Raven Accel Alt Data way off

Launched a couple of EX motors today (2gr 54mm red), pretty mild motors with about 3sec burn ~J420.
According to my Raven 2 the Baro alt was ~3700ft, but the Accel Alt shows a max of ~400ft with a starting -5875.

All the other prior flights (last time I few was at Red Glare in Nov) the Baro and Acc are very similar, do you guys think this indicates a problem or just that I need to go through the calibration procedure?



View attachment 01-20-2013 Like Smoke EX J420R 2nd.FIPa

The problem appears to be a combination of launch detect timing and post processing.

The accel data does not have any pre-launch values and the summary reports an average prelaunch value of over 3G's. Apparently the GUI is subtracting that from the data before running its own numeric integration. Which causes the insane result. Why it doesn't use the offset measured by the altimeter I have no clue.

UhClem said:

The problem appears to be a combination of launch detect timing and post processing.

The accel data does not have any pre-launch values and the summary reports an average prelaunch value of over 3G's. Apparently the GUI is subtracting that from the data before running its own numeric integration. Which causes the insane result. Why it doesn't use the offset measured by the altimeter I have no clue.

Click to expand...

That's mostly accurate. The Raven's prelaunch accel average is recorded into memory at the start of the flight, and it is used on-board for the velocity calculations. The FIP should use that value rather than attempting to calculate its own version since it may not have enough data available in the case of a slow start like this.

Zephhyr - Sorry about the confusion; yes it an issue with the GUI and not the altimeter or it's calibration.

David - You described the GUI issue correctly. Looking at it now, I don't think I have an answer for "Why it doesn't use the offset measured by the altimeter I have no clue." ... .... since FIP was written to also support the Parrot, perhaps I didn't have that value available from flash memory so had to calculate it myself and it just carried over to the Raven code as well (where the pre flight value is stored in flash)...

Adrian - I'm changing it such that if the pre-launch axial offset is known, then it uses it; else it calculates it's own (maybe needed for remaining Parrot users?). I have my code updated and the graph is not insane anymore (;-) and will work on an updated Devel build for us to test with.

Thanks!

How did you mount the altimeter in your rocket, was it able to move around a great deal?
If it's only a 2G case then, there is no way it's a J 420.
770 Ns / 3 gives you a J 256 or there a bouts....

JD

Zephhyr said:

Launched a couple of EX motors today (2gr 54mm red), pretty mild motors with about 3sec burn ~J420.
According to my Raven 2 the Baro alt was ~3700ft, but the Accel Alt shows a max of ~400ft with a starting -5875.

All the other prior flights (last time I few was at Red Glare in Nov) the Baro and Acc are very similar, do you guys think this indicates a problem or just that I need to go through the calibration procedure?

View attachment 114133 View attachment 114134

View attachment 114132

Click to expand...

The alt is mounted very solid, no movement at all, you can see the e-bay in my build thread, https://www.rocketryforum.com/showthread.php?25892-First-post-amp-Build-thread , got rid of the G-Wiz in favor of the Raven & Perch.

Ya I miss-typed it a few times before I realized it, and just went with it, I never actually calculated the Ns, Burn-Sim predicted J240 :wink:

JDcluster said:

How did you mount the altimeter in your rocket, was it able to move around a great deal?
If it's only a 2G case then, there is no way it's a J 420.
770 Ns / 3 gives you a J 256 or there a bouts....

JD

Click to expand...

Here's a flight from a few years ago:
https://www.mdra-archive.org/photos/esl82/BobUtley/Full/jdavenport20305.wmv
It was suppose to be a J 425... ...more like a J 200.


JD

Zephhyr said:

The alt is mounted very solid, no movement at all, you can see the e-bay in my build thread, https://www.rocketryforum.com/showthread.php?25892-First-post-amp-Build-thread , got rid of the G-Wiz in favor of the Raven & Perch.

