LabVIEW

cancel
Showing results for 
Search instead for 
Did you mean: 

How do you handle spare channels?

This is somewhat of a planning question regarding indexing and possible future utilization of spare inputs on CompactDAQ modules. What do you folks typically do when you are designing an application with a certain number of defined inputs and a number of spare hardware channels that may or may not be used in the future? Do you include and index the spares in the original application, or do you deal with them down the road if and when those channels are ever put into use? Is it better to have the spares integrated into the application scheme, or it it just as effective to tack them on at the end later? What are some of the pros and cons of both methods?

0 Kudos
Message 1 of 3
(2,310 Views)

Good Question. I normally won't include the un used channels into the application.

 

Facts

 

  • To avoid un-necessary sampling the channel as there is no actual use for the application
  • Saves CPU resource to be spent on sevicing the channel and thus eventually increases the frame rate (most of the time critical applications needs to avoid CPU usage on un-necessary things)
  • Can achieve high efficiency of the ADCs by reading only the selected channels
  • Yes its agian a time consuming one when you need to add channels in the future but when you design the application in such a way that it can be later added without much modification then no problem since when you add a channel obviously you need to add the calibration and other stuffs also so as a whole it will not be a big deal.

Hope this helps

-----

The best solution is the one you find it by yourself
0 Kudos
Message 2 of 3
(2,309 Views)

I think P Anand has listed several key points.

 

Only you can answer how likely the addition of channels will be.  If you feel that the extra work to add channels later will be greater than the work to include them in the original version, then it may be worth including them now.  Also consider whether the uused channels would be confusing to the end users.

 

Since you are considering the possiblity of adding channels, I think you should design the program so that doing so will be relatively straightforward. For example keeping the list of channels as an array and the data in multidimensional arrays (where one dimension is channels) simplifies adding channels. The data structures do not change, only the number of elements.  Make typedefs of controls (especially clusters and enums) which might need to be changed as channels are added. 

 

Lynn

Message 3 of 3
(2,303 Views)