank as I'm unable to use a normal length crank due to my knee not bending. I also use Rotor Q ring oval chainrings. Would a modified crank and oval rings result in inaccurate power readings?

This picture shows my current modified crank, which is used with a Shimano Deore XT chainset. I'd have to get a new one made to match the Rotor cranks. It has a 75mm crank length and a 50mm swing crank. The right hand crank is a standard 175mm crank with no alterations.

Swing crank explanation:
https://www.highpath.net/highpath/cycles/swingcrank.html

The swing crank allows you to pedal without bending your knee as much as normal.





.
 
See less See more
3
#28 ·
First of all I want to say how much this thread and WR304 has helped me in my search for a Powermeter.
Kudos and Thanks to you! I value your detailed style of reporting of new information to the Forum.
Being of similar mindset (detail oriented) I find it refreshing that you leave no variable unexplored and no idea untested. Some of your explorations into the subject of non-standard Powermeter / Crankset applications have significantly shortened my own timeline on becoming an owner.
I finalized my order with Power2Max today and will post an update once I have the parts in house.
Thanks for reminding me that I'm not the only one who thinks...
"Life is too short for off-the-shelf parts!"
 
#30 ·
Here are a couple of pictures showing how SRM fit their power meter spider onto a Shimano Dura Ace crank. SRM machine a standard Shimano Deore XT crank for their MTB crank so the idea of doing the same with a donor crank in order for it to accept a Power2Max spider looks to be sound. When you remove the spider from a Shimano Deore XT crank there is enough material there to add a spline.

These pictures are of a Shimano SRM Dura Ace crank with the spider removed so you can see how SRM have machined the Shimano crank:









This picture shows a SRM power meter fitted to a Rotor 3D crank. Instead of using the Rotor threaded lockring they have chosen to use allen bolts to attach the crank spider onto the crank arm. When making a custom crank this attachment method looks to be the easiest, saving the task of having to create a thread for the lockring



This picture shows a Power2Max spider (for an FSA crank) which has holes drilled and is designed to be attached via bolts.

 
#31 ·
It's been a long time coming but my Power2Max power meter is finally done!

It consists of a Shimano Deore XT M770 crank arm that has been modified to fit a triple Power2Max spider. First the steel axle was pressed out, and then the splines from the original Rotor 3D crank were measured using a CMM before being machined into the Shimano crank arm. It was decided that there was enough metal to cut a thread so a threaded lockring was used, rather than my idea of tapping holes. That has the advantage of not voiding the warranty on the Power2Max unit.





As this was a one off the splines machined into the crank and the splines on the spider match exactly with very tight tolerances, resulting in a closer fit than with the original mass produced parts.



When assembled this is how the power meter looks side on. The heat marks on the crank arm are from where a blob of weld was added in case there wasn't enough metal to create a spline. This proved to be unnecessary as the entire spline is original crank material.



This picture shows the chainrings installed and lockring. The outer ring is a BOR 44 tooth round chainring, the middle ring is a Fetha 33 tooth oval ring (with the same ovality as a Rotor Q ring but rotated bolt holes) and the inner ring is a Rotor 23 tooth oval Q ring. This gives me exactly the same gearing as I used before.



Here's a picture of the crank installed on my bike. Blue decals to match the Garmin Edge 500 head unit. It all seems to be ok. The power meter is detected and sending data to my Garmin Edge 500, the front shifting between chainrings works and the crank feels solid. I'm hopefully going to give it a proper try tomorrow to see what the data looks like.:)

 
#33 · (Edited)
The Power2Max power meter has an estimated left / right leg balance (bottom right figure in the picture of my Garmin below). When riding along without trying hard my balance is around 85% right leg / 15% left leg but at maximum effort, doing 400 watts trying not to get dropped on the club run for example, it evens out for short periods to as much as 65% right leg / 35% left leg. I can't seem to sustain that for long though, only a few seconds.

The Garmin picture is the summary from today's ride where I was just out by myself, rather than on a group ride. The elapsed time is only more than the ride time because I took the photo before saving the ride file after I got home. There wasn't actually any stopping during the ride itself.

