MPHV2 on UDB3 DCM problem

[stev...@live.cn]

Hi all,

I have set up the telemetry with UDB3 and ground station using
happykillmore's software. I previously play with Matrix Pilot 2.5.1
and the DCM sent from UDB3 looks fine. but when I loaded MPHV2
firmware onto UDB3 the DCM has some strange behaviors, e.g after I
swung the board and put the board back to its original position, the
gesture displayed on the groundstation software show some yaw drift
and sometime the roll drift as well. I investigated the rmat.c but
havn't notice any thing that could cause this problem. Could someone
help on this? thanks.

Regards,
Steven

[Marcus Fahlén]

How much drift do you have? (deg / min)

/ Marc

[John McClelland]

Steven

MPH V2 is not set up for the current telemetry formats. It is based on an
older version of MP.

I am preparing to release V3 soon (working on it now). I could look at
updating the telemetry for this as part of the release.

John

[John McClelland]

Which output format do you select in MP 2.5 to make your groundstation work?

----- Original Message -----
From: <stev...@live.cn>
To: "UAVHeliBoard" <UAVHel...@googlegroups.com>
Sent: Sunday, January 23, 2011 6:23 AM
Subject: MPHV2 on UDB3 DCM problem

[stev...@live.cn]

I set both Serial_UDB as the formats.



On Jan 24, 1:20 am, "John McClelland" <mcclelland.j...@gmail.com>
wrote:

[John McClelland]

Well just remember MP-H isn't using GPS. Without a magnetometer there is no
yaw lock. HH uses a magnetometer to get yaw lock. This will be available
in V3. Also, MP-H has a larger value of FILTERSHIFT than MP. FILTERSHIFT
is used to filter the gyros and accels. This effectively damps the
response. This was put in to better damp out vibration effects in the heli.
One of my TODO tasks is to lower FILTERSHIFT (and adjust the gains) to try
to get a faster response, while still avoiding vibrational effects. The new
Invensense gyros can tolerate higher vibrations, but tests are needed in a
flying situation to check it all out.

On the bench you could try running a lower filtershift (it is 6 in MPH and 3
in MP) and look at the respnse of the rmat elements to a rotation or
translation. I would not try to fly with a low FILTERSHIFT until you
carefully tune in the gains. You can get exponentially growing oscillations
right at take off. Not a nice experience. Typically you need lower gains
for lower FILTERSHIFT.

John

[stev...@live.cn]

Hi John,

Thanks. I did forget the MPHV2 does not use GPS signal. I tried to
lower the FILTERSHIFT to 3 and hoping for a correct response from the
UDB3 but just getting the same situation. The yaw kept increasing even
when the board is at rest for few minutes. Roll and Pitch still not so
stable...

Regards,
Steven



On Jan 24, 9:04 am, "John McClelland" <mcclelland.j...@gmail.com>
wrote:

[John McClelland]

Steven

It would help to see the response data you reference, both for MP and MPH
with the same units for the response and time axis. Are you looking at
rmat values or some visual indication of the attitude? I also assume you
are doing this on the bench and not flying it.

If you can send some plots or data files, I might be able to help you.
Include your options.h file.