Main Menu

Build log: Model 2

Started by naatumach, June 23, 2011, 10:34:20 AM

Previous topic - Next topic

0 Members and 1 Guest are viewing this topic.

Swapnil

Yup, seriously!
With so many tutorials and forums, it's really tough to find exactly what you need. :banghead: We searched for like 3 months before we got anything promising. There's more stuff in AVR than in PIC concerning robotics.
You are currently in 5thsem, right? You must not have had a subject called 'Control systems' yet. It makes you realize how easily you can apply 'PID' and 'Fuzzy' in robotics.
I'm assuming you've tried the bang-bang algorithm (If you haven't, try that first).
Now, the industrial standard PID control mathematical formula (containing integrals and derivatives) can be modified into a simple one (additions and subtractions) for a LF Bot. Then all you need to do is tune the 3 constants till you get the correct set. And this can be easily done in AVR as well as PIC.

Oops! Looks like your thread's drifting away from RC! I guess we could carry this on in chatter zone. :)   

naatumach

Quote from: Swapnil on June 26, 2011, 07:58:26 AM
Yup, seriously!
With so many tutorials and forums, it's really tough to find exactly what you need. :banghead: We searched for like 3 months before we got anything promising. There's more stuff in AVR than in PIC concerning robotics.
You are currently in 5thsem, right? You must not have had a subject called 'Control systems' yet. It makes you realize how easily you can apply 'PID' and 'Fuzzy' in robotics.
I'm assuming you've tried the bang-bang algorithm (If you haven't, try that first).
Now, the industrial standard PID control mathematical formula (containing integrals and derivatives) can be modified into a simple one (additions and subtractions) for a LF Bot. Then all you need to do is tune the 3 constants till you get the correct set. And this can be easily done in AVR as well as PIC.

Oops! Looks like your thread's drifting away from RC! I guess we could carry this on in chatter zone. :)   

I searched my whole 3rd sem i found lots of things but never a code on which i could reverse engineer and learn atleast.
I am now im 5th sem and im in mechanical there's no control system for us.
I have tried bang bang, i have tried to use a line sensor array and use it to give different speeds(PWM) for different values but no pid.
The biggest problem is i think that i know what needs no be done, but how exactly it is done eludes me.
Motto: build, fly, and modify.
Prefers working on designs than electronics.
No pre-made plans. Use blueprints, make calculations, build, fly, troubleshoot, fly again.

naatumach

Motto: build, fly, and modify.
Prefers working on designs than electronics.
No pre-made plans. Use blueprints, make calculations, build, fly, troubleshoot, fly again.

Swapnil

Quote from: naatumach on June 26, 2011, 10:35:22 AM
...The biggest problem is i think that i know what needs to be done, but how exactly it is done eludes me.

Haha! Thats the problem with everyone man!

O.K, before I give anymore details here, I'd like to know how good you are with electronics.
Like, 'can you manufacture your own PCBs'... 'have you worked with SMD'...'have you memorised your AVR uC datasheet  :)'...etc

This is important because... even a hobbyist can easily make a line-follower, but to make a really good one (i.e with PID) you need to know your stuff inside out. And it's good that you are in mech, 'cos you can properly fine-tune the required mechanical aspects like low inertia, low CG, proper traction, suspension (possibly!).

Which AVR are you using BTW?

Swapnil

Quote from: naatumach on June 23, 2011, 10:34:20 AM
... since i do have 10props i do not fear crashing a lot, so vineet you can follow this.
Which props are you using? And with which motor?

Swapnil

What's the progress with your plane? I'm gonna be sticking to deltas as they are simple and straightforward with less complex control surfaces.
Gonna try the 'Simple Delta' next.
http://www.rcgroups.com/forums/showthread.php?t=867083

And later the 'The Mugi Evo' perhaps.
http://www.mugi.co.uk/evo_intro.php

naatumach

Nice designs. Looks nice, worth a try.
What i mind is that the wind plays a big spoil sport in this design. But its a wonderful design.
.
I use 8060 props, with the same motor from http://www.rcindia.org/self-designed-diy-and-college-projects/first-flight/msg68033/#msg68033

But i am going to switch to either emax CF 2508 or aeolian C2822 and buy X-Motor mounts
Motto: build, fly, and modify.
Prefers working on designs than electronics.
No pre-made plans. Use blueprints, make calculations, build, fly, troubleshoot, fly again.

naatumach

Regarding that avr
No i know none, i can only code thats all. I never got into it that much.
Motto: build, fly, and modify.
Prefers working on designs than electronics.
No pre-made plans. Use blueprints, make calculations, build, fly, troubleshoot, fly again.