I averaged 279 watts (4.22w/kg) up the main climb (1.29 miles, 10% gradient, 10 minutes 24 seconds, 7.5mph average speed) and set a personal best time also, 33 seconds better than I managed in 2012 before crashing and having an enforced year off the bike. I was quite pleased with that.:)



Here's a more detailed look at the left / right leg balance in rubiTrack for IOS. The top line shows my power output in a short section of today's ride whilst the lower line shows my left / right leg balance as recorded by the Power2Max power meter. As I try harder the balance temporarily becomes more equal. rubiTrack seems to have a bug where left and right are the wrong way round on the graph it displays. I've marked it in text which is which.



Some Power2Max left- right balance notes:

"1) Our power meters don't measure left-right balance by using a reed contact to distinguish between 0 to 180 and 180 to 360 degrees, but goes from torque peak to torque peak: it compares the relative power of the left leg power phase with the right leg power phase. So it can be a bit different from Quarq because of the different approach.
2) The (Garmin head unit) calibration figure has no significance. Any offset changes during the ride (between auto zeros) are compensated by the temperature compensation mechanism.
...
4) No, we have tested oval rings and they don't inflate power - we measure at 50HZ and don't use discrete, event based cadence.

Best
Nicolas "
power2max

https://forum.slowtwitch.com/cgi-bin/gforum.cgi?post=5023818#5023818
 
#34 ·
I've got a few more rides in with the new power meter now. It's all been working ok and it survived its first wet ride today also.:)

This is only the second power meter that I've had. The previous one being a Powertap MTB hub. Something that I hadn't realised is that the information displayed by different power meters isn't the same. Due to the way a Powertap hub samples power the number displayed on the Garmin Edge 500 screen using a Powertap hub is very jumpy, making it hard to use for pacing without using a lot of smoothing, as otherwise the display will vary by 100 watts plus every time it updates.

"Likely the main reason for the variation in power you are seeing is due to the sampling rate (1 or 1.26 Seconds) of the Powertap. The samples are not synchronized to the crank rotation. So although the average power will be correct the individual samples will jump around. This is sometimes referred to as the precession effect. It doesn't affect SRM, Quarq or any other crank based powermeter.

The simplest solution is to increase the display averaging to 10S. It won't affect the raw data but it will settle down the display."
gregf83

https://www.cyclingforums.com/t/482272/powertap-measure-very-unstable-is-this-right

I thought it was normal and that all power meters would have a display as jumpy as a Powertap. That isn't the case however. The two WKO+ 3.0 graphs below are of the same route on different days, one ride from last year using a Powertap and then one ride from today using a Power2Max crank. What this is trying to illustrate is the difference in stability between what is displayed on screen whilst riding. The Powertap figure jumps around by 100 watts plus whilst the Power2Max is far easier to interpret as the basic information it sends is more consistent, even before it has some smoothing applied.

In use this is a massive advantage for the Power2Max if you want to use the power numbers for pacing whilst riding. The more stable the power number that you have on screen the easier it is to gauge your effort. I used to have 10 second smoothing on the Garmin display using a Powertap but have been using 3 second smoothing with the Power2Max.

Powertap hub:


Power2Max crank:


One of the concerns I had was that using a triple chainring crank based power meter with a mixture of round and oval chainrings might result in some odd readings. I did a 20 minute FTP test on Monday to try and get a feel for where I am fitness wise. My route starts off on the flat, goes up a hill and then I do a few minutes on the flat after the top to make up the time if necessary. On Monday I reached the top of the hill with two minutes left on the clock and kept the pressure on, shifting straight up to the big ring at the same effort as I had been putting in on the hill.

The WKO+ 3.0 graph below is an extract showing the final few minutes of the test. I climbed the hill on the 23 tooth Rotor Q ring inner ring at an average power of 279 watts, reached the top of the hill, shifted up and kept going as hard as I could, averaging 279 watts on the flat using the 44 tooth BOR round outer ring also. It's not conclusive but I don't think the mixture of chainrings is going to be an issue.:)

 
#39 ·
The more stable the power number that you have on screen the easier it is to gauge your effort. I used to have 10 second smoothing on the Garmin display using a Powertap but have been using 3 second smoothing with the Power2Max.
The weather was quite bad this morning so I only did a short road ride in the rain.

I decided to do the ride trying to stay at 200 watts throughout, using the Power2Max power display to pace myself. This graph shows the entire ride with 1 minute smoothing applied. By adding more smoothing to the graph it makes the overall trends more obvious. The yellow dotted line shows the 200 watts that I was targeting.



Although the blue speed trace varies the power output is mostly in the right area. I brought the pace up a little for the final 20 minutes. There were a few junctions, traffic lights and two u- turns (trying to find a gap in the rain) so it's not exact. On a turbo trainer you'd expect a perfect match but that's not really possible outdoors.



