[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Discuss-gnuradio] Parallel Port Interface
From: |
Joseph DiVerdi |
Subject: |
Re: [Discuss-gnuradio] Parallel Port Interface |
Date: |
Mon, 20 Jan 2003 14:40:58 -0700 |
>> 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 think you're giving up on the sound card too easily. IIRC you need 100 kHz
>of
>BW. There are soundcards which can provide that.
>
>> 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?
>
>If you are using an external sampling clock, you need the parallel port's
>sampling to be synchronous with that clock. Interrupt driven is the only way.
>Otherwise you'll sometimes have bits changing right when you sample, and some
>will have completed the change while others have not. However, I know of no
>machine which can handle 400-500k interrupts per second.
>
>The other option is to use a parallel port output bit as your clock. This
>would
>solve your synchronization issues, but would create others. RTlinux has an
>interrupt jitter around 10 us, IIRC, which is amazingly good, but not good
>enough for this. A 100kHz (10 us period) signal sampled with a jitter of 10 us
>is essentially useless.
>
>
>Going back to my soundcard suggestion, you could buy a 96ks/s stereo soundcard
>for less than $100 these days. Take your input signal, mix it (multiply) by a
>48 kHz sine and cosine wave, lowpass filter those 2 outputs, and feed to the
>soundcard stereo inputs. You'll have sampled the spectrum from 0 to 96kHz at
>24
>bits per sample.
>
>Matt
Dear Matt,
Thanks for your comments. For the record I haven't yet given up on the sound
card; I'm pursuing the evaluation in parallel. (No pun intended)
You've understood the requirement exactly regarding the interrupt-driven aspect
of this work and I share your concern about the effect of the high interrupt
rate. We'll see what happens with my measurements...
Regarding the suggestion of mixing and the use of a dual channel sound card:
Do 96ksample/s cards have 48kHz of actual bandwidth or will some analog filters
need to be disabled?
How wide of a hole surrounding 48kHz will this scheme have?
Can the sampling clock for a sound card be derived externally?
Your scheme gives me another idea to use the sound card: use a pair of
track/hold devices to sample the raw analog voltages, one at a time, at 192kHz
and offer these alternating time points to the two channels of the sound card
at 96kHz, generating the timing externally as I already have as a requirement.
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