Raven Altimeter Raw Data Interpretation

[jyawn]

I'm working with a university group who flew with a Raven last weekend, and they want to do more with the data than the FIP program allows. We have captured the raw flight data, but it is not obvious what the numbers mean. is there a reference somewhere? I suppose it could be reverse-engineered, but being "ADD-compliant," that would take more attention than I have left.

 

[Buckeye]

I'm working with a university group who flew with a Raven last weekend, and they want to do more with the data than the FIP program allows. We have captured the raw flight data, but it is not obvious what the numbers mean. is there a reference somewhere? I suppose it could be reverse-engineered, but being "ADD-compliant," that would take more attention than I have left.

Please elaborate. The Raven measures acceleration, barometric pressure, voltage, and temperature (I think). The FIP then generates more useful plots (like altitude, integrated velocity from acceleration, etc.) from the basic measurements. University students should have some sense of this data, right?

This may help:

https://www.rocketryforum.com/showthread.php?42957-Post-your-Raven-FIPa-files-here&highlight=fipa

 

[mikec]

You know about saving data from FIP into spreadsheet format, right? To quote Adrian: "you can right-click in the parameter selection area of the data graphing window and you'll get options to save the selected data to a file or copy it the clipboard, in which case it can be conveniently pasted into an Excel spreadsheet."

It's not that obvious in FIP and easy to miss.

 

[Buckeye]

Oh, yeah. The data export is probably what you are looking for. The Raven data is very complete and with lots of significant digits!

 

[jyawn]

Thanks to mikec and Buckeye for the replies... I had not been aware of the right-click option in the FIP program. That may indeed give us what we need.

But I'm stil curious about the true raw(ish) data available in the "raw" tab. I presume these are the numbers straight out of the hardware, from which the FIP program interprets acceleration, etc. Each "Frame#" seems to represent one sample at 20Hz, and the X, Y, X probably refer to accelerometer output, but I really have no clue so am wondering if there is a reference to the meaning of these bits.

What we are looking to do is to get a "thrust curve" for the motor from in-flight data. The ARTS software does that, but the curve from the ARTS we used does not closely agree with published thrust curves for our motor (J410) We presume that ARTS curve to be in error, since it disagrees with the four static test curves we have, one from CTI, two from ThrustCurve, and one we did ourselves, all of which agree pretty closely. The students would like to have other accelerometer data for comparison, as they would like to be able to explain the difference to the NASA SLI folks. Thus the Raven. Unfortunately, it may be a moot point, as I inadvertently gave them an old Raven that had been damaged in a CATO, and was flaky, and did not capture data during that flight. But they may want to try again on their next flight, so it would be good to have this capability in the future.

 

[mikec]

But I'm stil curious about the true raw(ish) data available in the "raw" tab. I presume these are the numbers straight out of the hardware, from which the FIP program interprets acceleration, etc. Each "Frame#" seems to represent one sample at 20Hz, and the X, Y, X probably refer to accelerometer output, but I really have no clue so am wondering if there is a reference to the meaning of these bits.

I've partially reverse-engineered these, but other than the resolution and sampling rate of each data type, there's nothing there that you won't get more easily from the data export.

Axial acceleration is sampled at 440 Hz, radial at 220 Hz, current at 40 Hz, and everything else at 20 Hz (I believe, if you look around there are some threads where this is discussed.)
Acceleration raw units are 0.01 Gs, barometric pressure in 0.00004 atm, voltage in millivolts.

Note that many of the newer Ravens have scaling problems with their accelerometers, so this might not be the best device for accurate thrust curve determination.

 

[Adrian A]

Thanks to mikec and others who pointed out the way data can be cut and pasted from the FIP to Excel. The pasting operation gives each measurement its own time reference column, too.

Note that many of the newer Ravens have scaling problems with their accelerometers, so this might not be the best device for accurate thrust curve determination.

The Ravens accelerometer scaling is the same as it has always been, which is subject to errors during calibration caused by non-linearity of the A/D converter built into the Zilog microcontroller. The non-linearity can cause the slope measured during +/- 1G calibration to be different from the slopes measured outside of that range. Accelerometer error on the order of 10%, and sometimes worse, is possible. When I did a lot of drag analysis using Raven data a few years ago, I used the baro data as a reference for adjusting the scale and offset during post-processing of the accel data, to get the accel data to match the baro data.

 

[jyawn]

I've partially reverse-engineered these, but other than the resolution and sampling rate of each data type, there's nothing there that you won't get more easily from the data export.

Axial acceleration is sampled at 440 Hz, radial at 220 Hz, current at 40 Hz, and everything else at 20 Hz (I believe, if you look around there are some threads where this is discussed.)
Acceleration raw units are 0.01 Gs, barometric pressure in 0.00004 atm, voltage in millivolts.

Note that many of the newer Ravens have scaling problems with their accelerometers, so this might not be the best device for accurate thrust curve determination.

Thanks for the clues... I'll look at he raw raw data again and see if we can possibly use it. More likely, I'll run this by the student group, some of whom are majoring in reverse engineering : ) But I believe that they will prefer to use the detailed right-click data, apparently obtained at a much higher sample rate.

 

[Adrian A]

Thanks for the clues... I'll look at he raw raw data again and see if we can possibly use it. More likely, I'll run this by the student group, some of whom are majoring in reverse engineering : ) But I believe that they will prefer to use the detailed right-click data, apparently obtained at a much higher sample rate.

The right-click data is the raw data, just conveniently scaled and formatted.

 

[jyawn]

The right-click data is the raw data, just conveniently scaled and formatted.

Adrian: Thanks for the assurance - it's now clear that we will not likely get better results from the Raw tab stuff, and will concentrate on the right-click data. Thanks too for providing us such an an excellent data logger for our flights, disguised at the Raven.

Jimmy