This shows the detail summary for the ride. If you look at the average power and normalized power figures they're quite close, 201 watts average power and 208 watts normalized power, showing that the ride was at a fairly constant pace. The more variable the ride (hills, descents, lots of sprinting etc) the bigger the difference between average and normalized power is likely to be.

The other thing to look at here is the power bests for each duration. Because I was aiming for a steady pace for just over an hour of riding the short duration power bests are close to the best one hour power. Any time the power and effort started going too high I would back it down. At the same time I'd try and keep the pace at the target on the downhills also, which means going straight up through the gears to maintain the wattage.



This shows the power distribution for the ride. I spent most of my time close to my target of 200 watts. I was quite pleased with the 0-20 watt figure too. 3 minutes 8 seconds, 4.3% of a 1 hour 13 minute ride spent freewheeling isn't too bad. That was mostly down to junctions too, rather than freewheeling for a rest.:)
 
#35 ·
The Power2Max has survived two wet rides now. 8am this morning, pouring with rain up in the hills whilst busting out some intervals. I'd swear I could hear a tiny violin playing.:lol:

If you want to ride in the UK getting rained on is a regular occurrence. It was wet enough that the barometric altimeter on the Garmin Edge 500 stopped and my shoes were full of water by the time I got home so it counts as a proper test. The power data from the Power2Max was fine throughout.
 
#36 ·
One place that I've been seeing some strange behaviour from the Power2Max is after stopping for a bit. This is something I've had happen a few times on rides now. After a fairly long stop, 10 to 15 minutes, when I set off again the power numbers will temporarily be a good 100 watts lower than they should be. It's not every time I stop though, only every now and again.

This was particularly noticeable today as I was out on the Tuesday club ride. It was dry, fairly warm and sunny. After 15 minutes spent chatting we set off. Immediately after setting off I was on the front of the group, riding along at 18mph and already starting to try a bit. I was probably doing around 200 watts. The power display on the Garmin Edge 500 was saying 100 watts though!

The problem was that there were riders sitting close behind so I was unable to safely do any extended freewheeling. For the next six minutes the power numbers stayed wrong until on a short downhill section I was able to do two four second freewheels to try and let the power meter sort itself out again. After that the power numbers returned to normal and it was ok for the rest of the ride.



You can see this clearly in rubiTrack when looking at the ride file. The left - right pedalling balance is normally dynamic and changes all the time. For the first six minutes after setting off the balance doesn't change. After I do some freewheeling the left- right balance starts responding again.

 
#37 ·
What I've been doing if I'm stopped for a long period of time is to spin the cranks backwards a few times before setting off, in an attempt to make sure the power meter wakes up and has a chance to settle down before beginning riding. That seems to be a procedure that works ok as I haven't had any of the odd power readings since then.:)

I've been getting out and doing more offroad rides with the Power2Max. If you've read some of DC Rainmakers comments about accelerometers and rough surfaces one of the things that I was wondering was how well the Power2Max accelerometer based cadence would work offroad.

https://www.dcrainmaker.com/2014/05/garmins-spd-cad-magnetless-sensors.html

Riding on the farm tracks and bridleway of the Cotswolds it seems fine, without doing anything notable or out of the ordinary. The graph below is a short extract from today's ride. The first part is on some woodland singletrack. It goes over a series of roots and is then a hardpacked dirt path running around the edge of a field. It's quite bumpy from horses but the bike's suspension deals with it. Part way through I made a mistake, slipping down from the off camber path onto the rough grass, losing some speed. I had to power back up the slope to get back onto the good line again.



The graph below is another extract from today's ride. This shows a section of farm track which consisted of overgrown dried tractor tracks in ruts. It was very bumpy and you couldn't see what was coming up or avoid the worst of it due to the grass covering the ruts. I had the bike's rear suspension set to full open with no brain platform to try and increase comfort.

Part way along the track I reached a short but steep slope, rode down it and the bike was bucking so much it was all I could do to hang on. At the bottom there were some deeper muddy ruts before it went uphill. I slowed down, picked my line carefully and then accelerated up the other side of the bank.



There are a few up and down cadence changes that I'm not completely sure I remember here heading down the bank. I would have spun the cranks whilst downshifting a few gears in preparation for getting through the mud at the bottom though so it's most likely from that. The part where I was definitely freewheeling at the fastest point down the bank has no cadence or power spikes, and that bit was properly rough, so it seems about right.
 
#38 ·
Here are my settings on the Garmin Edge 500. I'm using the 3.30 firmware on my Garmin Edge 500 and have an ANT+ speed sensor:

System Menu
Auto Power Down - OFF

Bike Settings Menu:
Auto Lap - OFF
Auto Pause - OFF
Auto Scroll - OFF
Start Notice - OFF

Data Recording
Cadence - NON-ZERO AVG
Power - ZERO AVG
Recording Rate - 1 Second (although it doesn't matter with a Power2Max because it automatically records at 1 second intervals when a power meter is connected)

Under the Named Bike Option -

Bike Details
Wheel Size - CUSTOM and use a manual tyre circumference so that the GPS isn't involved. Whenever you fit different tyres this needs changing. Don't leave it on auto.

Because an offroad tyre at low pressure compresses so much when you sit on the bike a measured wheel circumference roll out with no weight on the bike is going to be inaccurate (too large a circumference). I typically have a few favourite tyre types that I use most of the time and use a track pump so they always start off at the same pressure for consistency.

For each tyre at the air pressure that I intend to use I measure the moving wheel circumference whilst actually riding by putting a small blob of white grease on the centre of the tyre tread (something that is easy to identify, will leave a mark and show up on the ground) and then ride along on the bike for a few metres sitting in the saddle as I would normally. The front and rear tyres compress different amounts so you need to measure the one that your wheel magnet and sensor will be attached to.

Using a tape measure you then go back and measure the distance between two of the white blobs left on the ground as the wheel rotated. That distance is your actual moving wheel circumference to enter into the computer settings for that particular tyre. That ought to give you accurate distance measurements.

ANT+ Spd/Cad - YES (individual Spd Sensor)

ANT+ Power - YES

Sensors
For speed I have a VDO Z1 ANT+ speed sensor, although any ANT+ speed sensor should work. Bontrager sensors have worked well for me also.

https://www.vdo-series-z.com/downloads/zubehoer_en.pdf

Bontrager: node



Here's an example of why I have a speed sensor instead of relying on GPS speed from the Garmin Edge 500. This graph shows the GPS speed trace on a wooded climb under tree cover, where I was travelling at low speed with obscured GPS reception, overlaid with the speed trace from my old Powertap hub. The speed readings from the Garmin are way out and very inaccurate making it useless for any type of analysis.



Screen Display
When riding I like to have quite a basic display as my main screen, something that's easy to glance at and take in offroad. If you have lots of different items displayed it makes it harder to identify what you're after so you spend more time staring at your stem instead of the trail ahead. As you add more items each one becomes smaller also. With the Power2Max I have "Power 3s Avg" as the top line of the display. This makes it largest and most obvious.

Below that I have "Speed" and below that I have "Time". I like to have speed displayed to have an idea of when to change gear. I use time as a guide for making sure I remember to eat and drink regularly.



I then have a secondary summary screen, as shown in Post #33 . This is for looking at the ride summary when I get home before saving the ride as the History screen on the Garmin Edge 500 doesn't have much detail.
 
#40 · (Edited)
One of the things that I find a power meter useful for is stopping myself getting too carried away and overdoing it. I frequently have a habit of digging a deep hole with lots of riding but then not having enough rest to actually bank the fitness improvements:



In WKO+ 3.0 you have the PMC (Performance Management Chart). This attempts to give you an overview of your riding, including how well recovered you are.

https://help.trainingpeaks.com/entries/22684490-Performance-Management-Chart

Here's my PMC for May 2014 and the start of June 2014. May was a fairly solid month with 74 hours riding. The blue bars show the TSS for each ride.



If you look at the yellow line in this chart this is the TSB (Training Stress Balance). A TSB score of 0 or higher is supposed to show that you're well recovered. The lower the negative TSB score the more fatigue you have building up. I managed a negative TSB score of -94 somehow.:lol:

After a few weeks of this I decided it was time for an enforced rest. The green highlighted area shows where I took a rest week.

Here are a few extracts from my notes:

02 June 2014 - no riding
03 June 2014 - no riding
04 June 2014 - no riding

05 June 2014 - 2 hours 14 minutes riding, 120 watts average power
"gentle recovery ride, legs very sore, took it easy throughout, nice day, quite warm and sunny."

