Let’s start over. You said my PM was accurate both before and after the zero-offset I did. In that case, why should I ever zero-offset?
From the article I linked above:
Ignoring temperature, I see the ZO as a necessary step to account for the day to day use and possible equipment changes that happen in the typical use.
As much as we think all the gear we have is rock solid and stable, they experience subtle changes over time from use and even large temperature swings that affect the neutral state of related components. For example, Shane Miller likes to install his meters, do an initial ZO gets some data. Then he does some hard accelerations to “settle” or “bed in” the components, and he redoes the ZO. Even standard riding or addition of curb hops and other input to the bike could lead to shifts in the related components around the meter, and impact results.
The active temperature compensation works from the ZO to maintain general accuracy over a shorter period of time. But you will often hear people recommend to ZO as often as practical. There is no negative impact from doing so, and it can only act to improve the power data results.
This all goes to “good practice” even if it doesn’t make any difference most of the time. It is still the right approach to take and may well keep you from getting junk data at some point in the future when something out of the ordinary could occur.
1 Like
stevemz just lit the candle of a thousand zwift dopers…
The power meter is only ‘accurate’ when the correct offset ‘for the conditions’ is applied to the reported strain.
On the basis that the offset was 7 prior to zeroing the meter and is 12 after, the device cannot report the exact same values before and after ‘in the same conditions’. That is the whole point of the zeroing process.
When a manufacturer states that the power meter is accurate when the zero offset numbers are within a range, that means that the assumptions they have made regarding the effect of residual strain do not affect the validity of the slope of the strain to torque line and that offsetting the line will produce an accurate output. It does not mean that zero offsets within that range have no effect on the output across differing conditions.
As an aside…
Some power meters have no temperature compensation as it is built into the design of the gauges: In a Wheatstone Bridge arrangement of gauges that are symmetrically positioned around the neutral axis of bending of the crank arm, no active compensation is required since the temperature effects all gauges equally and they cancel each other out, just as they do when longitudinal or transverse forces are applied to the arm.
From an engineering point of view, measuring strain in a totally symmetrical crank arm is by far the best solution as it inherently only measures strain that propels the bike forward and can remove the need for active temperature compensation if the correct gauges for the material are selected.
Mike
Skimming the whole thread, it escalated quickly!!
I wouldn’t get too hung up on the calibration numbers displayed on the headunit. They are different for each powermeter, and as long as they are within the range quoted by the manufacturer just get on and ride!
The numbers differ greatly from one powermeter to another. Stages calibration figures are 840 - 940 IIRC, Powertap P1S usually between 7 and 14 (or there abouts). Any other number is usually an error code to troubleshoot. The numbers themselves will only really mean something for the manufacturer in the event of a problem.
Calibrate at the start of each ride and then forget about it.
3 Likes
“…Which one is the correct value in this case?”
Any of the values is going to be more accurate than values obtained days ago, in potentially different conditions.
Not that I’m concerned with accuracy - my original question, which has been answered, was simply which way are the numbers offsetting. I do welcome the fact though that this has morphed into a discussion of when/whether to zero calibrate.
Both.
2 Likes
Just zero offset my stages 10 times in a row and got exactly the same output each time. If you get a range of numbers larger that 1 there may be something wrong with your power meter.
Both can be correct since temperature is not the only thing that can affect the residual strain in the gauges.
Mike
It’s a Stages SI gen 2. The acceptable range is 890 ±50. Typically I see around 885 ±5.
For a Quark meter that range is -1000 to +1000, so a very small change in offset will have virtually no effect on the output values.
Small changes in the zero offset between the start and end of a ride (±50 from the initial offset value according to quark) will not cause the meter to stray outwith the ±1.5% accuracy specification. That does not mean that it is as accurate at the end of the ride as it was at the beginning when the zero offset was carried out.
Mike
I’m not sure what the acceptable range is for the Powertap P1 pedals from the opening post.
Mike
I’m sure we can agree that the reported values will be different but within the stated error of the device.
What we don’t know is whether a change in zero offset between 7 and 12 on a set of P1 pedals is still within the manufacturer’s range of error.
Mike
can I ask you a Q about your power taps ? I used mine for the first time today and cant see how to adjust power smoothing …can you help ?
Smoothing is set on the app or device reading the power meter.
- There is a setting in head units like Garmin and Wahoo. The exact menu varies, but can be found by searching.
- There is a setting in apps like TR.
I have a quarq power meter and changed the chainrings last weekend. I’ve been getting consistent calibration #s between 390-400.
After I changed the chainrings the calibration # is now around 500 for the last two rides.
I have a threshold ride pickering tonight. How would that change in calibration numbers effect he power output reading?
If the calibration was successful then it should have no effect whatsoever 
https://support.sram.com/hc/en-us/articles/360049664794-What-is-a-Good-Zero-Offset-Number-
Summary is you can expect big-ish jumps if you change somethng as long as its stable at the new value.
thanks
So when I was doing threshold work last week with the calibration around 400 but do threshold tonight with the calibration around 500
Is it going to change the feeling; easier, harder, or the same?
It should feel the same - that’s what the calibration achieves.
1 Like