Swapnil

Hmm...without good knowledge of uC peripherals, it'll be hard. Try getting acquainted with a good one. But then, starting with micro-controllers can be as difficult and time consuming as with RC. Good Luck!

naatumach

Then i think i'll pass. I am out of time for getting into uC because i only do rc to let off steam, i need to focus on college and gre now, baaki sab side mein.
Motto: build, fly, and modify.
Prefers working on designs than electronics.
No pre-made plans. Use blueprints, make calculations, build, fly, troubleshoot, fly again.

naatumach

After 'model-2' nose dived and broke. I did a nosejob on it, and left it. Then i went to b'lore. Now im back, i made the YF-22. But its too windy outside to fly, and i cannot sit idle so i took this nosejobbed-model2 for a sortie and the wind made it nosedive and the makeshift mount broke.

So i came back and made this modification.

The causes for trying out this design-
1. First model with the prop not touching the ground.
2. I never had much faith in my landing skills and structural integrity of landing gears. So i'll fashion some landing gears tomorrow, and learn to fly and land properly.
3. I can also use the ground for takeoff instead of a hand launch.
4. I tried to drain this battery and it lasted for 25minutes on full throttle on a ground run, i didnt assume it would run for so long. So i decided to make this and test my endurance.

And i love this type of motor mount.
Motto: build, fly, and modify.
Prefers working on designs than electronics.
No pre-made plans. Use blueprints, make calculations, build, fly, troubleshoot, fly again.

Swapnil

Wow, that bird looks weirdly impressive!
I'm working on a similar design.
Waiting for your flight report...

Swapnil

Is that mount from RCD? It's currently out of stock there. Where else can I get one from?

naatumach

Motto: build, fly, and modify.
Prefers working on designs than electronics.
No pre-made plans. Use blueprints, make calculations, build, fly, troubleshoot, fly again.

Swapnil

Oh, sorry, I meant to write RCB and not 'RCD'...typo...
Anyways, the 60 Rs. ones are out of stock there.
The one you have is for aeoilan motors, right? Will it fit a CF 2822?

naatumach

Yes it can.
Attached are the photos.
You can also use at as a prop balancer
Motto: build, fly, and modify.
Prefers working on designs than electronics.
No pre-made plans. Use blueprints, make calculations, build, fly, troubleshoot, fly again.

Swapnil

It's a bit pricey though...
Is it made of plastic or metal?
Also, what is the general shipping time of RCB?

naatumach

They are metal. Very strong and sturdy, worth the purchase.
I didnt use the shipping option, i picked it up myself. I wanted to see the store. Nice store.
Motto: build, fly, and modify.
Prefers working on designs than electronics.
No pre-made plans. Use blueprints, make calculations, build, fly, troubleshoot, fly again.

Swapnil

In the pic, there are holes for mounting a motor on both sides of the mount...or is it just me?

naatumach

There are holes for mounting the motor on both the sides. unfortunately they dont line up with this motor, they do line up with the aeolian.
Motto: build, fly, and modify.
Prefers working on designs than electronics.
No pre-made plans. Use blueprints, make calculations, build, fly, troubleshoot, fly again.

Swapnil

That's what I was asking when I said "Will it fit a CF 2822?".
I think I'll have to wait for them to restock those plastic ones...

naatumach

But this ziptie thing is just as good. the motor doesnt budge or vibrate atall.
Motto: build, fly, and modify.
Prefers working on designs than electronics.
No pre-made plans. Use blueprints, make calculations, build, fly, troubleshoot, fly again.

Swapnil

#47
Hmm...could give it a try.

ujjwaana

Quote from: naatumach on July 18, 2011, 03:32:32 PM
But this ziptie thing is just as good. the motor doesnt budge or vibrate atall.
Are you sure ? May be in other places, but definitely not as the only fastening points between the motor and the Mount!! These are plastic and not Kevlar. The ties would soon stretch just enough to give you enough viberations.

I would rather suggest you to use 'Stick' mount which are the best mount for you application, used by so many scratch builders:

http://www.rcdhamaka.com/index.php?main_page=product_info&cPath=56&products_id=410

The best part is that the bell of the motor, which is inside the inserted mount, would go right inside the hole in the plastic mount.
Futaba 8FG Super | HK-450v2 | FA-22 Raptor |AXN Floater-Jet | FunJet | Black Horse Edge 540 | Amp Master 015 | 2.3M Big Brother

foamybuilder

The zip ties are good. I have been using that in all my planes since two years, probably not on highrpm motors though.
Only issue with that is once its tied and glued if there is some  problem with motor its a pain to remove the ties and to refit it you need to remove the wooden mount which is already glued to the foam.