Boarding the microcontroller-train

During the past two years I've been thinking of - but never got the grip
to - starting to learn how to use microcontrollers. For my next hobby
project, it seems to be a necessity to learn so I thought now - once and
for all - is The Time To Learn.

The next project is going to be a something that looks and acts like a
chess clock, with some few additions. While not completely impossible to
do with a bunch of 555-timers and a lot of logic gates feeding some
7-digit LEDs, the system would probably be a messy harness real quick and
blocks the ability for improvements. So I'm thinking a uC will take care
of all the timing / scoring, and use some external chips (like a 7-LED
decoder and the like) as helper chips.

Now, so far so good in the planning, here's the real reason that's stopped
me for learning microcontrollers: I have absolutely no idea where to
start. Some people say "Go Atmel!", others say "Go PIC!", others say "Go
Foo!".

I have googled around for tutorials, beginners' pages, introduction to /
comparison between Atmel/PIC and simliar pages, I've seen plenty of
"starter kits" every here and there, some way more expensive than others
making me confused of which to buy, so I still don't know where to put my
foot.

I have 20+ (software) programming experience. Around 1990-95 I programmed
system assembler for the MC68000, meaning I have atleast some of the
required skills. I think.

So, considering the next project of mine and my background, which kind of
uC might be good enough for me? Or atleast for starting out in the area.

 

Add a "years of" after "20+".

 

Elektor has a 5 part course, "Elektor Microcontroller Basics Course -
Complete.pdf", that floats around on the web.



--
..
..
..
..
..
..
..
..

 

If you've already written assembler for 68000, you're already well on
your way. For little projects like you're describing I like to use
assembler wher you have total control and responsibility for what
happens. I haven't used PICs but quite a bit of 8051 and 6805/65908. I
find switching gears between writing for one vs the other is pretty
trivial - even to the point of transplanting MCS-51 code directly into
6805. Rename Intel MOV to Motorola LDA/STA as needed and fix the
assembler mnemonics and named Motorola RAM location to the names of
Intel registers. Worked fine.

As for the 'helper' 7 segment decoders, I used look-up tables to do
the decoding in software which also allowed text using '7-segment
font' so the clock can be set for month/day in text along with
reporting time 'fail' and restore times.

http://www.twyman.org.uk/Fonts/

You'll have fun with whatever processor you choose and wonder why you
waited so long.

GG

 

don't know if you have seen this or not:
http://www.dontronics-shop.com/pages.php?pageid=23
it may help

Don...


--
Don McKenzie
E-Mail Contact Page: http://www.dontronics.com/e-mail.html

Crystal clear, super bright OLED LCD (128x128) for your microcontroller.
Simple serial RX/TX interface. Many memory sizes.
http://www.dontronics-shop.com/product.php?productid=16460

No More Damn Spam: http://www.wizard-of-oz.com

 

If you're a software guy, then you will appreciate how much better the
Atmel AVR instruction set is than the classic PIC. Others before you
have appreciated it in the form of a working GCC target for AVR. So if
you want to sit down and code in C with no cash outlay, go AVR. If
Atmel is still selling the Butterfly board for $20 from Digikey, get one.

 

You could download MPLAB from MicroChip and write some RISC code for
their "PIC" type (RISC) processors and run it (simulate it) on your
PC.

Then you could buy the actual hardware i.e. a PIC and some other
components (for a few pounds) and solder them together and transfer
you code into the PIC with the "PICSTART" programmer (maybe a £100 ?)
and try it out.

With your experience, you will *enjoy* writing assembler using a
minimal, instruction set, free from redundancy and you will have total
control.

If you were to go for a CISC then you should use C and maybe a third
party operating system but then you *will* get bogged down in a morass
of systemic-setting-up with no extra gain for a small project.

Robin

 

You could download MPLAB from MicroChip and write some RISC code for
their "PIC" type (RISC) processors and run it (simulate it) on your
PC.

Then you could buy the actual hardware i.e. a PIC and some other
components (for a few pounds) and solder them together and transfer
you code into the PIC with the "PICSTART" programmer (maybe a £100 ?)
and try it out.

With your experience, you will *enjoy* writing assembler using a
minimal, instruction set, free from redundancy and you will have total
control.

If you were to go for a CISC then you should use C and maybe a third
party operating system but then you *will* get bogged down in a morass
of systemic-setting-up with no extra gain for a small project.

Robin

========================================================================

I would suggest the PICkit1 for $36 or less, which allows you to program
smaller flash devices of 8 and 14 pins. I used mine to develop a project
recently. I also have an ICD2 debugger tool, which cost about $150. It can
program almost any PIC, and can run the target in realtime with a
breakpoint. Both devices connect with USB. I even accidentally damaged my
ICD2 with a 120 VAC spike, and I got a free replacement in 2 days from
Microchip.

There are many players in the market, but it is important to go with a
company that can provide good support, and has a large base of users who
can give advice. Microchip has a very good on-line service ticket which
gives you quick response to problems, a good user forum, and a liberal
samples system, all of which are free. Of course, they hope you will be
ordering 1000 pieces soon!

Paul

 

You've everything going for you so no clear choice for the micro.
I'd suggest the PICs. Primarily because they are cheap, available
everywhere, all the devices are similar inside and every man and his dog
offers hardware and software for them.
Downside is that the PIC hardware internals are pukey and the microchip
manuals and support stuff have been written by lunatics from some
programmer's care home. They're easy to use once you've got to the flashing
LED bit, though for a complete beginner, the AVR would get you to that point
quicker.

 

Actually I think the datasheets are one of their strong suits. If you
want to drive some PIC's I2C hardware, you could learn how to do it from
their datasheets having not previously known what I2C was at all.
They're much better than the ones for the early AVR parts.

 

It's the size of the things that scares me. God knows what they do to a new
starter.
Just starting to use their 128K PIC18F? thing, 8Mbyte PDF, 500 pages, no way
am I printing it out. Spent 10 minutes leafing through looking for
monumental improvements in line with their marketing blurb. Nah, realised I
was just looking at cut and pasted 16F628s or 16F877s etc.
Seems we're now upto 80 instructions (very nice!). Think it's high time they
lost the "R"isc marketing aspect
john

 

I just find it a little hard to believe that there are people who actually
seem to _like_ bank-switching.


Thanks,
Rich

 

Thank you for the input.

There is probably no risk for painting myself into a corner - I could
always change the architecture if I see that there's something queaky. But
as of the time of writing, I have zero experience with any kind of
microcontroller, and I feel that I must atleast start *somewhere*.

I've decided to start with PICs. Ebay had some cheap USB-programmer from
China (probably pirated, but who cares) so I ordered it yesterday. In the
meantime until I get it, I'll eat some more documentation.

 

Check out the Microchip site. You may be able to
get free samples to experiment with.
http://www.microchip.com/stellent/idcplg?IdcService=SS_GET_PAGE&nodeId=64

Ed