a SoR project for all

[benji]

Has anyone gotten started on their article yet?

not me ,, now im  very busy with my final year project (a hexapod) and the exams are 2 weeks from today
so ill do all the SoR stuff once i finish my exams

[TrickyNekro]

Me nop... five days to exams...

Also, if samples from analog arrive I'll be doing a autopilot system...
Get a hold on...


Cheers,
Lefteris

[superchiku]

me also not started working for the $200 robot contest , trickynekro autopilot using what??

[TrickyNekro]

me also not started working for the $200 robot contest , trickynekro autopilot using what??

A dual rotor coaxial helicopter...

It's a little bit fat dream but with work I'll get to it...
I already have some gyros and I wait for the accelerometers...
Also I'm thinking of buying STK600 or AVR dragon... Probably the STK...
Although I haven't searched for pricing for the STK1000 it's probably gonna be too expensive...
And I'm not interested in AVR32 yet... I want to see what the AVR16 (ATxMEGA series) got to offer...
Surely the 32MHz speed well suit my needs for the autopilots since I saw autopilot with adruino...

And I also need a oscilloscope...
A guy can give me his old Tektronix for 500E... It's still a very good price and it's in good shape too... (I know the guy
and he hasn't really used it too...) (and he said that particular model was used in US military some years before...)
It's at 100MHz so I still think it's good... perhaps I'll lower the price too... I'm tough at trading...

And I will also intro myself to Eagle 5.0. There aren't many differences but it's hell more pleasant to use...

Also I'm thinking of buying a drill and a UV lamp for my boards...

You see if I get to university I'll do some serious business and I'll really need many stuff like this...
So... I gotta spent some money first to improve the quality of my work first...

Cause I really only run on a inexpensive but at least good multimeter right now...
First of all I got to go get gfs, some rest, some Aegean isles wind and some sun,
some gym practice first of all, some drinks, perhaps a motorbike... and of course a job... part time....
Hell all that in a summer... puff... I need rest... someone spear me...

I know you don't really mind about all these stuff... but I gotta tell someone anyway...

So... Cheers...
Lefteris...

BTW... I need rest....

[superchiku]

trickynekro i also thought earlier of working on an autonomous helicopter but there are problems , u see some of my seniors made this project which was rc not autonomous and they had much difficulty controlling it , if U want i am also interested in such a project may be we can share our ideas and all but my suggestion it make it semi automatic coz if anything goes wrong on the top while ur testing it then say bye bye to everything.. by the way if u are using the accelerometer then it i sufficient to give u tilt , acceleration and angle wrt gravity then why are u using a gyrometer?

[benji]

just a question about this helicopter thing,, are you guys tottaly aware of FLYING?
i mean the aeromechanics,, how much should you push air?,,,,aero stuff in general,,any studies about that?

[TrickyNekro]

just a question about this helicopter thing,, are you guys tottaly aware of FLYING?
i mean the aeromechanics,, how much should you push air?,,,,aero stuff in general,,any studies about that?

Me nope... But I got some RC heli flying hours...

The first step is to learn how to counter counter-rotation with a gyroscope...
Here I need some libraries for Eagle... (any help here?)
I'll be using the ADIS16100ACC... And I don't have a board for it... I must build it...

Then I do studying how to hold position by a dual axis accelerometer.

I also know some physics on the helis... how torque is produced, and generally how stuff works...
And I'll be doing also more and more research...

If anyone can offer more help he is more than welcome...

Lefteris

[superchiku]

i cant offer u help but iam also thinking of making such a thing if u want we can share some theories do some research together and make the thing what u say

[benji]

well good luck with that ,,expect too many crashes,, how big the helli is gonna be?how much weight?

[Private Reid]

Wait a second?

I havn't posted here in while, but that doen't mean i havn't been reading


How is making a helicopter KEEPING IT SIMPLE?

[superchiku]

makin a helicopter isnt simple at all and making a autonomous one is ultrakill , there are lots of risks and difficulty involved also the cost factor is high that is why i am not making it mainly because of the cost but if tricknekro wants to do it then why not help him may be we can make it together

[TrickyNekro]

well good luck with that ,,expect too many crashes,, how big the helli is gonna be?how much weight?

Small and agile able to hold some batteries and the boards... The batteries usually come pre-installed and the boards
are light so no worries here...

To make this clear... total autopilot is not an overkill but it's very difficult and I would need much budget for sensors...
I was talking about a co-pilot autopilot... or at least some systems that would help you from crushing...
And maybe I will also involve some simple sensors for object detection...

And the remote and the receiver will be made by me and only... I guess a lot of people would like that!!!

Anyways... I have nothing more to say....

Cheers,
Lefteris...

[superchiku]

hmm are u using an already made chassis of a toy helicopter then  think it wone be very difficult but if u want full automation then iam interested in it ..

[TrickyNekro]

hmm are u using an already made chassis of a toy helicopter then  think it wone be very difficult but if u want full automation then iam interested in it ..

Not a toy... a full RC cause in toys you can't find pitch control and such...

And every single electronic part will be made by me...
And also not full automatic... you see still I need a GPS and such things... making it quite difficult...
cause of limited budget...


I'll see man... I'll see and do alike...
You see I mostly need help with parts, especially finding them...


Lefteris

[superchiku]

yes gps will be very costly also u need to learn how to use them, not good if u have limited time but if not fully automatic then i think ull be able to achieve it.... also i want to lern abt the electronics ur going to use if u need any kind of research just tell me i can get them for u

[benji]

is it totally autonomous or remotly controlled?

[TrickyNekro]

semi autonomous...
Some control will exist in first states... it's impossible not to...
Then it comes automatic take off and most importantly landing...

And final stage will be way pointing with GPS...

And all these will for the SoR contest...


