| Subject: |
|
Re: starting a project |
| Name: |
|
Rob |
| Date Posted: |
|
Aug 24, 04 - 9:43 PM |
| Message: |
|
Hi,
Where to start and what to do first? Do LOTS of research - parts catalogues,datasheets, websites, magazines, books etc. Read and UNDERSTAND other people’s projects and what problems they had and how they solved them. Carry a pen and pad around with you at all time and religiously scribble any ideas/notes/bits of circuits that come into your head no matter how ridiculous they seem at the time – don’t rely on “Oh I’ll remember that and write it down later”, plus it’ll make writing up your project really easy it you date them – I’ve got to say the note book was one of the best things I did - it ended up bigger that the actual assignemt.
I like to break my projects down into blocks – draw the blocks (literally - as labelled blocks e.g. microprocessor, motor controller, sensor module etc) and show (with arrows) and how they connect together on a piece of paper. You will notice, if you’ve read the report for the project, that the robot was built on several individual circuit boards. A big project is less daunting if broken down into small achievable parts, and should the worst happen and the project fail, at least you will have something to show and be able to say well 4 out of the 5 modules worked as planned.
I suppose it also depends on you experience; one of my problems was that I hate analogue electronics, so designing the sensor module was difficult because I didn’t want to have to build an analogue amplifier circuit for the sensors but doing research I found some sensors that gave logic level outputs and so negated the need for me to get involved analogue circuitry. Remember - there is always a way!
I always find when starting any project I like to do the most difficult bit first because I always have the most enthusiasm at the beginning of any project.
If you do the bit(s) you think you’re likely have problems with first:
a) When you succeed with that part, everything else will be downhill from that point (very important if you’re doing it to a dead-line).
b) If you get stuck you can always do a part you know you can do and then come back to the difficult bit (just working on something else for an hour/day/week can yield the solution immediately when you come back to it).
b) If all else fails, it allows you to find a different solution (or additional help) for your problem.
Things to consider for a robot project (anyone - please feel free to add any I may have missed out):
What is going to control it? – Will it be pc controlled or microprocessor controlled? – are your interfacing/programming skills up to the task?
How will it detect the line?
Where will I get the motors/gearboxes from?(try eBay!) - How do I control them?
How will it be powered?
How do I make the chassis?
Personally I would start with the sensors (I would try Omron EE-SY410, Optek OPB760T or OPB704 sensors if available), get some (more that you require, just in case of accidents and there will be some) they’ll be cheap and have a play with them with some insulation tape (as a track) on a bit of paper – measure the output voltage as the sensors sees the white paper then the dark tape – you’ll get an idea how far they need to be from the track and how sensitive they are; then try to interface them to a control system (pc/uP) – try to get their status (light/dark) displayed on some leds or on screen for a start, then so on and so on – you get the idea. Don’t try to build the entire thing on one board on your first go, it probably won’t work and you’ll rapidly loose interest also don’t forget to look at the Resources and Bibliography page in my report for additional sources of information.
Estimate how long you think the project will take and double it! |
|
|
|
|
|
|
Reply
Home
bravenet.com