Can I do frequency measurements on more than 2 inputs of the USB-6255?

[Tobin]

Hi,

I'm planning to purchase the USB-6255 multi-function DAQ. I will be using almost all of the 80 single ended analog input lines and all of the digital input lines. I want to use up to 6 digital inputs for frequency measurements. These will be square wave signals from zero to 2200 Hz.

Can this be done with this hardware? Or, am I limited to just the 2 counter/timer inputs for frequency measurements?

Thanks.

[NI-Bongo]

Hi Tobin,

You are only able to do the frequency measurements with the counters.  Because you will be doing up to 6 channels of frequency measurements, I'd recommend trying a PCI-6601 (4 Counters, plus the 2 on your M Series) or the PCI-6602 (8 Counters so you could do all your frequency measurements on one device).  Here's a link to the data sheet on these devices.

Counter/Timer Devices
http://www.ni.com/pdf/products/us/4daqsc386-393_386-387.pdf

Let me know if you'd like to speak with a technical representative about the solutions we have to offer.


Message Edited by NI-Bongo on 02-20-2008 09:00 AM

[Tobin]

Thanks for the response John.

That throws a big monkey wrench into my plans. The DAQ that I'm setting up is for use in a test vehicle with a 12vdc power source. My plan was to use USB devices connected to a notebook computer. PCI and PXI are not preferred for this task.

[stilly32]

Hey Tobin,

While the 6255 and all USB M-Series devices only have 2 counters, there may be other options. You could use the correlated digital inputs (if you're not already using them) to acquire at up to 1 MHz (sometimes more on some systems), and then in post processing get the frequencies of the digital signals. You would only have 1 MHz timing resolution rather than 80MHz, but for your lower frequencies this may be ok.

If you want the timing resolution and to stick with USB, you could look into the Bus Powered M-Series, like the NI USB-6210. They have the same counters (actually slightly better) as the externally powered, cost less and wouldn't take up that much space. They'll only have 2 counters as well, but you could buy 2 more for a total of 6 counters.

Hope this gives you some other options,

Andrew S 

Message Edited by stilly32 on 02-20-2008 12:28 PM

[Tobin]

Hi Andrew,

Let me see if I understand this. Are you saying that I can gather my 70+ analog signals at around 750kS/s (aggregate) while gathering my digital input signals at around 1MS/s and then use some processing in my VI to calculate a reasonably accurate frequency measurement on 6 of those digital inputs? Say enough to distiguish 2000 Hz from 2020 Hz?

I'm less concerned about absolute accuracy than I am with relative accuracy. I will be comparing each input's frequency to the other inputs and looking for differences. Four of the frequency inputs are of wheel velocities. I want to be able to recognize vehicle turning radius based on wheel speeds along with detecting if a wheel has broken traction (over or under speed).

Tobin G.

[stilly32]

So you should be able to do AI and DI at those rates at the same time - at least I can on my Dell GX280 (~3 year old machine). As far as frequency accuracy, it helps me to think in terms of period accuracy, because that's what we would measure directly. At 1 MHz, you would have 1 microsecond (us) resolution, so you wouldn't see a change less than 1 us, and when the actual change in a digital line occured would be uncertain to 1 us.

For 2000 hz vs 2020, the period of 2000hz would be 1/2000 = 500us and for 2020, 1/2020 = 495.05 us. So for a digital sample rate at 1 Mhz, Basically what you would see is (500us period/1us per sample = ) 500  samples between rising edges at 2000 hz, and 495 samples for 2020. Based on the alignment of the measurement and exactly when the measured pulses come with respect to the clock you would want to figure in +/- 1 sample clock tick of uncertainty, but it looks like this would be ok.

Hopefully this gives you an idea of the accuracy you could expect, and whether it will work for you. Please post back if you have additional questions, or I need to clarify any of what I stated above.

Cheers,

Andrew S  

[Tobin]

Thanks for your response Andrew.

This gives me tremendous hope that I can perform the task with a single USB device.

I was starting to consider adding a second device that has 10 counter timer inputs (USB-4303 from Measurement Computing, which, as I understand it, is owned by NI)

Thanks again to all that responded,

Tobin