Re: [Discuss-gnuradio] Parallel Port Interface

[Chris Albertson]

You will never get good, or even decent sampling stability from
Linux unless you go with one of the two real-time varients.

Another idea would be to generate the sampling rate in hardware
and use a large FIFO buffer.  But see below it _can_ be done
in software.  Your sample rate is near the limit for a modern
ECP parallel port.

I've used "real-time Linux" (RTL)to write a camera driver where the
pixels must be sampled using very exact timming continuely for
hours.  (It was a "drift scan camera" ask for details if
interrested.)

RTL is a modified Linux kernel with features added for "hard" 
real-time applications, where mincroseconds count.  Users of this
system write theirapplications as loadable kernal moduals.
It's not hard except that a bug in your modual can crash the
system.  In fact "A: goto A" will lock it up and require a
hardware reset.  

It would work well for your task.  I'd put the sampling code in
a loadable modual so that it would run at true real-time
prioroity (The modual runs __in front__ of Linux itself.  The
Linux OS runs in the backgrouund of the real-time modual.)
You can write normal user level code to take the real-time
data and do "whatever" with it.  

I did this work years ago with a 1.x version kernel.  Lately
I notice RT Linux has "forked" and there are two of them.  So
now you have a chose.  

You can see what did.
http://129.21.188.9/tass/software/tmget3/index.html
Here is the hardware it ran on
http://129.21.188.9/tass/hardware/beser.html

you can also go to Google and find links to both RT Linux
projects.  They are semi-comercialized it that while free you
can buy support.  


--- Joseph DiVerdi <address@hidden> wrote:
> As I wrote recently, I am embarking on a project to apply SDR
> techniques to ionospheric examination via VLF monitoring and still
> haven't settled on an interface. Today, I'm thinking about the
> parallel port.
> 
> I'm planning to perform the experiments myself very shortly but
> expect that the results are already known somewhere so I'll ask the
> question:
> 
> What are the limits on sampling rates when data is pumped in to a PC
> (running Linux) through its parallel port? 
> 
> Assuming no subsequent processing, can the port be read (8-bit data
> only) 400-500k times per second? Continuously? What is the load on
> the various subsystems (CPU, etc)?
> 
> As the sampling times are generated externally (to the PC) via
> hardware, is interrupt driven data sampling appropriate? Advisable?
> Effect on maximum sampling rate? Effect on subsystem loads?
> 
> Any data, experiences, comments would be greatly appreciated.
> 
> Best regards,
> Joseph
> -- 
> Joseph A. DiVerdi, Ph.D., M.B.A.          
> http://diverditech.com/           970.980.5868 (voice) 
> http://xtrsystems.com/            970.224.3723 (fax)
> PGP Key ID: 0xD50A9E33
> 
> 
> _______________________________________________
> Discuss-gnuradio mailing list
> address@hidden
> http://mail.gnu.org/mailman/listinfo/discuss-gnuradio


=====
Chris Albertson
  Home:   310-376-1029  address@hidden
  Cell:   310-990-7550
  Office: 310-336-5189  address@hidden
  KG6OMK

__________________________________________________
Do you Yahoo!?
Yahoo! Mail Plus - Powerful. Affordable. Sign up now.
http://mailplus.yahoo.com