06 June 2014 - 2 hours 12 minutes riding, 116 watts average power
"gentle recovery ride, felt better than yesterday, legs weren't sore whilst riding but still a bit sore afterwards, put in a few efforts during last five minutes or so and they were ok, nice day, quite warm."

07 June 2014 - no riding

08 June 2014 - 2 hours 25 minutes riding, 108 watts average power
"gentle recovery ride, felt ok, having to try to concentrate on keeping pace down, legs were ok no soreness."

It's surprisingly hard to keep the intensity right down. You have to really pay attention to what you're doing as otherwise the power output starts creeping upwards. This graph shows part of Sunday's ride where I was trying to keep the pace down. It meant going as slowly as 5.3mph in bottom gear up this rise.

 
#41 ·
This is an amazing thread. I haven't even read all of it,, but I can see tons of interesting information in it. That crank modification is a thread all on it's own. I know what's involved, and I don't even know how they held the part. Great stuff!
 
#42 ·
My Power2Max power meter has been working well so far. It's done 225 hours riding, including plenty of wet rides in heavy rain, and I really don't have anything bad to say about it.

I did have an interesting occurrence in today's ride though, a power spike. I was freewheeling, hit a rough patch at speed, with the bike slightly unweighted, and the power meter reported a 878 watt power reading in error.



That's notable in that it's the first time I've had it do that. I haven't seen that happen before which is why I'm mentioning it. The rest of the ride file looks fine with nothing else odd or out of the ordinary.

The power meter battery is supposed to last 400 hours before needing replacing. Apparently you get a low battery warning appear on the Garmin Edge 500 screen also with a Power2Max. I've bought a few replacement Renata CR2450N batteries off Amazon to be on the safe side.
 
#43 ·
I was so disappointed by getting dropped on the club ride on Tuesday that I did the last part of the route again the day after by myself. The club ride was in a group of four riders (including me). They were having a steady ride whilst I was trying fairly hard to keep up. On the flat main road section, despite sitting in at what should have been a do-able power output, I blew up completely. I wasn't happy with that and decided a second try was needed.

The two graphs of the same route viewed together give quite a nice summary of what power output on a group ride looks like compared to a solo ride. You can see how on the group ride my power output goes up and down repeatedly with frequent spikes well above 235 watts, and also frequent drops below 235 watts. I didn't spend much time on the front, the speed was dictated by the other riders.

On the solo ride riding at my own pace I was aiming for around 235 watts and riding as fast as I could. If you look at the two graphs alongside one another the main thing to note is that there are fewer high power spikes on the solo ride graph. Even though I was trying hard that meant maintaining a constant tempo. If you put in a short hard burst it takes time to recover from that. At the same time there are fewer low power sections as I was trying to keep the pace high on the flat and downhill sections.







On the first climb I was 13 seconds slower by myself, due to spinning a lower gear on the steepest section to not go too far into the red. With the club I'd put in a big effort near the top of the climb to keep up. I then pushed hard on the rolling roads over the top, 20 seconds faster than the club ride. From the top of the hill down to the main road I pulled out another 22 seconds, so that by the time I reached the main road I was 31 seconds ahead of the club ride.

That still left the 9.55 miles of the main road to go though! With a strong rider riding tempo on the front the average speed along the main road to where I was dropped had been 7 miles in 18 minutes 53 seconds at an average power of 205 watts and an average speed of 22.2mph. On the club ride I'd been sitting in drafting and that pace would be hard to match by myself.

Going as hard as I could by myself along the main road to the point where I was dropped I managed 7 miles in 19 minutes 59 seconds at an average power of 229 watts and an average speed of 21mph. 66 seconds slower than with the club.

On Tuesday's club ride I then blew up spectacularly. The remaining 2.55 miles to the junction took me 10 minutes 2 seconds at an average power of 140 watts and an average speed of 15.6mph.

Wednesday I was ok and kept the hammer down all the way. 7 minutes 40 seconds at an average power of 238 watts and an average 20 mph. It shows how much of a difference an extra 100 watts makes as I made up 142 seconds in just 2.55 miles.

Overall I was 105 seconds quicker solo compared to the steady club ride but that was a big effort. I then gently rolled home, getting caught in the rain and soaked again...
 
