how do i create AO voltage waveforms of about 1MHz with high speed M series DAQ?

[scarface]

i have a M series high speed multifunction DAQ card, and i need to continuously generate waveforms of approximately 1MHz in frequency through multiple voltage AO channels. i then need to continuously aquire these waveforms through multiple AI voltage channels and write the data to graph and file.so far i have only been able to generate frequencies of about 10-100KHz.i am quite new to programming in labview with DAQ cards and i'm not sure about how to access and use the different clock rates to generate such a high frequency while maintaining a high resolution in the aquisition of the data at the end.can anyone help?

[YatinM]

Hi Scarface

May i ask which M-Series card you are using? Unfortunately no M-Series card would able to cope with the requirements you stated in the last email. To generate 1Mhz across multiple channels can not be achieved by any M-Series card. There are only four M-Series cards that are capable of generating upto 1.25Mhz sampling rate and two of them cards do not have analog output capabilities. Also it is 1.25mhz aggregate which means across all the channels. (ie. 1Mhz across 4 channels means 250kHz per channel). The only way to get close to what you would like to achieve is by using an S-Series card, although the most AO channels they have are 2.

Here is a link to the NI website for S-Series cards

Hope this helps

YatinM
National Instruments UK

[scarface]

hi yatim,

thanks for the reply to my question.the card is a NI PCI-6259 M series DAQ.would it still be possible to generate a single 1MHZ AO and then read this waveform (sine wave) through a single AI?also, ultimately, what we want to do is to generate continuous single digital pulses (with specific amplitude and frequency and also of a certain width) at MHz frequencies through 2 voltage AO's, simultaneously, and then read in these pulses, simultaneously, through 4 voltage AI's.would it be possible with digital pulses instead of analog waveforms, or would they still be limited to a maximum frequency of about 250kHz?

regards

steve

[YatinM]

Hi Steve

To generate the 1MHz across the 2 AO channels should be ok as they work off the same ref clock. The fact you are transfering so much data there is a risk that accessing 4 AI channels at such hi-speed might prove difficult to keep an appropriate resoloution. You can simulate doing this in LabVIEW with the use of DAQ Assistant. When i try simulating the 4 AI channels the memory buffer isnt fast enough to keep up with the data and falls over. Using a S-Series 6120 should fulfil all your AI and AO issues.

Hope this helps,

Best Regards

YatinM

NIUK

[scarface]

hi yatim,

thanks again for the help.i think 250 KHz should be ok for the AI.could you give me some tips for writing a program with 2 AOs and 4 AIs?

thanks again and best regards,

steve

[Tom_W_[DE]]

Hi All-

A quick correction to one of Yatim's earlier posts- the NI 625X Specifications list maximum AO update rates on page 3.  These rates describe the per channel maximum rates when the indicated number of channels are in use.  So, for the PCI-6259 you can operate all four channels at 1.25MHz.  The rates specified are per-channel rates and do not represent an aggregate rate to be shared across multiple channels.

Hopefully this helps-

[scarface]

cheers guys, i've managed to rectify those problems with your help.most appreciated.

steve

[Kevin_Price]

Highlighting the portion that hasn't been addressed:

---

the card is a NI PCI-6259 M series DAQ.  would it still be possible to generate a single 1MHZ AO and then read this waveform (sine wave) through a single AI?also, ultimately, what we want to do is to generate continuous single digital pulses (with specific amplitude and frequency and also of a certain width) at MHz frequencies through 2 voltage AO's, simultaneously, then read in these pulses, simultaneously, through 4 voltage AI's.  would it be possible with digital pulses instead of analog waveforms, or would they still be limited to a maximum frequency of about 250kHz?

---

You could generate digital pulses either with 2 counters or with 2 bits worth of digital Port 0.  It would be possible to control both frequency and width to sub-microsecond precision, but the amplitude would be fixed at ~5 V.  Using the counters would likely be a much "cleaner" solution than digital output if there are long precision-timed delays between pulses.

I'm not sure why you'd use 4 input channels to read 2 output channels.  Nevertheless, depending on the purpose, you *might* be able to do this using a different 4 bits from digital Port 0.  It's possible to run timed digital output and timed digital input simultaneously.

-Kevin P.

[scarface]

cheers kevin,

i think using a counter would be a better way forward.i have another question, is it possible to set up an initial time delay (~ms - microseconds) between the 2 counter outputs (counter outputs are still of the same frequency and amplitude) and then have them run as per usual?

steve

[Kevin_Price]

Short answer: yes, the two counters can have a different value for initial delay.  You would also need to configure both of them to use the same digital start trigger so that their different initial delays begin at the same t=0 mark.

Caveats:

1. You keep referring to "amplitude."  Signal amplitude is NOT something you can program using counter or digital outputs.  You can only get standard logic high and low outputs of ~5V and ~0V.

2. Both counter pulsetrain outputs will need to be continuous rather than finite.  (1 finite pulsetrains output would consume both counters.)

General: can you take a couple steps back and describe the overall purpose of this app?   What decision needs to be made and how do these signals help to inform that decision?

-Kevin P.