Pre Sale - Static Torque Test

Thanks to a member here I have a potential buyer already for my current Powertap SL+ wheel and even though he has already emailed me and expressed a desire to purchase this used wheel I wanted to give him some comfort about this used wheel set. It also gave me an opportunity to try something a little different than my usual at home bucket test that I have always used in the past. (Please know that I am not at the level of DC Rainmaker, Rchung, Tom A. and others when it comes to precisely evaluating equipment. Not even close to their level. So take my little test as just something entertaining, but at least I can give the buyer in a different state an idea that the hub does function and I think it is in good shape. The wheel and bearings are good too.)

Here is that rough test using a weighed bucket.
http://thecyclingaddiction.blogspot.com/2010/05/powertap-torque-test-rough-initial-test.html
http://www.thecyclingaddiction.blogspot.com/2012/01/2012-torque-test-garmin-800.html

The disclaimer to the test today was still not an accurate method because I did not use a certified weight. The weight I used was a 45 lb plate measure on the gym scale. I used that number which was 46.5 pounds. I tried this method that I had read in the past by placing the weight on the pedal at the 3 o clock crank arm position and rolling the bike slowly and by just a little back and forth and take the highest number shown on the computer. My old method was a fixed crank arm position as close to level as possible. I think I still like that method to be honest. Any little shift in the bike the numbers were constantly moving so it was hard to balance the bike in the most upright position and trying not to let is lean from one side to another. So I am still not sure if I got this part of the test perfect either. So now we are at two potential errors creeping in with a non-certified weight and keeping the bike in perfect position to get the prime number. Anyway here is the pictures from the test this morning.

Just a bit more information
I used my training bike for the test, C-dale six13 with an Ultegra 170 crank arm, 34 ring was used and a SRAM 11-28 cassette. For the computer I used a Garmin Edge 800 and reset the zero offset after each cog tested. The wheels tested were the PT SL+ hub that I am selling and the new G3 hub that I had built from a friend and picked up yesterday afternoon.

Saris method to the consumer to verify the hub is checking the zero offset value (Garmin display "Current Calibration") and I think their prime number is 512 with a certain acceptable range above and below by a few digits.

SL+ = 514
G3 = 507

Because I am selling the SL+ I will just state the results of this morning's test for that wheel. Both wheels tested fine as I expected with the zero offset values being within range of 512, but it never hurts to get some more practice testing the power meters.

SL+ test readings were: 22 Cog = 22.2 / 17 Cog = 17.6 / 13 Cog = 13.3












Balance not set yet, but just a picture of the scale and weight used in this test


The bike and two Powertap hubs being tested at my gym


I used my mountain bike pedal so that I could slip the plate on (my regular Look Keo pedal will not work)


Zero offset value for the PT SL+ hub


Notes from the data collected on each cog on the SL+ (sorry for the blur)

 

I forgot to ask my question

Besides having a certified weight what else could I do to improve this personal power meter test?

 

Nice writeup and photos on your PM testing. Yeah verifying the weight is pretty important but looking at your results I'd guess your gym balance scale isn't too far off. Also FWIW I've had three Olympic style 20 kg weight plates measured on postal scales over the years and they all came in on the high side and in the 45 to 46 pound range vs the marked 44.1 pounds so your weighing seems typical.

I like the roll back method, but realistically it isn't that critical. You've got something like a +/- 8 degree range of crank angles on either side of horizontal before the loading really begins to drop off. IOW if you set the cranks thinking they're dead horizontal with respect to gravity but are off by +/- 8 degrees your actual applied torque would be ~ 1% lower than you think and if you're only off by +/- 4 degrees you'd be within +/- 0.25% of the expected torque so the crank leveling does not have to be dead on and errors in crank leveling will always reduce the applied torque, never amplify it relative to horizontal cranks.

It looks like your hubs are very close to expected torque accuracy.

-Dave

 

Do it many times to get your technique dialed and to demonstrate repeatability. That and with a PT it's a good idea to test in different gear combinations (which you did) but I'd personally test big ring and small ring and focus on combinations that give fairly straight chain lines as sometimes you get low results with cross chained combinations. If you don't go to sharp cross chained angles the different gears just applies different levels of hub torque for the same crank torque so that demonstrates linearity across a range of hub torques, not essential but it's nice to verify that your PM is accurate across a working range.

But realistically for testing hubs and feeling solid about the results the big things are to be very confident in your testing weight and to be confident in your technique and the repeatability of your test technique and that just means testing a lot of hubs till you're very confident with the whole process including feeling solid that you actually observed the highest applied torque.

-Dave

 

Thanks Dave for the advice.

 

???? Those are test reports for whole wheels as it's really hard to use your bike to test a hub not mounted in a wheel. But from a power measurement standpoint it's a hub your testing regardless of the wheel it's built into.