03-10-2008 11:31 AM
03-11-2008
04:46 PM
- last edited on
05-03-2024
05:12 PM
by
Content Cleaner
Hello Tobin,
The reason you're seeing this error is because on the PCI-6255 only 8 of the 24 digital lines are clocked. On the remaining 16 lines you can only do software timed input or output. You can set the sample rate for the 8 clocked (hardware timed) lines, but if you want all 24 lines to read in at the same rate you'd need to configure your acquisition to be software timed on all 24 inputs. To do this you can choose "1 Sample (On Demand)" in the DAQ assistant and put it inside of a while loop to acquire one sample at a time from each line with software timing.
The exact sample rate will be highly dependent on your computer as well as and any other analysis you may have in the loop, but all of the lines will be acquired at the same rate. If you require a sample rate of 15.625 kS/s/channel on all 24 digital input lines then you may need to consider getting another board. I wouldn’t expect you to get a sample rate much higher than 1 kHz, if that, using software timing. You can however set the rate as you would on the analog input if you only use the first 8 lines which are clocked.
I hope this helps explain the cause of this error. If you require 24 clocked digital IO lines as well as 80 single ended analog inputs you’ll likely need to get 2 separate devices, but if you can lower the number of analog inputs to 32 or the number of clocked digital lines to 8 there are solutions available in a single card. The NI PCI-6254 has 32 single ended analog inputs and 32 clocked digital lines while the NI PCI-6255 you have will work if you only need 8 clocked digital lines.
Cheers,
03-12-2008 07:55 AM
Brooks,
Thank you for your detailed response.
If I were to add another device, say the USB-6259 to my system with the USB-6255, along with a high speed CAN interface USB-8473s (all on the USB of my notebook computer), would I be trying to get too much information through the USB?
Here is what I would be attempting: 70 to 80 SE analog inputs at between 10kS/s/channel and 15.625kS/s/channel plus 8 clocked digital inputs of the 6255 plus all 32 clocked digital inputs of the 6259 (sampled at the same rate as the analog inputs) plus a high-speed CAN monitor.
I show more than the total of 24 digital inputs here because I really have 45 digital signals that I want to look at, but, originally, I was going to repeat my tests with different digital signals monitored.
The notebook pc is WinXP with a dual core processor at 2.4GHz and 3.5GB Ram. My biggest concern is the USB throughput. There are 3 USB ports on the notebook pc, but I don't know if they are all sharing the same host controller, or if each one has the full bandwidth available to it.
03-13-2008 01:33 PM
Hi Tobin,
I just wanted to give you a heads up that the USB-6255 has an aggregate (multi-channel) maximum sample rate of 750 kS/s. To calculate the maximum sample rate/channel you divide 750 kS/s by the number of channels you're scanning. That means that if you used all 80 channels on the USB-6255 your maximum sample rate would be 9.375 kHz. I'm assuming you knew this because if you got the USB-6259 as well you could get this rate up to 15.625 kS/s by using 32 analog inputs on the 6259 and putting the remaining 48 on the USB-6255. I just wanted to make sure you hadn't arrived at that number by random chance.
You bring up a valid point with concerns about the USB bus. For USB 2.0 the maximum rate for the standard is supposed to be 480 Mbit/s. It’s difficult to say the exact data transfer rate of these USB products, but generous overestimates still put it well beneath the theoretical limit. Here's what I calculate:
80 analog inputs and 40 digital inputs = 120 total channels.
Sample rate of 15.625 kHz on each channel * 120 channels = 1.875 MS/s.
Assume 24 bits/ sample (it is actually probably closer to 16bits/sample) = 45 Mbits/s.
Add on the transfer rate of the high speed CAN card you have and if you're still beneath this limit it should work.
This is a theoretical limit and various computer manufacturers implement the standard to different levels of accuracy, but if you're 40-60% under that limit I wouldn't expect you to run into any trouble.
Let me know if anything was unclear and have a good night.
Cheers,
03-13-2008 02:33 PM
Thanks Brooks,
I was being sloppy with my numbers. At one time I had calculated the 9.375kS/s/channel number and I accepted that limitation. As long as these DAQ devices support the Hi-speed (480Mbps) USB as apposed to Full-speed (12Mbps), then maybe I'll be ok.
Thanks again.