How long does it take your power to "fall" to zero when coasting?



ridiculous

New Member
Apr 6, 2006
13
0
0
39
How long does it take your Powertap to "fall" from whatever wattage you are putting out to 0 after you stop pedaling?

What about how long does it take your torque to do the same?

Normally, after I stop pedaling, my Powertap will next read some lower wattage number (5-100W), then 0. Torque will read something low too (5-30) then zero. Although my torque also often just hits zero right away for the next reading.

Here's where I start wondering though: sometimes my torque will read: 50-70 (whatever I'm riding at), then I stop pedaling and I see: low torque number (ie: 5-30), 1 (or 2), then 0. Likewise, my power will sometimes read 1W (or 2 or 3, something very low) before returning to zero.

Does the fact that my torque/power sometimes settle on 1 or 2 before hitting zero mean that my torque was drifting? Is this normal and should just be expected (nothing to worry about)?

I've checked my torque functioning and all and the auto-zero does work. But the CPU can't auto-zero while you're pedaling and after doing long extended threshold intervals, if the watts/torque are "falling" back to zero but settling at 1W/1 torque right before hitting zero watts/zero torque (ie: an extra second or two of positive W/torque data), does that mean my watts were off for that interval because of torque drift? Anything I can do about this? I always make sure my torque is zeroed at the start of the interval but you can't do anything while you are doing the interval of course...
 
Depends on the averaging interval that the display is set to (mine is set to 3 seconds and it takes a good 3-4 updates to get down to zero), and also when in the 1.26sec sampling window you stop pedalling.

I didn't read anything alarming in your description of the PM's response. Do you have reason to think that the torque/power readings may be drifting or incorrect?
 
frenchyge said:
Depends on the averaging interval that the display is set to (mine is set to 3 seconds and it takes a good 3-4 updates to get down to zero), and also when in the 1.26sec sampling window you stop pedalling.

OK, yes mine is set to 1 second display and it takes about 2 updates, sometimes 3 to get down to zero.

I didn't read anything alarming in your description of the PM's response. Do you have reason to think that the torque/power readings may be drifting or incorrect?
I don't think so. But my question then is, how do you know when you are experiencing torque drift? Is it that you see a steady power/torque reading even when you are coasting (ie. sustained display of positive power watts/torque while coasting) or that it just takes more updates for power/torque to come to zero (ie. maybe 4-6 updates instead)?
 
ridiculous said:
I don't think so. But my question then is, how do you know when you are experiencing torque drift? Is it that you see a steady power/torque reading even when you are coasting (ie. sustained display of positive power watts/torque while coasting) or that it just takes more updates for power/torque to come to zero (ie. maybe 4-6 updates instead)?
If the power and torque settle to zero within ten seconds or so (assuming you have the display averaging set to 2,3 or 5 seconds as frenchyge pointed out) then you're fine.

You can also check your files when you get home. Make sure your power files show periods of zero torque or export the files into Excel and sort on the torque column to make sure your torque actually goes to zero during your ride. Remember, the display averaging only impacts what is displayed on the bike computer, not what is recorded in the file for later review. So if you have display averaging set to 5 seconds the response will be damped as you increase or decrease power(helpful for maintaining steady efforts like long intervals and time trials) but the recorded data isn't smoothed.

-Dave
 
My CPU displays power over just 1 second. About 70% of the time, it
hits zero after two display updates - this goes for power and torque.
The rest of the time, it will take at most three display updates. It
usually hits some low power/torque number (like 1-5W or +1 torque)
before then hitting zero. Is this "quick" enough or is there cause to
worry?
The other thing to mention is that this is how the data is displayed
but when I look at it in raw numbers after downloading, for the most
part, I see zeroes in torque and power when I'm coasting. Sometimes
though, there are exceptions where there might a positive power
reading with zero torque (at the beginning part of the coast or right
at the end as power is applied again) but that is just for one data
set - and correct me if I'm wrong, this is just because power and
torque didn't record exactly simultaneously together (or something
along those lines)?
 
ridiculous said:
... Is this "quick" enough or is there cause to
worry? ...
Your power meter is perfectly fine. That extra 1 watt or 2 watts for one or two sample periods on the display means nothing relative to the accuracy of the power meter and as you say it's only on the bike computer display and not in the actual data. You're splitting very fine hairs with this one. Think about what that extra 1 or 2 watts means relative to say an average power of 200 watts. It's really not important and most likely related to the difference between the PowerTap sampling interval and the computer display update rate.

If you've validated periods of zero torque in the recorded file and the display regularly reaches zero within a few seconds then forget about it, it's not a problem.
 

Similar threads