Unstable & Spiking Data (using FSUIPC 3.4.4.0)???

[InMotion]

Peter,

I have a customer in South Africa that purchased a motion simulator from us a few months ago and he's using FS2004 with FSUIPC. We noticed the ptich (Z axis) acceleration data was not behaving correctly in the simulator. After I looked at the Z axis data from FS2004/FSUIPC, and I tested several addresses ($3070, $3078, $31D0), I noticed there were spikes in the data. This is obviously a problem with a motion simulator. The motion will act "jerky" because of the spiking.

I then tried to write my own acceleration data using [delta v / delta t] and unfortunately, the velocity data was also spiking and couldn't use it.

I ended up using a homemade formula for acceleration that doesn't spike. But I would like to know if there is a way to prevent the spiking. I don't see the spiking problem with the other axis'. Thanks!

Regards,

Mark Barry

InMotion Simulation

928.759.0631

http://www.inmotionsimulation.com

[Pete Dowson]

After I looked at the Z axis data from FS2004/FSUIPC, and I tested several addresses ($3070, $3078, $31D0), I noticed there were spikes in the data. This is obviously a problem with a motion simulator. The motion will act "jerky" because of the spiking.

I then tried to write my own acceleration data using [delta v / delta t] and unfortunately, the velocity data was also spiking and couldn't use it.

I ended up using a homemade formula for acceleration that doesn't spike. But I would like to know if there is a way to prevent the spiking. I don't see the spiking problem with the other axis'. Thanks!

Sorry, I really have no idea why FS would do that. FSUIPC is only passing on the information it gets. In fact I think that those locations are mapped directly into FS's own data areas.

You mention 3070, 3078 and 31D0 and call them "pitch" accelerations, but the Z axis in FS is the longitudinal one -- forward/backward. The FS team name them in terms of screen-type orientation, Y up/down, X left/right, Z forward/back. Both 3070 and 31D0 are Z-axis, not pitch values.

3078 is a pitch acceleration. Is that the one you mean?

Maybe you need to add some form of digital filtering into the values to smooth out the spikes?

Regards,

Pete