To ADMIN:
I know member pages are not for blog like situations BUT... That project can only be completed in stages...
So as far as I have a stage done, I will provide dates, how to, building info, coding info, physics info general info, how I got the parts and so on...
I may also provide my real name here cause it's obvious this project is nice to be heard of... (no contact info)

But I insist on stage like project appearance, cause the whole thing is very complicated...
I wait for confirmation....


Lefteris

[superchiku]

gps waypointing will be cool may be u can make a tutorial abt it after u complete it

[airman00]

To ADMIN:
I know member pages are not for blog like situations BUT... That project can only be completed in stages...
So as far as I have a stage done, I will provide dates, how to, building info, coding info, physics info general info, how I got the parts and so on...

Why?

It should be written like any tutorial , take the 50 dollar robot tutorial as example. Remember the tutorial should have a focus on the program aspect of it , and only a bit of background info to understand the concept , for the robot itself

[TrickyNekro]

To ADMIN:
I know member pages are not for blog like situations BUT... That project can only be completed in stages...
So as far as I have a stage done, I will provide dates, how to, building info, coding info, physics info general info, how I got the parts and so on...

Why?

It should be written like any tutorial , take the 50 dollar robot tutorial as example. Remember the tutorial should have a focus on the program aspect of it , and only a bit of background info to understand the concept , for the robot itself

You really think this can be done with such a project... You know how many things will be missing that way...
And how didn't liked the big spider bot video series (BTW I haven't seen any movement so far... Am I missing something???)

I won't be chatting I know what to do...
But this can't come in a package... It's impossible both for the writer and the reader...

The further the steps go the further will the difficulty be...
Mounting electronics as first steps....
battery management...
microcontroller communication...
gyroscope interface...
accelerometer interfacing...
countering the heli's natural torque...
controls interface...
tests... ( that's unavoidably gonna be blog like)
coding...
advanced coding....
way pointing...
battery consumption and travel distance...
advanced way pointing...
RC remote building...
take off...
landing...
proper place for landing...
tree detection...
videos...
videos...
videos...


Lots of work to do...

Lefteris

[airman00]

ohh I see

You are absolutely right!

but for the other tutorials which deal with less complex robotics , they will not need that much.

Can't wait to see the heli

are you gonna make a quadracopter or just a regular heli?

[superchiku]

making an helicopter isnt complex its just tough coz u have to do research in so many things , if trickynekro makes a semi autonomous one, ill be making a fully autonomous one coz i already have the idea and the required resources just needed some motivation

[benji]

making an helicopter isnt complex

i think landing would be

[TrickyNekro]

ohh I see

You are absolutely right!

but for the other tutorials which deal with less complex robotics , they will not need that much.

Can't wait to see the heli

are you gonna make a quadracopter or just a regular heli?

Actually I'm will to buy a coaxial dual rotor helicopter RC like the Russian KA-50 (NATO reporting Hokum) and mode it...
Buy I know a guy who plays a lot with RCs and may be able to give me a regular heli in a extreme low price...
I'll see...

And Yes that's a challenge a big one I know... but just imagine it crossing the skies...

[TrickyNekro]

making an helicopter isnt complex

i think landing would be

in fact landing may be more difficult than way pointing....
The helicopter not only must find an flat area to land and decent but also
it must check the surroundings... you see the rotors have mass... meaning
that if object exists in the way it may collide with blades even causing a crash...
for tackling this problem I'm aiming to use a heli to ground sonar... some trigonometry
and some sophisticated code...
And the bad thing is that can cost RAM and processing power...
And for better results in decent process I would need a accuracy 3 axis accelerometer...
But I think I'm good with a two axis one...

So that's for now...

Lefteris

[benji]

how is it gonna see whats infront? to not let it bump into a wall or somthin

[TrickyNekro]

how is it gonna see whats infront? to not let it bump into a wall or somthin

Sonar or Sharp IR...
I have also read about optic flow but I don\t think it's practical outdoors....
And IR may have difficulties with the constant movement of the helicopter let alone the sun at outdoors...

But it's still an option for celling finding as the beam can pass through the spinning blades...
And generally quite many sensors.... if it's gonna be fully auto...

But first things first... tomorrow I start my exams finally... essay... pppoooffff....

Cheers,
Lefteris

[benji]

considering fast movement of the helli i dont think you have any time to do scans.. so the sensor should be stabely directed at one point,
the ir wont do it under the sun,, a sonar is better for this

[TrickyNekro]

considering fast movement of the helli i dont think you have any time to do scans.. so the sensor should be stabely directed at one point,
the ir wont do it under the sun,, a sonar is better for this

I agree with ya... but I was talking about landing... the heli should be fixed in a position and then choose
if the position is correct to land... if not go elsewhere...

And a sonar with a accelerometer with do the job...
Consider the the Doppler effect and knowing the speed of the heli (or just something close...)
You just have the distance of the object ahead...

you see it's really all about physics and maths... so... don't really worry...
Budget is the big hit...

Lefteris

[Admin]

Ok I finally read this really really really really long post

The $50 Robot is a great starting point for tutorials, as its a common open source platform many of you already have. And you wouldn't need to write a tutorial for everything, just on your special module you made for your own uses.

Yea benji, you can do more than one entry to the contest. The ad revenue I make from your tutorials pays for the prize money, so more entries means I hold the contest more often. But its always better to have one really good tutorial instead of two so-so quality articles.

But I insist on stage like project appearance, cause the whole thing is very complicated...

yeap Tricky, as you start your article you can add as you go. But the final version should be clear for someone to copy your robot design.

oh and tricky, i got myself one of those RC helis . . . crazy hard to control . . . but then a gear broke like 15 minutes after i started using it =(  if you make something that mixes the RC signal with stability control, I think that'd be a good first step!