#44 ·
Doing a fair comparison for accuracy between different models of power meter is notoriously difficult. See the first section of this dcrainmaker review where he discusses the issues:

https://www.dcrainmaker.com/2013/06/stages-review-update.html

Something I've been meaning to post for a while now is my Mean Maximal Power Chart for 2012 and 2014. As my PowerTap hub broke I didn't have a chance to ride with both the PowerTap and Power2Max recording at the same time to see if there were differences between them. Whether the same effort I was putting in would appear as the same power output on both power meters.

I've done around 250 hours riding with the Power2Max power meter with plenty of hard rides and interval sessions now. The Mean Maximal Power Chart is a logarithmic chart that shows your best power outputs. In WKO+ 3.0 you can overlay this with best power outputs for different dates. In 2012 I only used a PowerTap hub whilst in 2014 I've only used the Power2Max crank based power meter.





What's immediately obvious is that there's a big difference between 2012 and 2014 at short durations whilst sprinting, up to around 30 seconds. The PowerTap reading much higher than the Power2Max. This may be because I'm not putting out as much peak power (due to having damaged my left leg even more in 2012 and having to go to an even shorter left hand crank) but it is quite striking. My suspicion is that this might be a difference between the power meters as I feel that I'm sprinting close to how I was in 2012.

At longer durations the two years are closer, the PowerTap consistently reading higher but not by a massive amount.

In terms of speed I've been setting climbing personal bests and quicker times by several minutes over longer routes than in 2012 (on the same Specialized Epic and Ground Control tyres).
 
#45 ·
Regarding the peak power, I think you're correct about the difference between powermeters. I have a Powertap on my moutain bike and SRM and Quarq on my road bikes. I've always been able to make about 50w more for peak power on my mountain bike than my road bikes. I'm pretty sure it's just the Powertap reading higher than crank based, as my two road powermeters are always pretty much dead on (5-10w) for peak power.
 
#46 ·
That is one disadvantage of having multiple bikes and several power meters, knowing whether the data that you record from each one is comparable or not.

Ideally they'd all read exactly the same but in practice there always seems to be some variation between units. It's not a huge deal for me fortunately because I use the one bike for everything. Whether I'm out getting dropped on the road club rides or dodging cattle offroad on a muddy bridleway up in the hills it's the same power meter. So far it seems consistent with itself anyway.:)
 
#47 ·
Power2Max have announced their newer Type S power meter for mountain bike cranks now. The Type S has been around for a while on road cranks since the start of this year so it was about time. The main differences are that the Type S power meter is less bulky and should be around 100g lighter than the Classic Power2Max power meter that I have.

It's available as a 1x crank also and is supposed to be available in October:

https://www.dcrainmaker.com/2014/08/power2max-introduces-additional.html

 
#48 ·
When the Power2Max battery begins to get low you get a "Power Meter Battery Low" message appear on the screen of the Garmin Edge 500. I had it appear around 2 hours into today's ride. There must be a safety margin built in as the power meter kept working. The battery lasted around 315 hours, less than the claimed 400 hours.

Unfortunately at the 3 hour point of the same ride my right hand crank snapped. I'd just gone off a speed hump in the high street and was overtaking a car when my right pedal started feeling very odd, as though the pedal axle was bent or my cleat had come out of the shoe. I stopped and realised the crank had failed.

From the beginning of May 2014 to 22 September 2014 that crank had done 315 hours riding / 4681 miles. Not a huge amount of miles but I do put full power through the right crank. When it broke it didn't come apart completely fortunately so I was able to stop safely and ring for a lift home, my second mechanical failure needing a lift home in two weeks. Ironically, it was only half a mile or so from the roundabout where I crashed and broke my leg in 2012.

The Shimano crank was modified with material machined from it and some welding done (which would soften the aluminium also) to get the Power2Max spider on originally. It failed as a result of the modification. Where it's snapped the metal looks very thin. I'm going to take it over tomorrow to take it apart and have a think about how the next version can be improved.:)





 
#49 · (Edited)
I took the broken Shimano Deore XT crank over to my friend's factory today to look at. He cut it open to examine the cross section of the crank where it had failed.

In the picture below you can see how the original crank casting was uneven in thickness, before it was modified, with one half of the crank being thicker than the other half of the crank. In normal use this would probably have been fine but during the process of fitting the custom spider material was machined off the flat on the inside of the crank. This left the inside of the crank very thin, less than 1mm thick at the narrowest point.

It's at this narrowest point that the crank eventually failed yesterday.