Ya I miss-typed it a few times before I realized it, and just went with it, I never actually calculated the Ns, Burn-Sim predicted J240 :wink:

Click to expand...

kjs said:

Adrian - I'm changing it such that if the pre-launch axial offset is known, then it uses it; else it calculates it's own (maybe needed for remaining Parrot users?). I have my code updated and the graph is not insane anymore (;-) and will work on an updated Devel build for us to test with.

Click to expand...

Thanks, Kevin. I think you're right about the origin of this with the Parrot.

Yes I was thinking back to when I was deriving the original algorithms from the Excel spreadsheet you had and there was only 'data' there - none of the state of the altimeter at flight time (which the Ravens have)...

If anyone wants the Devel build that has this fixed in it, they can download the stub installer from: https://www.here-and-beyond.com/FeatherweightIP/FIPSetup-DEVEL.exe - note that after install, it will immediately tell you that a later version exists and you should agree to the update as it is the update that has the fix in it. After it is installed, it will create a separate icon for FIP-Devel that can be run independently from FIP Production. I'm getting a little time to work on this again now so may make some more changes before moving it to production...

Thanks!

As an aside, I was thinking about changing the summary page to be a text formatted output instead of that grid control - with one advantage being that it is easier to copy / paste the results into an email / etc.

Any thoughts?

kjs said:

David - You described the GUI issue correctly. Looking at it now, I don't think I have an answer for "Why it doesn't use the offset measured by the altimeter I have no clue." ... .... since FIP was written to also support the Parrot, perhaps I didn't have that value available from flash memory so had to calculate it myself and it just carried over to the Raven code as well (where the pre flight value is stored in flash)...

Click to expand...

Why am I reminded of the firmware bug that destroyed the first Ariane V?

UhClem said:

Why am I reminded of the firmware bug that destroyed the first Ariane V?

Click to expand...

That is an interesting wiki read - thanks! Luckily the FIP stuff in this thread only has to do with the display of the data post flight so no firmware flight risks... but I believe the devel build now correctly supports the Raven and remains backwards compatible with the Parrot... There are a couple things more I am looking into before making it the 'production' release.

Thanks!

Thanks for the info and update to the FIP.

I think I like the text page format over the table format.

kjs said:

As an aside, I was thinking about changing the summary page to be a text formatted output instead of that grid control - with one advantage being that it is easier to copy / paste the results into an email / etc.

Any thoughts?
View attachment 114805

Click to expand...

Thank you sir, may I have another? (IOW, "YES please!")

sschnege said:

Thank you sir, may I have another? (IOW, "YES please!")

Click to expand...

Scott, glad you like it! The link to the devel build earlier in this thread has it formatted that way so you are free to download it and play with it. Longer term I am thinking of a better 'print' that outputs relevant graphs and the summary in a more 'binder ready' format.

Thanks!

When I try to install the FIP/devel build, Norton flags it as a Trojan virus??

Discovered:
January 20, 2011
Updated:
October 9, 2012 5:08:10 AM
Also Known As:
Bloodhound.Sonar.9 [Symantec]
Type:
Trojan, Virus

I take it this is just a mix-up by Norton?

Zephhyr said:

When I try to install the FIP/devel build, Norton flags it as a Trojan virus??

Discovered:
January 20, 2011
Updated:
October 9, 2012 5:08:10 AM
Also Known As:
Bloodhound.Sonar.9 [Symantec]
Type:
Trojan, Virus

I take it this is just a mix-up by Norton?

Click to expand...

Bloodhound.Sonar.9 means that it was "detected" by the heuristic engine of the scanner. Heuristics are designed to catch yet unknown threats, but they are prone to false alarms.

This is somewhat similar to a security guard who has no idea who you are, but who thinks that you look or behave suspicious.

Reinhard

I'll try to look at it this evening but started a google for "Bloodhound.Sonar.9" and google started auto filling in "false positive"... I saw this before on FIP for a different build and they confirmed at the time that it was false. I was disappointed in their method which tends to 'target' small app development as they don't see them very often... at the time, I wrote this https://www.here-and-beyond.com/FeatherweightIP/DisablingNortonSonar.htm to document how to disable their 'sonar'... [but I will check to be sure this evening.]

Thanks!