Re: negative voltage reading problem with cmd.c and ni_pcimio

Hi!

Sorry, I meant to say that sampl_t in comedi is 16bits where as my ADC gives
12bit data (maxvalue = 4096). so I don't think I want to XOR the top 4 bits.
Is that right? When I ^0x0800, the negative numbers turn out fine but the
positive ones shoot up.

Naoko

Quoting Bernd Porr <BerndPorr_at_f2s.com>:

> Hi!
> 
> If it's 16 bit, just try ^0x8000. That should do the job. A 0 in comedi 
> means the most negative value and a 65535 is the most positive value if 
> you have 16 bit. Can't remember the functions exactly but there is a 
> function to get the maxvalue (here 65535) and there is one which 
> converts raw data to real physical values.
> 
> /Bernd
> 
> www:	http://www.berndporr.me.uk/
>          http://www.linux-usb-daq.co.uk/
> Mobile:	+44 (0)7769 506746
> Work:	+44 (0)141 330 5237
>          Department of Electronics & Electrical Engineering
>          Room 519, Rankine Building, Oakfield Avenue,
>          Glasgow, G12 8LT
> 
> 
> Naoko Kurahashi wrote:
> > Hi,
> > 
> > Thank you for your advice. I adventured my way through the sourcecode
> > ni_mio_common.c. (I'm still confused about the difference between high
> and
> > low data which are divided after ni_readl)
> > I've been trying XOR-ing as you suggested. Since 'data' is 16bits, I
> tried
> > data^0xffff and this changes negative values to positive values and
> gives
> > something between 2048 - 4096 (upper half of my 4096 resolution, not
> lower),
> > but I still don't know how to change it to the comedi equivalent of
> negative
> > voltages (0-2047 in my case). also, if I do this operation to 'data',
> if the
> > input was truely positive, the value of 'data' again shoots up to the
> > 63000's. Is there a good way around this? Did I understand your advice
> right? 
> > 
> > I don't have a formal background in signal processing or engineering
> for
> > that matter, which means I'm misunderstanding many things, but I
> thought 2's
> > complement data format means that I have to subtract 1 before I invert
> the
> > 1's and 0's. Is this still the case here?
> > 
> > Thank you for your patience in advising me. 
> > Naoko
> > 
> > Quoting Bernd Porr <BerndPorr_at_f2s.com>:
> > 
> > 
> >>Hi!
> >>
> >>Seems so that the sign bit of the converter is not interpreted in the 
> >>right way. Comedi uses unsigned data but some A/D converters give the 
> >>data in two-complement form. The driver should "undo" this by XOR the 
> >>appropriate bit(s):
> >>
> >>comedi_buf_put(s->async,(this_usbduxsub->inBuffer[i])^0x800);
> >>
> >>Might be a bug in the driver.
> >>
> >>/Bernd
> >>
> >>www:    http://www.berndporr.me.uk/
> >>         http://www.linux-usb-daq.co.uk/
> >>Mobile: +44 (0)7769 506746
> >>Work:   +44 (0)141 330 5237
> >>         Department of Electronics & Electrical Engineering
> >>         Room 519, Rankine Building, Oakfield Avenue,
> >>         Glasgow, G12 8LT
> >>
> >>
> >>Naoko Kurahashi wrote:
> >>
> >>>Hi, 
> >>>I've been battling a weird problem today of not being able to acquire
> >>>negative voltage values with my NI PCI MIO card. When executing the
> cmd
> >>
> >>demo
> >>
> >>>program, for example, every time the input voltage is < 0, the
> >>
> >>acquired
> >>
> >>>value shoots up to 63000 or so (max res for my card is 4096). I have
> >>
> >>my
> >>
> >>>range set to 3 (bipolar +/-1V). when I set it to unipolar 0-5V, the
> >>
> >>negative
> >>
> >>>values just give 0 and everything looks fine.
> >>>
> >>>Maybe my problem is hardware related, or I'm just missing something
> >>
> >>very
> >>
> >>>simple because I'm very tired, but I am out of things to try. Please
> >>
> >>let me
> >>
> >>>know if you have any suggestions or have experienced similar
> problems.
> >>>
> >>>Thank you,
> >>>Naoko Kurahashi
> >>>
> >>>_______________________________________________
> >>>comedi mailing list
> >>>comedi_at_comedi.org
> >>>https://cvs.comedi.org/cgi-bin/mailman/listinfo/comedi
> >>
> > 
> > 
> 
> _______________________________________________
> comedi mailing list
> comedi_at_comedi.org
> https://cvs.comedi.org/cgi-bin/mailman/listinfo/comedi
>

Received on 2004-06-15Z22:20:09