Development progress, bad news, good news
5 views
Skip to first unread message
Crashmatt
Nov 28, 2010, 4:55:08 PM11/28/10
to RC servo interface
Bad news: With the CAN interface and all the extra channels, the UDB
is running out of memory
Good news: The next version of the code moves the mixers to the
interface, reducing the resource requirements on UDB.
Bad news: The UDB no longer has all the radio input information so
telemetry of radio inputs will not work
Good news: The interface board will soon support the telemetry
Bad news: This is all going to be more confusing and more difficult to
use than before.
Good news: There are ways to fix this but it will take a little more
time (Documentation update to follow)
Bad news: Some careful testing needs to be done to prove airworthiness
Good news: This all works with HILSIM so the testing should not be
too difficult.
is running out of memory
Good news: The next version of the code moves the mixers to the
interface, reducing the resource requirements on UDB.
Bad news: The UDB no longer has all the radio input information so
telemetry of radio inputs will not work
Good news: The interface board will soon support the telemetry
Bad news: This is all going to be more confusing and more difficult to
use than before.
Good news: There are ways to fix this but it will take a little more
time (Documentation update to follow)
Bad news: Some careful testing needs to be done to prove airworthiness
Good news: This all works with HILSIM so the testing should not be
too difficult.
Peter Hollands
Nov 28, 2010, 4:58:11 PM11/28/10
to rc-servo-...@googlegroups.com
Hi Matt. That sounds like good and useful progress. Thanks for the update. Pete
Crashmatt
Nov 29, 2010, 8:40:57 AM11/29/10
to RC servo interface
An alternative way to do this is to run all of the MP code on the CAN
interface.
UDB becomes a remote GPS connection and IMU/ADC board.
CAN would transport raw GPS UART and ADC filtered samples.
This would be the easiest scheme for users.
Matt
On Nov 28, 10:58 pm, Peter Hollands <peter.holla...@gmail.com> wrote:
> Hi Matt. That sounds like good and useful progress. Thanks for the update.
> Pete
>
interface.
UDB becomes a remote GPS connection and IMU/ADC board.
CAN would transport raw GPS UART and ADC filtered samples.
This would be the easiest scheme for users.
Matt
On Nov 28, 10:58 pm, Peter Hollands <peter.holla...@gmail.com> wrote:
> Hi Matt. That sounds like good and useful progress. Thanks for the update.
> Pete
>
aero-analysis
Dec 7, 2010, 12:00:52 AM12/7/10
to RC servo interface
Hi Matt,
First, I would like you to know that I appreciate all the hard work
that you are investing into your product. I have CAN board #5 and I
am excited to use it! I have been holding off on using the board
until the code reaches some what of a baseline. I am very
inexperienced with hardware, and the lessons learned from you and the
others here in this group, will help me greatly. My objective, or
goal, with your board is to use an Air Data sensor to obtain more
accurate airspeed and altitude, and provide the UDB navigation
equations with a separate air speed and altitude reference which is
mathematically orthogonal, enabling me to write a simplified Kalman
Filter. What plans do you presently have of supporting this sort of
external hardware? I know that the UDB4 has plans to support this,
but I would like to begin my development on this filter as soon as I
can. This is similar to the work that I do in the aerospace industry,
and I would like to try this new algorithm on smaller and simpler
equipment.
Regards,
Paul Roysdon
First, I would like you to know that I appreciate all the hard work
that you are investing into your product. I have CAN board #5 and I
am excited to use it! I have been holding off on using the board
until the code reaches some what of a baseline. I am very
inexperienced with hardware, and the lessons learned from you and the
others here in this group, will help me greatly. My objective, or
goal, with your board is to use an Air Data sensor to obtain more
accurate airspeed and altitude, and provide the UDB navigation
equations with a separate air speed and altitude reference which is
mathematically orthogonal, enabling me to write a simplified Kalman
Filter. What plans do you presently have of supporting this sort of
external hardware? I know that the UDB4 has plans to support this,
but I would like to begin my development on this filter as soon as I
can. This is similar to the work that I do in the aerospace industry,
and I would like to try this new algorithm on smaller and simpler
equipment.
Regards,
Paul Roysdon
Crashmatt
Dec 7, 2010, 12:48:11 AM12/7/10
to RC servo interface
Paul,
I think I counted you twice when I did board orders. This answers
much. Good news is we now have a spare board.
I am heading in exactly the same direction as you and I need your help
on the software.
There is a baseline software in the gentlenav repository. This will
do servo and radio interface only.
It is good software to get you started with the hardware
It is useless for long term development.
The next version of software will support the extra hardware.
I have had the following running:
MPXV5004 differential airspeed sensor
SCP1000 barometer on SPI, (sensor now obsolete but available)
HMC5843 Magnetometer
The BMP0805 I2C barometer looks like a good alternative to SCP1000.
The I2C magnetometer sensor code takes over the I2C port. It needs
splitting into port and sensor driver to allow another sensor to be
added.
At the moment, only the magnetometer has support for data transfer to
the IMU.
The airspeed and barometer need calibration routines writing for them.
I suspect that you can help with this.
Kalman style auto calibration would be very good.
Whatever data melding process you implement, it goes in, or is called
by functions in rmat.c
There is a similar conversation here about combining data from the
sensors
http://diydrones.com/profiles/blogs/improved-altitude-estimate?xg_source=activity
Bad news is that adding the hardware is not so easy. I will try to
make this as painless as possible,
The triple axis magnetometer breakout from DIYdrones is the easiest to
use. It has an onboard voltage regulator and signal level translator.
Thats all for now
Matt
I think I counted you twice when I did board orders. This answers
much. Good news is we now have a spare board.
I am heading in exactly the same direction as you and I need your help
on the software.
There is a baseline software in the gentlenav repository. This will
do servo and radio interface only.
It is good software to get you started with the hardware
It is useless for long term development.
The next version of software will support the extra hardware.
I have had the following running:
MPXV5004 differential airspeed sensor
SCP1000 barometer on SPI, (sensor now obsolete but available)
HMC5843 Magnetometer
The BMP0805 I2C barometer looks like a good alternative to SCP1000.
The I2C magnetometer sensor code takes over the I2C port. It needs
splitting into port and sensor driver to allow another sensor to be
added.
At the moment, only the magnetometer has support for data transfer to
the IMU.
The airspeed and barometer need calibration routines writing for them.
I suspect that you can help with this.
Kalman style auto calibration would be very good.
Whatever data melding process you implement, it goes in, or is called
by functions in rmat.c
There is a similar conversation here about combining data from the
sensors
http://diydrones.com/profiles/blogs/improved-altitude-estimate?xg_source=activity
Bad news is that adding the hardware is not so easy. I will try to
make this as painless as possible,
The triple axis magnetometer breakout from DIYdrones is the easiest to
use. It has an onboard voltage regulator and signal level translator.
Thats all for now
Matt
Crashmatt
Dec 7, 2010, 2:13:29 PM12/7/10
to RC servo interface
Just to be clear, the magnetometer code was designed and built by Bill
Premerlani. I am not capable of such masterpieces.
> sensorshttp://diydrones.com/profiles/blogs/improved-altitude-estimate?xg_sou...
Premerlani. I am not capable of such masterpieces.
Reply all
Reply to author
Forward
0 new messages