Problems using IMAQ snap with PCI-1405 and JAI CV-M50 Analog Camera

[Doug1974]

I am currently using a PCI-1405 to capture video data from a JAI CV-M50 Analog Camera.

I've been experiencing frequent error messages as follows:

IMAQ Error 0xBFF60022, "Did not receive all expected video data within the timeout period."

Using MAX 3.1.1, I've configured my camera for CCIR mode (supported by the camera). 

I've tried various width/height combinations which should be supported by this mode.  

When I set up the camera using a new profile and use default CCIR settings, then leave MAX without editing anything, I can grab and snap images

without issue in my application.  After some arbitrary amount of time, I begin experiencing the aforementioned error. 

Upon entering MAX to see what's up, I notice that the width/height have been changed to 8 x 1 and I can still snap images (at full size, not 8 x 1).

but I cannot grab as the operation will timeout after 3 frames or so..

I have absolutely no idea how to deal with this issue and I'm in dire need of help.  It should be noted that 'yes' I have tried changing the timeout settings.

I've tried 30 seconds.  I get 3 or so frames, then after 30 seconds, I get the same aformentioned error.. This occurs after some time in Both MAX and my application.

Also, is there a location on the NI website that contains actual camera files?  I cannot fine one for my camera.  Any attempts at loading any of the camera files installed

with IMAQ results in "currupt file" error message..

Please HELP!

[JesseD]

Hello,

The MAX behavior where you see the acquisition window set to 8x1 is a known issue that is actually caused by having DAQmx 7.4 or Motion Assistant 1.2 installed.  Our latest IMAQ release, 3.1.3, installs a new component that fixes this behavior.  You can download version 3.1.3.

The PCI-1405 is an analog framegrabber that only works with standard cameras.  Since the timing and configuration parameters are standardized, the only camera file options include CCIR, NTSC, PAL, and RS170.   You can access these by right-clicking on the channel in MAX and selecting the appropriate option from the "Camera>>Standard" menu.  If the wrong camera file is selected, then the error you mentioned could occur.  In addition to the error, you may also notice vertical and horizontal synch issues (where the image wraps top-to-bottom or right-to-left).  If you try to load any other camera file, you will most likely receive an error.

I highly recommend upgrading to IMAQ 3.1.3 to fix the window behavior first.  After upgrading, double check that the acquisition window is correct.  The default for top, left, width, and height for CCIR are as follows: 150, 22, 768, 576.  If you see black bars on the left or top of the image, you may need to increase the top/left offsets.

Let me know if this helps out.

Best Regards,

Jesse D.

Applications Engineering

National Instruments

[Doug1974]

Thanks for the quick response.. I had contacted tech support prior to reading your reply and was forwarded to the same conclusion regarding the IMAQ driver (ver 3.1.3).  My problems with funky image boundries seems to be resolved now.  I am still encountering the timeout errors however.  Is there a way to reset my card to continue grabbing frames if I encounter the error?  I've browsed through the various properties for camera control in the IMAQ vision builder VIs but haven't been successful in finding a solution.  I'm assuming if I can't solve the situation programatically, I'll either have to purchase a more flexible interface, or find a standard camera.  Any 'more' suggestions you or anyone else may have would be greatly appreciated.  We were blaming the PC we were using and I managed to convince the guy that pays the bills that I may have to procure another PC and have been granted a significant budget so I think it would cover the purchase of a new card or camera. 

I also have some concerns about the architecture of the frame grabbers available from NI.  The PCI-1405 card I have now is keyed for operation in standard PCI slots (my PC has 1 of these) as well as PCI-X slots (my PC has 4 of these).  My daq card is currently populating the standard slot and the imaq card is in one of the pci-x slots.. I have access to a PCI-1411 card but it's keyed for standard PCI usage (I can't sacrifice the daq card) -- If I purchase a new card, will it be keyed correctly?

Thanks again for the advice..

[JesseD]

Hello,

Another thought...is it possible that your camera is actually RS170?  JAI makes both CCIR and EIA (RS170) versions of the camera.  I have been successfully using the RS170 version here at NI for a little over a year with multiple IMAQ boards (including the 1405 and 1411).  If I change camera files to CCIR, I also get a few frames and a timeout.  Could you try changing the camera file to RS170 to see if the behavior changes (better or worse)?

As for resetting the board, there is no direct function to accomplish this.  When the error occurs, the session becomes invalid.  The way to recover is to close the current reference(s) and reinitialize the acquisition (Init, Setup, etc...).

