PT stomp test - dodgy results...

GettingFaster said:

...The measured torque was higher than the expected value by 8%, 7%, 6%, 9% and 9% in the 5 gears I measured. ....BTW knocking 8% off the PT values I obtained in the interval workout I posted previously would put all the values within 2% of the reading I got from the ergomo. ...

Those tests have enough consistency that I would consider them pretty darn reliable. You can get low measurements from a stomp test if you have to lean on the bars or hang onto something for balance while using your body weight, but getting consistent high values sure makes it look like your PT is reading ~8% high across a range of torques. The fact that it pulls your PT into line with your Ergomo is just a bit more evidence that your PT is a bit out of whack. Drop Saris a note to find out if they'll help you out. Hard to say, but those numbers seem a few points out of spec so they may take care of it for you.

Tough call about the PT vs. Ergomo. I had an Ergomo and liked it a lot for its extra features like the altimeter and real time TSS/NP/IF/kJ display as well as an interval display that showed AP, NP, HR and cadence all at the same time. The one big downside of the Ergomo is that you can't run anything equivalent to the stomp test so if things drift with time or you reinstall your BB and things get funny, there's no way to test short of borrowing another PT or finding a calibrated bicycle dynometer(not likely but a computrainer or velodyne or even KK can do in a pinch but then you get into all the issues of tires, inflation, roller pressure...). Anyway I swapped to PT after getting some good deals on wheels but it's a real toss up in terms of which to use now that you have at least a baseline reference that your Ergomo is well installed and pretty accurate.

GettingFaster said:

so following the advice I received in an earlier thread I did the stomp test on my powertap SL today. Mounted it in the turbo, got a friend to hold the back brake on and measured the torque standing on the right pedal in 5 different gears. I then calculated the expected torque using my weight, crank length and the gear ratio.

Sounds good.

Just wondering how you were standing on the pedal. You might get slightly different results depending upon how you shift your weight around while standing on the pedal? Did you notice this at all?

Sorry for chiming in late but I would also suggest checking it with another known weight just as a sanity check and to see if there are any scaling differences. So maybe get your friend to stand on the pedals while you hold the brake. Or your wife/g.f./partner/s.o./child. Someone who might stand on the pedals slightly differently and whose weight is different enough from yours is a good person to ask. See if you get the same results on a percentage basis.

I know it's a pain in the backside to do this again so I can understand blowing this off. I spend a large percentage of my professional life dealing with instrumentation in a laboratory so I have some idea of things that can go wrong.

When I checked mine, I was getting pretty large errors. I dont remember if the error was high or low. Like you, I was standing on the pedals with the brake locked. I then hung a heavy weight from the pedal. In my case I had an old car battery. Then I could make sure the bike was vertical and the weight was steady. Lo and behold, the torque value was spot on.

Steve_B said:

Sounds good.

Just wondering how you were standing on the pedal. You might get slightly different results depending upon how you shift your weight around while standing on the pedal? Did you notice this at all?

Sorry for chiming in late but I would also suggest checking it with another known weight just as a sanity check and to see if there are any scaling differences. So maybe get your friend to stand on the pedals while you hold the brake. Or your wife/g.f./partner/s.o./child. Someone who might stand on the pedals slightly differently and whose weight is different enough from yours is a good person to ask. See if you get the same results on a percentage basis.

I know it's a pain in the backside to do this again so I can understand blowing this off. I spend a large percentage of my professional life dealing with instrumentation in a laboratory so I have some idea of things that can go wrong.

Thanks for the advice. I did see small fluctuations in the torque value as I was balancing but they stabilised once I'd managed to get still.

I agree that using my bodyweight is a bit iffy (I'm also a lab scientist!!) but it seems to me that pretty much anything that would conceivably cause an error (leaning on anything, not having cranks dead level etc) would cause the PT torque value to be lower than expected, not higher. There's no way I am capable of generating greater than the calcluated torque for any length of time (I would need to be instantaneously accelerating upwards.)

Using a range of different weights seemed an obvious good idea but surely just using the same weight in different gear ratios has the same effect? The absolute torque values changed by a factor of x2 from the highest to lowest gear ratio so I've measured the torque/torsion coefficient over a decent range.

I can't help thinking it's kaput 🙁

GettingFaster said:

Thanks for the advice. I did see small fluctuations in the torque value as I was balancing but they stabilised once I'd managed to get still.

I agree that using my bodyweight is a bit iffy (I'm also a lab scientist!!) but it seems to me that pretty much anything that would conceivably cause an error (leaning on anything, not having cranks dead level etc) would cause the PT torque value to be lower than expected, not higher. There's no way I am capable of generating greater than the calcluated torque for any length of time (I would need to be instantaneously accelerating upwards.)

Using a range of different weights seemed an obvious good idea but surely just using the same weight in different gear ratios has the same effect? The absolute torque values changed by a factor of x2 from the highest to lowest gear ratio so I've measured the torque/torsion coefficient over a decent range.

I can't help thinking it's kaput 🙁

You've made a good point about gears and yes, I suppose you'd be more likely to be low - not high - if you weren't getting all the weight on the pedal. Sorry, I wasn't paying attention to all the details you gave.

My original thought was that I spend most of my time at torque values < 90 lb-in at the hub and putting my full weight on the pedals in some gears would produce a torque value greater than that in some gears (I haven't done the math though so don't beat me up on this🙂). Not that it's acceptable to have bad values when sprinting, I was just wondering if it was spread across torque ranges but it seems like you've answered that.

Too bad. 🙁 I'd say call up Saris and tell them the story.

doing a stomp test by standing on the pedal is really unreliable.

don't assume it would only give low readings, because if you use your arms at all you may be applying more torque by leverage.

use some weights even if you have to improvise; i had to get creative to get a heavy test weight, so i put some steel chains in a burlap bag and hung it on the crank.