Phase (4)

I was kinda of hoping to write more details in phase (3) before moving on to (4), but let's wrap it here and move to the next phase...

Phase (4)

Integrating.......

This phase defiantly bring memories!

Anyways,

Naturally, the whole team was involved in this phase and we started to assemble the whole thing together. Adding the control part to the mechanical was not a piece of cake at ALL!

we had MAJOR problems...

• we didn't know how to connect the linear motor
• the printed circuit was not tested
• we only had one shot!
• the program was not tested

and the very best part :D, the motor driver chip exploded with the very first try (I mean we could actually see flame coming out!)

Which didn't exactly give us much of a choice but to re-implement the circuit on a breadboard, which we didn't do. we tried to test the rest of the circuit and add an external chip and connect it.

What we didn't know back then was that the printed circuit already had a problem with the ground/VCC connections and was solved by making a separated vcc/ground connection to a part of the circuit, which people forgot to mention earlier !!!

Oh, and that was just 2 or 3 days before delivering the project :)

problems,stress,deadline........These are the times u learn in the most, not only about the project, but about yourself and the surrounding people!

I am being drifted off point here

tell you what happened next in the upcoming post isA :)

stay tuned ;)

Continue Phase (3)....control

Back to the Control Team....

we basically used a microcontroller, the inputs were; sensors and RTC chip. The output was to the motor driver and from it to the two motors.

we searched online for the appropriate interface circuits with the addition to a switch and that was the output.

I wasn't deeply involved at this particular phase really,  but this is how I perceived it...  

Continue Phase (3)....mechanical design

The pictures as promised in the last post...

those were the first parts we made :)

 That one was at 3ezbet el na7'l, at the workshop where we assembled the parts for the first time :)

This is actually one of the happy moments; when you see what you draw and thought about became reality !!!












Some close up looks...

 

Rack-Pinion arrangement;

 transferring motion from the linear actuator to the tracker (linear to rotational) 

One of the major difficulties that we encountered during the assembly phase of the tracker was that the non-availability of the type of the rotational encoder we planned to use (disk), we had to use another type and the design wasn't made for this one! so we had to improvise. 

Trial one

Someone in a different team suggested that design for us. Unfortunately it didn't work as the encoder will not exactly rotate with the same angle as the bearing which is VERY not acceptable as the encoder job IS to measure the angle of the bearing movement :S 

Trial two

After several trial and errors, we came up with this design, which is far more accurate and it worked just fine thanks to Allah :)

Final look

That's the final design, after it was polished and the dummy load (glass sheet) was fixed to replace the cells.

Continue Phase 3....

The Mechanical Team

Implementing the tracker on real life was a bit different than on papers. It was our first time to actually do something for real!. The amount of knowledge we gained was GREAT but the task was tiring.

Difficulties:
_________

1. When you drew on inventor a working drawing that says a part should be 32 mm it's easy, but when you tell someone to cut it 32 mm that's another story :D. We had to adjust most of the measurements we had to make it easier to implement
2. The names you study for things in college whether it's the English term or even the Arabic one is so not what workmen use. For example, what we would normally call a ball bearing with it's seating; is commonly known as "balya wesh 2ora!!!!"
3. You could make the best design in the world, but when you go do it someone might tell you it's not doable, and they could be very wrong, but if you insisted you could have it your way which is good or the workman might simply tell you "Fine, it's doable but I don't know how to do it!". WHY?? cause they have other stuff to do (mass production) and can't careless about your once in a life time small piece. so you may have to readjust your design for no apparent reason :S (Though I'll have to say that a good group in communicating with workmen would easily get pass this obstacle isA :) )
4. Rest of obstacles are comfort wise :D (bad place, weired people, horrible weather, no place to park if you had a car, and no transportation if you don't , no bathrooms, barely mosques to pray......yet, I had fun :D:D )

I will spare you the details of the journey and wrap it up with pics  :)
Stay tuned for the next post isA ;)

Phase (3)

As mentioned before we were divided into two main teams: control - mechanical

let's start with the control:

first we choose PIC18F452 as our microcontroller because it was:

• Available
• Easy to program
• Low power
• High performance
• Large memory
• Re programmable
• And some technical specs 

we wanted to have two different control methods:

1. was the normal controlling using the angles equations
2. Test case: to try it on the presentation.....for that we added 4 sensors in order to adjust the solar panel in the required place.

we will talk on details about these two controlling methods later...

The idea 
was to get the best out of the tracking system....and the most accurate tracking was not the one done by equations, yet it was the most one used....why??

because if we tracked using the sensors we will have a major problem; what if the whether turned cloudy??!!

the scenario will be something like that:

it's cloudy, so the sensors reading will go crazy for a while or will read equally. after a while the sun will have moved from it's place to who knows where. when the clouds are off, the panel will be facing wrong direction, and either will be able to readjust itself if the the sunlight was still seen by any of the sensors. or the sunlight will already be gone leaving the panel in its place till next day....and that's a major waste of possible energy which make it a major problem............however under normal whether conditions.....sensors are better trackers than angles......that's because it is a real time sun tracker, instead of pre-calculated angles which might be not accurate from a day to another.

our aim was two develop both tracking methods and then as a future adjustment to the traker; AKA if we had time :D, we would add a cloud sensor and use it's output as a condition to shift between the two tracking systems.

however we didn't implement this part for two main reasons:

1. lack of time of course
2. lack of the required sensor (we needed to have it deliver from another country) and that wasn't available because as I said........no time, plus not all the team was excited about the idea to be honest...most of them just wanted the whole thing to get done we 7'las :D

Well, after this quick outline, lots of brainstorming and online searching, they designed the circuit, picked the components required and two of the team went shopping :)

the remaining team member (as the control team consisted of 3 members) started to develop the code.

to be continued isA.....

Our Final Design :D

This is the final design I mentioned in the last post

Advantages:

• Simple
• cheap
• no calibration required
• no gear backlashes
• motors used have self locking mechanism

Phase (2)

Conceptual Designs....

picking up where we left at the end of phase (1), we started sketching our own designs. some where some how illogical and got rejected on the spot. others however were seriously taken into consideration; those were the one we picked up for the next level where we implemented 3Ds conceptual designs of them on Autodesk inventor

The first design was a gear-gear mechanism

main reason why we thought of something else was because of the gear backlashes which will lead to inaccuracy in the angles.

Second one was a gear-belt mechanism

we were trying to avoid the backlashes by using a toothed belt instead

third one was a belt-belt mechanism

it was less complicated and no gears used, but I don't' have an available pic for that one though

fourth one was a lame trial for something different

it was when we decided that these mechanisms were still too complicated, so we shifted our thinking completely and decided to make a swinging mechanism. we were already trying to avoid this cause we will then have to calibrate the angles and we didn't wanna do that :D , still trying to simplify the programming phase. yet, we gave it a shot.

fifth one was the swinging mechanism

in that one we enhanced the mechanism before it, more logical dimensions also :)

Sixth one

that was our final design that we actually implemented.... it was redesigned several times in the actual implementing but I will be posting it's picture and some details on the next post isA