The plan for version 2 is to use a Shimano Deore M615 chainset instead of a Shimano Deore XT chainset. The cheaper Deore M615 chainset uses solid crank arms, rather than the hollow crankarms of the Deore XT chainset. This should avoid the problem of it breaking in the same place again as there will be more metal there.

The machined spline and threads on the crank looked great with no signs of any problems in those areas.

 
#50 ·
Between the torque due to the pedal about the crank arm, and the crank about the axle, it looks like the shear forces were highest at the thinnest part of the wall. It's an odd fracture, that I can only explain due to shear forces. But then you knew you were working with a hollow structure. I'd have been nervous riding that. I actually thought it was a solid crank until seeing this photo. Another additional contribution to failure could have been stresses due to notches from machining.

It's often interesting to modify things.

I was just looking at plans for a house that is going to be cantilevered out over an 80 foot high bluff over the ocean. The whole house sits on one 3 foot diameter column. It's really cool. We were mulling over the way it might behave in various seismic modes. I wouldn't want to be the engineer responsible for that thing.
 
#51 ·
My thinking behind the original Shimano Deore XT M770 crank was that as SRM sell a modified version (see the SRM modified Dura Ace crank arm pictures in post #30) it would be ok to modify it for this also. The wall thicknesses vary across a Shimano crank. I'd seen the top cross section of an XT crank previously but hadn't seen a lower cut cross section.

This picture shows the upper section of a Shimano Deore XT M770 crank arm. It's much thicker at this point.



I didn't crash so it's a case of moving on and learning to avoid repeating the same mistakes again.:)

My Shimano Deore M615 crank turned up today. Here are some pictures showing the rear of the solid crankarm. If you look at the shape of the crank arm this is a solid crank, rather than the hollow crank arms on the higher end models. It still has the two piece spider and crank arm so will come apart for modification.



One slight concern is that the Deore M615 crank is shaped and has quite a deep cutout underneath the spider behind the crank arm. That might possibly interfere with machining a spline into the crank if there isn't enough metal there.



It's going to be a case of having to wait for at least a few weeks before my friend can fit the work in at his factory so I've been using my old crank and riding on feel (who needs a power meter anyway? ) ;)

One thing I find interesting is how riding with and without a power meter affects me psychologically. With a power meter I tend to gauge my effort and pace to it. That works well over longer rides in particular. At the same time it also acts as a "rev limiter" sometimes where I could go harder and it's a distraction.

Without a power meter on the stem I tend to just blast up everything as fast as possible if I'm feeling ok. That works well up to around the three hour mark.:lol:
 
#52 ·
That's the same reason I'm looking forward to a power meter. I tend to ride race pace everywhere.

This project reminds me of when I was fitting a Datsun alternator onto a Norton Commando. I remember searching motorcycle junkyards for the right alternator. Before that, I was taking alternators apart, and machining special fittings to get cases and cranks and rotors all lined up. I had all but forgotten about that until now. But I think what you're doing has way more merit. I graduated from motorcycles to mountain bikes a long time ago.
 
#53 ·
I picked up my new version of the crank today. It's had the same machining done as on the previous crank. This one uses a solid Shimano Deore crank arm with no welding involved however, which should avoid the problem that I had with the hollow crank arm breaking.

I haven't tried it yet but it looks to be solid and should hopefully hold up better. :)

 
#54 ·
I did a few hours riding with the new crank today and it was great. No strange noises or play from the modified crank, it all felt aligned whilst pedalling and the front shifting was better than with version 1. Version 1 (using the Shimano Deore XT crank) didn't have quite the right chainline (the rings had to be moved further inboard relative to the crank arm) but on this crank the chainline is very close to ideal.

I'm not a huge fan of BOR chainrings though. My outer BOR 44T chainring seems to be made of cheese. I think BOR have made the ring too flimsy as it's prone to bending out of shape. I trued it up with a spanner which seems to have helped.

I'd fitted a new Renata CR2450N battery in the power meter also. When doing this on the classic Power2Max the three screws for the battery compartment are small. I'd recommend considering taking the crank off the bike and doing it on a table to make certain you don't lose them. When you've changed the battery there is a small LED inside the battery compartment which will flash green when you move the crank to show it's functional.

The power readings seemed to be the same as normal today. I did have one strange occurrence. A 1000 watt power spike whilst freewheeling 1 hour 50 minutes into the ride. Possibly it was like the other spikes I've seen in the past where I was changing gear, turning the cranks round but with no pressure on the pedals. It was only once in a 3 hour 11 minute ride though.

 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top