Re: [Paparazzi-devel] Problems getting an overview of Paparazzi

[Christian Merfels]

Yeah, this is an undergraduate group project, but no, novelty is not a
requirement. The thing is, the project was plannend for serveral
consecutive groups, and we are now about the third taking over the work
of the over two groups. That's why already some stuff is bought, and
we'd like to build up on their work. But honestly, right now I don't see
how, as the board they have constructed can not perform the calculations
and is probably not compatible with Paparazzi. I will talk with the
others and our tutor, and maybe we need to redefine the project task.

> Marcus's suggestion is a good one if you have a good level of
> familiarity with Linux and embedded systems.
> 
> Assembling your own tiny is not really 'engineering' (it's just
> difficult to do well) - is this an undergraduate group project, and if
> so is there a requirement to have an element of novelty in the design?
> Assembling a UAV is a much more difficult task than people always
> assume!  I've seen 10-20 undergrad groups who in the past have tried to
> build a Paparazzi-powered UAV and not one of them actually got anything
> working autonomously.
> Paparazzi do not 'support' parts - we just have a list of stuff that we
> know works!  There are a lot of design decisions you will have to make
> alone.
> The makeup of your group will also influence this - if there are more
> 'structures' than 'systems' guys, you might want to think about changing
> the airframe design instead?
> --G
> 
> On Tue, 16 Nov 2010 10:46:51 -0000, Christian Merfels
> <address@hidden> wrote:
> 
>> Hi Gareth,
>>
>> thanks for your reply, it cleared up some things. I managed to get the
>> simulation running in Paparazzi and I read a lot on the wiki (even
>> though i didn't understand everything).
>>
>> So, for a real flight, one would - after configuring Paparazzi - compile
>> the necessary code and flash the board with it, in order to enable the
>> UAV to do all the algorithmic work?
>>
>> Well, if that is how it's done, than one would normally need to use the
>> transmitters, the board and everything else supported by Paparazzi (as
>> suggested by you). But as our task is not to just assemble an UAV - we
>> need to do some engineering work - the only possibility would be to
>> build an own Tiny2 board. Is this correct??
>>
>> Thanks for the warm welcome!
>> Best regards,
>> Christian
>>
>>
>>> None of your current hardware will be much use I'm afraid.
>>>
>>> You need the autopilot unit itself which is called a Tiny.  This goes in
>>> the aircraft.
>>> http://paparazzi.enac.fr/wiki/Tiny_v2#Architecture
>>>
>>> You can't stabilise a UAV via the ground, there is too much latency.
>>> The paparazzi board (Tiny) does all the work, the ground station simply
>>> displays what is going on and allows changing high level flight
>>> parameters (current waypoint etc).
>>>
>>> If this is your first UAV project, forget the gyro - Paparazzi
>>> traditionally uses infra-red level sensors, which are described on the
>>> wiki.
>>>
>>> Have any of you got experience with model aircraft?  You will need
>>> someone who is competent at flying model aircraft.  Your Twinstar should
>>> be fine.  You almost certainly do not want to build your own Tiny board
>>> - I would highly recommend purchasing from one of the suppliers, listed
>>> on the wiki.
>>>
>>> An example of everything you need is here (I'm not promoting this
>>> particular supplier, they are just the only one I'm currently aware of
>>> with a photo of all the stuff you need!):
>>> https://mini.ppzuav.com/osc/product_info.php?cPath=2&products_id=68&osCsid=jug5m75orjn3rvq182qhmq8vfaquggjc
>>>
>>>
>>>
>>> Also, read getting started on the wiki:
>>> http://paparazzi.enac.fr/wiki/Getting_Started
>>>
>>> I would recommend maybe reading through the wiki a couple of times, and
>>> then trying to run a simulation as described in the getting started
>>> section just to see how everything fits together.
>>>
>>> All that said, welcome to the mailing list and to the Paparazzi
>>> community!
>>> --G
>>>
>>> On Mon, 15 Nov 2010 17:03:14 -0000, Christian Merfels
>>> <address@hidden> wrote:
>>>
>>>> Hello together,
>>>>
>>>> I'm an engineering student from Lyon who got currently involved in a
>>>> project of building an UAV. We took over the project from another
>>>> group,
>>>> which already put some work in this. Right now, some hardware was
>>>> already bought:
>>>> - a microcontroller (PIC16F876A, 20Mhz)
>>>> - a receiver/transmitter set (ARM-U8, frequence 868Mhz)
>>>> - a GPS chip (EM406A)
>>>> - a quartz (20Mhz)
>>>> - a gyroscope (IDG 500Dual)
>>>> - a batterie
>>>> - some cables
>>>>
>>>> (As we think that the receiver/transmitter doesn't work with Paparazzi
>>>> (?), we are probably going to change them.)
>>>>
>>>> Our goal is to buy a Twinstar II, build the necessairy card, wire
>>>> everything together and than control the drone from the ground via
>>>> Paparazzi.
>>>>
>>>> Now as I have described the setting, do you guys think this is
>>>> possible?
>>>> I'm not quite sure about the true function and work done by Paparazzi;
>>>> am I correct when I say:
>>>> * Paparazzi does all the aerodynamics calculation
>>>> * It than sends the current speed and direction orders to the plane
>>>> * The card on the plane needs to interpretate the orders and do the
>>>> actual adjustments of control signals / power levels
>>>> * The plane itself sends back GPS / gyroscop signals to Paparazzi
>>>>
>>>> So our main work would come down to the above described interpretation
>>>> of orders. Is our microcontroller sufficient for this?
>>>>
>>>> I'm sorry for this quite lengthly explanation, but I tried to keep it
>>>> short anyways :-). When reading the Paparazzi Wiki, talking to the old
>>>> group and our tutor, we got a lot of different infos about Paparazzi's
>>>> mode of operation + capabilities, so we thought it would be the best to
>>>> ask you guys directly.
>>>>
>>>> Best regards,
>>>> Christian