Tss keep increasing when not pedaling

tss increased by 0.1, check following pictures, I was not pedaling.

Are you stopping and resuming your activities or use auto-pause? Are you including zeros when computing the average power (there is a setting for that)?

Training Stress Score computation involves Normalized Power (kind of a rolling average of your power which additionally boosts changes in this rolling average) and Intensity Factor (a relation between your Normalized Power and Functional Threshold Power).

Ideally, I agree that when you get off your bike, TSS should not change. But just from a power timeseries perspective, there is no reliable way to distinguish whether you're off the bike/trainer sipping coffee, or just coasting downhill fighting for your life between intervals. Well, you can use other input like your speed when available, which is where the auto-pause is coming. Or you can manually indicate when you are taking a break. Make sure your timer stops when you stop your workout. If the TSS was to keep increasing while your workout is stopped, I'd consider this to be a bug.

I am not exactly sure if including/excluding zeros in power averaging affects NP/IF computation, but just make sure that you include (rather than ignore) zeros if you want to properly account for coasting in your statistics.

What you say really make sense! I will try to include zero in the settings of my 1040, and after this weekend give feedback here.Thx.

I found the include/exclude setting in my 1040, and it is including zero by default! So...This might possibly a bug of 1040.

TSS will increase if the timer is running even if power is zero. There are long discussions on this on the Web, but the formula uses the activity time in the calculation.

TSS =((Activity time X Normalized Power X Intensity Factor)/(FTP X 3600)) X 100

NP on the Edge includes zeros. There is no way to exclude them from the NP calculation. The exclude zero was an option just for the average power calculation.

I played a bit with the formula and I think there is an easy way to explain the observed effect.

T - Activity Time
NP - Normalized Power
IF - Intensity Factor
AP - Average Power
TP - Total Power
C - Time Constant (1/36 s)

According to TrainingPeaks, the TSS formula is actually:

TSS = IF^2 * NP / FTP * T / 3600 * 100

Remember that:

IF = NP / FTP

Let's define:

TP = AP * T

Let's simplify by replacing Normalized Power with Average Power. For fixed power it's the same, if there are variations NP will be slightly higher but close.

Then:

TSS = C * T * NP * IF^2 / FTP ~= C * T * AP^3 / FTP^3 = C / FTP^3 * TP^3

Now if you are not producing any power, your TP is constant along with your FTP and C. Which means TSS should be constant, except NP will normally be slightly higher than your AP to compensate for power variations. So I guess for a little while your TSS will be growing if your power was higher just a while ago.