Marsa Systems Tilt Gadget Preview

The Rocketry Forum

Help Support The Rocketry Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

jderimig

Well-Known Member
TRF Sponsor
Joined
Jan 23, 2009
Messages
5,989
Reaction score
4,097
At BALLS 27 the Marsa Tilt gadget (Marsa Gyro1) will be flight tested. Assuming no major issues this will be available as add on option for Marsa54L, 54LHD and 33LHD altimeters

The Gyro1 uses two 3-axis gyros so that optimum rotation rate resolution can be obtained for the spin axis which can happen at high rates and the yaw and pitch axes which happen at a very low rate. In addition the tilt gadget has a 100K barometric sensor to backup and extend the range of the native barometric sensor on the main board of the altimeter.

The Gyro1 uses an ARM processor running at 48mHz so that the processing of the rotation mathematics is dedicated on the gadget and not burdening the altimeter processor.

(Click photo to enlarge)
marsa_gyro1.jpg

The Gyro1 is connected to the expansion header of the Marsa altimeter. Once connected the Marsa firmware recognizes and configures itself for tilt-aware processing and event options.

Programming of tilt-aware events is done on the Marsa altimeter user interface. For any channel which is programmed at a motor timer tilt criteria can be added to the logic of that channel event firing.

gyro_setup_screen.jpg
The first line sets the maximum allowable tilt angle for the timer channel to fire.

For flight profiles where a coast down to a target speed is desired, this maximum speed can be set as shown on line 3 of the display. Two options can be applied to this speed criteria.

The "Try" option will delay firing the channel until the speed drops below the target speed OR the maximum tilt angle is reached, whichever happens first.

The "Require" option requires the target speed to be reached before the channel fires. If at that time the tilt angle exceeds the allowable value then no event will occur.

Finally the "Ignore" option ignores the speed criteria for the event and the event will fire at the specified timer delay as long as the angle criteria is satisfied.

The goal with this product as with all Marsa products is to make tilt event programming as easy, straightforward and reliable as possible.

More to come, stay tuned.
 
Cool

I have just begun reading about Kalman filtering and combining multiple sensors to determine the most likely value.

Care to share what information is used to determine orientation? Is it just the gyroscopes?

Care to make any comparisons to a sensor fusion device like this: https://learn.adafruit.com/adafruit-bno055-absolute-orientation-sensor/overview

Thanks for producing great products!!!
 
Last edited:
Cool

I have just begun reading about Kalman filtering and combining multiple sensors to determine the most likely value.

Care to share what information is used to determine orientation? Is it just the gyroscopes?

Care to make any comparisons to a sensor fusion device like this: https://learn.adafruit.com/adafruit-bno055-absolute-orientation-sensor/overview

Thanks for producing open product!!!

Only the 2 gyroscopes (6 axes of them) are used to determine orientation. In a rocket flight there isn't a gravity reference to fuse like more terrestrial IMUs use.
 
+1 product desired/needed in HPR community. Please keep us updated on your results at BALLS 27 and tentative release date.

Fred, L2
ICBM member
KG4YGP
 
Aside from dual stagers, I suppose this could be used for delaying air start lights for maximum altitude correct?
 
Initial test at MWP16 today very good. Did a flight with a VMAX to stress test acceleration sensitivity of the gyro accuracy. There was no shift in the gyro outputs at the 0 to 20g and 20g to burnout transitions. More in depth data analysis needs to be done but it looks like the product release is on track.

Plan to do more test flights next weekend.
 
Last edited:
Very nice...

When I first developed the RockeTiltometer, I considered adding in additional components but decided it was in the best interest to get it out quickly to market and so relied on integration with 3rd party flight computers, specifically in my own personal use case with the Raven to achieve altitude and velocity trigger options for staging and/or adverse-tilt conditions...

It is great to see the technology continue to evolve...high rotation effects can be disturbing...in the case of the second iteration of the RockeTiltometer, I added a magnetometer to offer some help to deal with high spin rates...processing for the gyro conversions was also obviously off-board, but I also used DCM rotation-matrix calcs rather than Kalmann filters for greater processing speed/accuracy and null-point avoidance (though cycles per second are no longer much of a cost issue!)...

My initial goal was toward increasing staging intervals by allowing comparisons of tilt and velocity bleeddown, but the safety/easier-retrieval aspects of adverse-angle-abort ultimately became more significant...

By providing such easy access to the various conditional triggers is quite elegant...

Good work John!
 
Frank, thank you so much for your post! I stand on you shoulders on this project to follow the ground breaking work you achieved with the tiltometer. I appreciate your encouragement

The gyros are pretty good these days but there is the issue of acceleration cross axis sensitivity with the low cost consumer level chips. Hopefully I can contain that to less than a few degrees of error. Hence the flight tests that are ongoing.
 
John, if you are looking for a launch site TMO starts its season on 11/10, providing the weather co-operates.
 
Back from a great trip to MWP16. Great launch and great time.

Here is a plot of the data from a test flight. Flew a WM Darkstar on a K2000 Vmax. Motor was chosen to provide big acceleration to look for any tilt transients in the data. The motor hit hard for 1.2second at 25g.

Graph is shown below. No evidence of an acceleration effect. Tilt is relatively stable until the apogee turnover which is apparent in the data.

tilt_mwp.PNG
 
Will the Marsa Tilt gadget interfere with the ability to use the Marsa Net Gadget, or will you be able to connect both gadgets?
 
Will the Marsa Tilt gadget interfere with the ability to use the Marsa Net Gadget, or will you be able to connect both gadgets?
There will be 2 versions, this one and a wireless Mnet version. So if you do not need the Mnet the wired config will be simpler.
 
How can I buy one?

I don't see purchase options or price on the website.

I've been blind and oblivious before though.
 
Thanks John, I wasn't seeing it on my mobile browser... or maybe you added it in between browsing at the desk, to browsing at home.
 
The gyros are pretty good these days but there is the issue of acceleration cross axis sensitivity with the low cost consumer level chips. Hopefully I can contain that to less than a few degrees of error. Hence the flight tests that are ongoing.

What is its sensitivity to misalignment between the sensor and vehicle roll axis?
 
Does that manifest itself as an acceleration sensitivity?

If they are not aligned perfectly (for some value of perfect) then roll will appear in the pitch and yaw axis data. Probably not a big problem at low roll rates but since you are using two sensors specifically to support high roll rates...
 
Back
Top