RPM/s²

[jsmcortina]

Still wondering why RPM/s^2

F=m.a

James

[slow_hemi6]

Agree that rpm already has factors of distance (traveled in 1 rotation) and time (in mins so /60 for S) so that covers Velocity.
Acceleration is deltaV/t so, rpm/s seems to be appropriate for deriving m/s^2.

[piledriver]

I suspect some additional filtration for RPM or vss is needed.
The sensor lags might be sufficient for RPM, haven't tried.
The vss input likely lacks those software filters, so its raw data.

When I first tried using the TQ/HP info in TS it was all over the place, so I stripped out the same RPM data from the logs and pasted it into another dyno "program" that runs in excel or libreoffice... VDR?

Had to play around with its filter a bit but it then produced sane numbers and very consistently.

I suspect vss data really needs a filter to be useful in this way.

[LT401Vette]

I suspect vss data really needs a filter to be useful in this way.

Yes a lot of filtering, especially when using RPM.

[Laminar]

That helps confirm your configuration. You're worried about unreasonable calculation results, but without knowing your configuration it's impossible for anyone else to know if the calc results are reasonable.

[Yves]

voila

[Laminar]

How often does your GPS speed update? Having a delay in speed updating of even half a second will throw off the correlation between speed and RPM and mess up the power calcs.

[Yves]

I don't use a gps for speed, I used the VSS.

[Laminar]

Why is your msq set up for GPS then?

[Yves]

Why is your msq set up for GPS then?

The results with GPS were even worse than vss. No GPS was connected since the summer of 2015.
What's even more strange is that last time I drove while logging speed, I had changed that to VSS. No idea why that changed back.

[Laminar]

So this is or is not your most recent configuration?

[Yves]

Look at the date

[whittlebeast]

I think what you are looking for is something like

([RPM]-[RPM-25])/([Time]-[Time-25])

Here is a screen shot of a dyno pull

http://www.nbs-stl.com/CRX/Dyno%20Sweep ... %20Sec.png

I do the same trick with MPH, KPH or VSS to get acceleration. 20 MPH Per Sec is accelerating hard.

Andy

[jsmcortina]

That's RPM/s (i.e. first derivative of RPM) the OP says he wants RPM/s² (the second derivative)

James

[subwoofer]

Which would be "jerk" or 1/smoothness. Could maybe be useful for drivability tuning, like overrun fuel/spark return etc.

[Yves]

Which would be "jerk" or 1/smoothness. Could maybe be useful for drivability tuning, like overrun fuel/spark return etc.

I don't understand this.

[jsmcortina]

Which part don't you understand?

RPM/s² (RPM dot dot) is likely related to TPS/s (TPSdot)
i.e. the rate of change of acceleration is related to rate of throttle opening.

James

[Rick Finsta]

Yves, you're on speedtalk, right? The derivatives are described in detail in many cam motion discussions there. SchmidtMotorWorks had a great thread going for a while with a program to play with, even.

You have an object with a position.
If it is changing position that change is described by velocity.
If the velocity is changing, that change is described by acceleration.
If the acceleration is changing, that change is described by jerk.

So you are not using acceleration in your proposed calculation, you are using jerk, which describes how the slope of the acceleration rate is changing.