continuous 8 channel acquisition using win32 api and DAQ

[electrobody]

Hi all,

I'm trying to acquire signals from multiple channels in my win32 API application using traditional DAQ and win32 callback procedure (the same I use to handle my GUI
events). The acquisition should use double buffering, write results to a file and stop only on a GUI event. Could
anybody give me any code that would be a starting point for that? (the important thing is that I'm not using CVI, etc., just DAQ). Even a short code showing how to implement acquisition's callback as a part of win32 callback procedure would be helpful.

Regards,
Simon

[Tom_W_[DE]]

Hi Simon-

What type of DAQ device are you using?  What you're trying to accomplish may be easier using NI-DAQmx. 

At any rate, a great double buffering example for Traditional NI-DAQ that installs with the driver is "DAQdoubleBuf.C" and can be found in ( C:\Program Files\National Instruments\NI-DAQ\Examples\VisualC\Ai ).  How exactly are you trying to implement the callback procedure (i.e. are you trying to only read from the backlog at specified acquisition intervals, etc)?

Thanks-

[electrobody]

Tom,

Thank you for your reply. I'm trying to use NI 4472 for continuous multi-channel signals acquisition (double buffering) as a part of my masters dissertation. The callback function should only write data to file whenever the half-buffer is full.

I have found a "Using DAQ Event Messaging under Windows NT/95/3.1 - Application Note 086) document, which includes basic example of using windows callback function to handle acquisition process messages (attached to this post), but when I execute it, I get "Error code -10080 at SetupNIDAQEvents) message.

Could you please give me a hint what's wrong or where I should look for some code examples? Or maybe DAQmx code examples for Visual Studio C++ would suit my needs better?

Regards,
Simon

[Joji]

 If you are using a DSA device, modify the iGain varialble to be in units of dB. Valid gain settings for DSA devices are -20, -10, 0, 10, 20, 30, 40, 50, and 60.

See the attached example or at this link: http://sine.ni.com/apps/we/niepd_web_display.display_epd4?p_guid=B45EACE3D85456A4E034080020E74861&p_no...

Most of the examples usually are set up for E series devices which have different  gain settings which would be invalid for a DSA board.

Buddy Curtis

Applications Engineer

Modular Intruments

[electrobody]

Hi,

Thank you for your help. The gain setting was the answer! Now I'm trying to continuously acquire data from two analog channels using double buffering (as a starting point for 8 channel acquisition). My code seems to be working fine: the double buffer is being filled up with the right amount of data. But the acquired samples have strange values, eg.:

0   -190267 71522
2   -197805 -112550
4   -188074 -86492
6   -198216 -138331
8   -183551 -42830
10  -185196 26336
12  -187800 -88986
14  -190952 -32296
16  -182181 56830
18  -184373 12752
20  -183277 -74987
22  -185059 52672
24  -186292 40890
26  -179302 -28414
28  -179714 41167
30  -177658 -36177
/.../

There are some positive values at the second channel, which seems strange, as the values should be, as far as I know, in Db (so should be negative)?

Best regards and thank you for help,
Simon

Message Edited by electrobody on 01-23-2006 07:49 AM

[Joji]

Please refer to the Traditional DAQ C Reference on the Scan_Start.  It specifically states that "For DSA devices only, buffer should be an array of i32. These devices return the data in a 32-bit format in which the data bits are in the most significant bits."

You are only using i16 (or 16-bit buffer), so you are getting strange data back.

P.S. You can find the C Reference from Start Menu >> All Programs>>National Instruments>>DAQ>>NI-DAQ

[electrobody]

Buddy,

Thank you for your reply. If you take a look at the code attached to my post:

static i32 ipBuffer[1999] = {0};
static i32 ipHalfBuffer[999] = {0};

and in scan_start the buffer being is casted to i16*

iErr = SCAN_Start(iDevice, (i16*)ipBuffer, 1000, 0, 0, 0, 1);

Isn't that enough? I still cannot sort out the strange scan results.

Another question I have is:
If I acquire data from 8 channels using double buffering, should I configure the event of scanning the right amount of samples for one half-buffer transfer operation just for one channel, or for all the channels?

eg.
iErr = Config_DAQ_Event_Message(iDevice,(i16) 1,"AI0" ...
or
iErr = Config_DAQ_Event_Message(iDevice,(i16) 1,"AI0:7" ...

Best regards,
Simon

[Joji]

It still seems to me that it is the alignment of 16 bit to 32bit.

Can you take a look or even run this C code in the link below

<a href="http://digital.ni.com/public.nsf/websearch/B82DB701019383B18625673C006835AC?OpenDocument">PCI-6110E, PCI-6111E and DSA boards Return Corrupt Data on Analog Input when Transferring Data to an Unaligned 32-bit Buffer</a>

I also saw another way doing the same thing that you might want to try.

static i16 acqBufPtr[2 * NUMSAMPLES + 1]; // Allocate buffer to be at least one word larger than required

if((u32)acqBufPtr % 4) // If buffer is not aligned on 4-byte boundary, adjust pointer
acqBufPtr += 1;

iStatus = DAQ_Op(iDevice, iChan, iGain, acqBufPtr, ulCount, dSampRate); // Pass aligned buffer to acquisition function

[Joji]

It still seems to me that it is the alignment of 16 bit to 32bit.

Can you take a look or even run this C code in the link below

PCI-6110E, PCI-6111E and DSA boards Return Corrupt Data on Analog Input when Transferring Data to an...DIV>

I also saw another way doing the same thing that you might want to try.

static i16 acqBufPtr[2 * NUMSAMPLES + 1]; // Allocate buffer to be at least one word larger than required if((u32)acqBufPtr % 4) // If buffer is not aligned on 4-byte boundary, adjust pointer acqBufPtr += 1; iStatus = DAQ_Op(iDevice, iChan, iGain, acqBufPtr, ulCount, dSampRate); // Pass aligned buffer to acquisition function

Buddy Curtis Applications Engineer National Instruments