[phoenix-project] Re: libcomedi

Hello,

Pramode C.E. a écrit :
> I am forming the outline of some user-level 
> documentation - any suggestions regarding use of
> specific tools for writing the docs (plain old LaTeX,
> plain HTML, LyX,
> Scribus, OpenOffice, DocBook ...). 

Hmmm.... the more structured is the tool, the more flexible
and portable your document may be. There is always a tradeof between
high structuration and ease of access to contributors.

OpenOffice may be a good bet, provided you really use the structuration
tools included with it. Don't hesitate to create the necessary markup
tools. For example, if you include code lines, do use the Stylist
(Format->stylist...), use the tool icon with a popup help "New Style
from Selection" and create a style "code" from the style "default". 

Then you may later parse the XML produced by OOo to output well-formatted
documents in every case. As soon as you defined the new style "code", it
appears in the Stylist, and you can by right-clicking it, define specific
attributes, like a particular font, an color, etc. (for example Courier
pitch 10, brown). So when you use your markup, there is a wysiwyg
behavior.

The style can be exported and used in more than one document.

> identifying experiments which can be 
> done *without* using expensive and often
> hard-to-obtain
> sensors - the only kind of sensors which can be
> purchased without too much difficulty here in India
> are (a) basic heat sensors (b) light - LDR,
> phototransistor, stuff like that) - any suggestions
> in this regard? 

some inexpensive sensors I already used :

- a screen with regularly distributed rectangular holes (or the same
  printed with opaque ink on a transparent medium) ==> speed sensor when
  passing in an optical detector.
- a black/white linear gradient (make it with sketc or any other SVG
  editor), printed on a transparent sheet ==> one-dimensional position
  detector when put between a light source and a light sensor area
  greater than the pitch of the printed gradient. I use the printed
  gradient clipped in a slide frame, the frame has holes to bind it to
  other devices.
- coils around a tube ==> speed sensors and detectors for moving magnetic
  small dipoles, for example magnetic stirrers used in chemistry labs.
  For example is the tube is vertical, and the stirrer falls inside it
  (with air or any liquid), each coil around the tube emits a signal
  (the zero crossing may be a date for the event), whose peak-to-peak
  voltage is representative of the speed.
  
> Would it be possible to identify 
> experiments in the fields of Earth/environment
> sciences (Phoenix controlled weather station - a 
> simple wind speed measuring instrument which generates
> rotational movement the speed 

- a rotating magnet near a coil yelds a signal whose peak-to-peak
  voltage accounts for a speed. Of course optical detection is
  straightforward if the rotating body is fitted with a little mirror.
  
> of which can be measured
> using some photo sensor + digital input pins; some
> setup to measure humidity - things of that sort),

- any resistive device, used with a capacitor and a HFE4093 chip gives a
  reliable oscillator, whose frequency can be related to the resistor
  value. Take in account the necessity to regulate the supply voltage
  (LM78L05) since it also modifies the frequency.
- humidity sensor : you may try digitated electrodes made of conductive
  ink on a paper saturated with a Cobalt chloride salt. If the
  conductive ink is not that easy to use, you may use other electrodes,
  but they must be of a metal more noble than cobalt (for example
  silver.. you may try the alloys of some coins also)
  
> Chemistry, Biology, Industrial control and
> even basic electronics (software `gates' using 
> digital input+output pins, `analog computers' using
> OP-AMPS ...)

wide, very wide fields. If you can describe more precisely one
measurement to be done I may try some trick. Cheap sensors are seldom
linear and long-lived, but they are handsome to teach useful concepts.
However their conception is often challenging.

Best regards,                   Georges.

Other related posts: