[HH] Seeking help on Programming I2C bus in python on Raspberry Pi

[Jon Evans]

Hi Ming,

The << operator is a *logical shift left*.  See the wiki article here:
http://en.wikipedia.org/wiki/Logical_shift

According to the MCP23017 datasheet, the I2C registers 0x12 and 0x13
control the outputs.  There are 8 bits per register, so each bit controls a
single output: if the bit is high (1), the output is also high, and the LED
is on.

What the example code does is loop through the range 0 to 7, and toggle one
bit on at a time (to turn each output on in sequence).

If you take a look at the wiki article for logical shifting, can you work
out what happens when (for example), you set I2C address 0x12 to (1 << 5)?

Best,
Jon



On Thu, Nov 15, 2012 at 3:38 PM, Ming kuo <mingtkuo at gmail.com> wrote:

> Thank you Tom, Jon, and Greg,
>
> I am grateful for all the help while I go through this challenge.
>
> Tom, I have previously responded to Jon and Greg privately trying to avoid
> too many emails to the group.
>
> But yes, your tracking of my progress of the project is right, I basically
> have proof of concept with hardware, circuits and now it is just the
> software programming..
>
> Following the code in the previously mentioned example, I am trying to
> adapt it to working with the Bar LED..
>
> I have isolated the critical piece of code is
>
>
>
>    for x in range(0,8):
>      a = 1 << x
>      set_led(a,0)
>      time.sleep(delay)
>    set_led(0,0)
>
>
> in code :  set_led(x,y) x is the data and y is the bank.
>
> so I am confused on x or data piece of the parameter..
>
>
>
>
> in the above code set_led(a,0)
>
> *a= 1 << x what does this statement do?*
>
>
> *and I tried just setting a to various numbers 1,2,3,4,5,6,7,8,9 and the led seem to light up in no understandable pattern, sometimes individually and sometimes multiple LEDs go on.. *
>
>
>
> So basically, once I get to know how to control the leds on the bar graph, then I can start deciphering the code for letters and words
>
>
> Thanks to the group
>
>
>
> Ming
>
>
>
> On Thu, Nov 15, 2012 at 2:36 PM, Tom Metro <tmetro+hhacking at gmail.com>wrote:
>
>> Ming kuo wrote:
>> > Hi Guys, I was recently introduced to this group, so here is my first
>> post.
>>
>> Welcome to the group.
>>
>>
>> > I figured out the wiring to get the bar to light up using the same codes
>> > as above.
>>
>> Sounds like you've got the hard part done. If you've gotten that much to
>> work, then you've proven the hardware works, and the basic software
>> works. Now you just need to tweak the behavior.
>>
>>
>> > I am stuck...on the python code to control the I2C.
>> > This is my immediate challenge to figure out how to program the
>> > I2C for the LED bar...
>>
>> Can you clarify?
>>
>> You said above that you succeeded in getting the bar to light up using
>> the same setup as the article you referenced. That implies you have a
>> working I2C driver and some Python code using it.
>>
>>
>> > ...I need to program a python code to POV display the messages which
>> > i can imaging that it get pretty challenging given how to form a
>> > character with LED while timing the rotation.
>>
>> So POV is persistence of vision? You're looking to build something like:
>> http://www.youtube.com/watch?v=fJBYOU-Y-ZU
>>
>> How do you plan to move the display?
>>
>> To build on what Jon Evans said in his reply, it sounds like what you
>> have left to do is less hardware hacking and more basic Python coding.
>> Looking at the sample code:
>> http://www.skpang.co.uk/blog/wp-content/uploads/2012/06/ledchaser.py
>>
>> they provide a set_led() function, so you'll need to remove the two
>> chunks of code used to strobe the LEDs ("# Move led left " and "# Move
>> led right") and replace them with your character generating code.
>>
>> Not trivial but given that POV displays seem to be popular, likely you
>> can find some existing code for the character generator. If not, you'll
>> need to dig up a library or lookup table that maps characters to dot
>> patterns. Basically a dot matrix font. Then build a function that when
>> given a character, translates that into the equivalent array of dots.
>>
>> I'd be inclined to use an architecture where rows of dots are pushed
>> onto a stack, and then some background process is responsible for
>> pulling a row from the stack and lighting up the LEDs on a steady clock
>> rate. (I'm not sure if that's easy to accomplish in Python.)
>>
>> I'm guessing that with most of these POV displays, there isn't any real
>> synchronization between the display motion and the strobing. The builder
>> simply tunes the strobe rate until it works most of the time for
>> whatever RPM their display ends up running at.
>>
>>
>> > So any help through email or in person would be greatly appreciated.
>>
>> Post some follow up questions as you get stuck and we'll see what we can
>> do.
>>
>>  -Tom
>>
>
>
> _______________________________________________
> Hardwarehacking mailing list
> Hardwarehacking at blu.org
> http://lists.blu.org/mailman/listinfo/hardwarehacking
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.blu.org/pipermail/hardwarehacking/attachments/20121115/fc7381df/attachment.html>