Re: Maximum scan rate & external reference for AO

From: Yoshiya Matsuzaka <matsuzay_at_mail.tains.tohoku.ac.jp> · Date: Fri, 15 Jun 2007 16:59:24 +0900

Hi Marco,

  I am studying neurophysiology, too, and am developing an acquisition 
system for multichannel recording . Currently, there are several 
commercial multichannel recording systems in the market (e.g. Plexon, 
NeuraLynx), but they are terribly expensive (perhaps cost more than a 
Benz). So I decided to develop one on my own using a relatively 
inexpensive DAQ board.
  Regarding your question No.1, I found it easy.

comedi_cmd c, *cmd = &c;

    cmd->start_src            = TRIG_NOW;
    cmd->start_arg            = 0;
    cmd->scan_begin_src  = TRIG_TIMER;
    cmd->scan_begin_arg  = 0;   
    cmd->convert_src        = TRIG_TIMER;   
    cmd->convert_arg        = 0;
    cmd->scan_end_src     = TRIG_COUNT;   
    cmd->scan_end_arg    = N_CHANS;
    cmd->stop_src            = TRIG_NONE;   
    cmd->stop_arg            = 0;

  Then call, comedi_command_test for 3 times in a row. Then, the COMEDI 
library sets the field convert_arg so that the board collects the data 
at its maximum sampling frequency (I tested this technique only for NI's 
PCI-6071E, though). The real challenge is how to save the data to the 
disk. I read the data in kernel space under RTAI, send it over RT-FIFO, 
and let a user space program save it. This is much faster than reading 
the data by  read(comedi_fileno(comedi_t*), char*, size_t) in user space 
(in my box, the latter method causes the board's FIFO overflow when the 
sampling frequency exceeds 400kHz while the former allows sampling at 
1.25MHz).
  For question No.2, I haven't tested analogue output yet unfortunately.

  By the way, what software are you adapting for neuronal recording? RT-Lab?

Yoshiya Matsuzaka

Marco Hackenberg wrote:
> Hi,
>
> I am adapting a software for neuronal recordings to comedi in order to 
> obtain universal support for any kind of DAQ board.
> There are two open questions to me:
>
> 1. Is there any possibility within comedilib to get the maximum 
> scan/sampling rate supported by a board for asynchronous analog 
> acquisition ?
>
> 2. For my NI-E card, external reference for analog output is supported 
> by comedi via the actually not existing range indices 3 & 4.
> Is this the general strategy for all comedi-drivers to support 
> external reference?
> Can I somehow find out generically which output ranges are reserved by 
> the current driver for external reference?
>
> Thanks in advance for your help!
>
> Greets,
> Marco
>
> _______________________________________________
> comedi mailing list
> comedi_at_comedi.org
> https://cvs.comedi.org/cgi-bin/mailman/listinfo/comedi
>

-- 
----------------------------------------------------
Yoshiya Matsuzaka
Dept. of Physiol., Graduate School of Medical Science,
Tohoku Univ.
2-1 Seiryo machi, Aoba ward, Sendai 980-8575, JAPAN
Tel: +81-22-717-8073, Fax: +81-22-717-8077
----------------------------------------------------