The new revisions of the PCI-1411 are universally keyed (for both 3.3V and 5V), so it should work in PCI-X slots.

Best Regards,

Jesse D.

Applications Engineering
National Instruments

[Doug1974]

I'll give RS170 a try.. Just for fun, I set up a very simple loop to run over night where I'm grabbing and displaying frames and elapsed time while monitoring the error cluster.  The loop will stop when it encounters the error I have been experiencing.  The loop ran all night and was still running this morning when I got in (this is using CCIR btw..).  Then when I tried running my standard application (which makes use of machine vision analysis VIs) I eventually get the timeout errors after about 30 minutes to an hour later.. Anyhow, if I find RS170 yields the same results as with CCIR, would upgrading to the 1411 card be worth my while?  I'd like to continue using the same camera (CV-M50).

Thanks again Jesse..

[Doug1974]

I forgot to mention in my last reply -- I have tried RS170 and CCIR.  Both eventually run into the same problem (this was back before I upgraded IMAQ drivers to 3.1.3).  I've tried to repeat testing with RS170 and the same problem occurrs.  Switching camera types does not correct the problem -- the PC must be reset to restore frame 'grabbing' functionality.  Disposing the reference and reinitializing and setting up a new acquisition reference does not help either.  One lower-level observation I have made is when I enter this error state, the number of frames returned before timeout occurs toggles with each attempt.  First attempt will return 3 frames, second will return 13, then 3, then 13, etc.. I'm on an extremely tight schedule so I'll have to continue developing the software around a disabled IMAQ framework (which will be a nightmare as the main objective of the application is to analyze the images I'm grabbing).  I need reassurance that ordering the 1411 card (or some other card you could suggest below $2000CDN) will guarantee performance without these errors..

[RManion]

Hi Doug,

Can you clarify a few points regarding the errors?  First, is it true that you are able to acquire an image using a basic grab or snap program?  Assuming this is the case, I would also assume that you can acquire in Measurement and Automation Explorer with no problems?

However, when you attempt to acquire in your application (which does some sort of processing) you will receive timeouts after 30 minutes or so?  And this always takes roughly 30 minutes?  How does the 3 frames, 13 frames, 3 frames tie into the 30 minutes before timeout?

What kind of processing are you performing on the image in your program?  If the error is only occurring when you run the code that performs processing, this could be important information.

Finally, unless I have missed something, I do not believe purchasing a 1411 will change the situation in any way. 

Thanks, I am interested in discovering more about this error.

Robert

[Doug1974]

Hi Robert..

To clarify, the timeouts I experience would occur after some random amount of time (I think I mentioned 30 minutes to an hour).  When I enter this error state I cannot grab frames successfuly (with the exception of the 3 or 13 frames before timeout).  I can 'snap' successfuly in this state however.  I was in on Saturday (Aug 6) doing some work and had absolutely no problems grabbing frames all day (even with my machine vision VIs running).  The only way to recover grab functionality is to completely shutdown and restart the PC.  Powering down the camera, unplugging the and reconnecting the co-ax cable does nothing.  Even switching to another camera does nothing -- I think I had asked if there were any low level VIs that would allow me to reset the PCI-1405 that I'm using.  I'll just have to keep working with what I've got until I've received a better suggestion with regards to hardware. 

Do the video capture cards have reprogrammable firmware?  Just curious incase there is something wrong with the embedded software in my 1405 card..

Doug.

[Aaron B.]

Hello,

There is no way I know of to upgrade the firmware. I will double check to make sure that I have not overlooked anything.

Regards,

Aaron B.
National Instruments

[Doug1974]

I think I've come up with some clues as to why I've been experienceing timeout problems with my 1405 card and JAI CV-M50 camera..

As I mentioned in an earlier post, the PC I'm using has 1 legacy(?) PCI slot and the remaining slots are PCI-X.  I've removed my 6035E DAQ card from the standard PCI slot, and moved my 1405 card from the PCI-X slot to the standard PCI Slot.

I ran my application with DAQ features disabled -- the application ran error free for 2.5 days and is still working.  I am going to try to swap the hardware back to the way it was and just disable the daq functions and see if the same results are realized. 

Reading the manual for my motherboard (I'm sure this applies to most systems like this) -- when multiple devices are present on the PCI bus, the bus will clock itself down to the capabilities of the minimum card -- most likely the daq card.  I don't know what effect this would have on performance -- any advice? 

In the meantime, I'm going to try swapping PCs to one with all standard PCI slots and see if the same problems occur..

Doug.