glitch on dc signal

[sripapa]

Hello,

I am using Analogue module PXI 6713 A[5] pin in PXI chassis 1044, to generate a Glitch( amplitude 1.2V with a depth of 0.5V)  on a DC signal. The glitch on the DC signal through A[5] pin is achieved in the following way.

Method 1: (Please refer Glitch_2ms.JPG and VMON_Ramp.vi files)

1) Start generating the DC signal from 1.2V.

2) Once it reaches the 1.2V start decreasing the voltages in steps of 0.1V until it reaches 0.5V.

3) Once it reaches 0.5V start increasing the voltages in steps of 0.1V until it reaches 1.2V.

Repeat steps 1 , 2 and 3 ...

This way I am able to achieve the glitch width of 2ms with a decrease of 0.1V /interval. But I would like to achieve the glitch of width 1us or 10us. I followed below steps to generate the minimum glitch width.

Method 2: (Please refer Glitch_180us.JPG and VMON_Ramp_debug.vi files)

1) Start generating 1.2V DC signal.

2) Immediately generate 0.5V DC signal.

Repeat steps 1 and 2.

But the glitch width I could able to achieve is only 180us which doesn’t meet my requirement of 1us.

The Labview source code for  both the  methods  are attached.

Could you please give your inputs in order to achieve the glitch width of 1us order.

Thanks.

Sincerely,

Srinath.

x0069190@ti.com

816 728 2134

281 274 2988.

[kellis.wiseman]

Hi sripapa, 

With such a complicated application, you would probably be better served by calling in for support if you have a contract. That being said, I will help you as best I can. If I understand you, you need 15 steps within 1us, which means you need to be able to write at least 15MSamples/second. Your 6713 is not able to write that fast (it writes at 1MS/s). If you need a glitch width of 1us, you need will need another card. Look at the following advisor page (towards the bottom of the page) and anything over 15MS/s should work. 

Hope this helps. Good luck.

~kgarrett

[sripapa]

Hi,

Thanks for the reply.  I do agree that 6713 module has only 1M/s , but my concern is generatino of a ramp or glitch signal as shown in the figure. It need not have 15 steps. It could be any. All i need is jump from  1.2V to configurable dip( in this case in my earlier waveforms i configured for 0.5V). All i need is a small programme  which reads 1 Msamples per second , which does generates a gitch width of 1us and which jumps from 1.2V to 0.5V. I want to repeat this procedure with a configurable time period. Bottom line is 

1) Need to generate a glitch widht of 1us, with a dip from 1.2V to 0.5V, with samples of 1M/s.

2) Should be able to repeat at the configurable time period.

Please let me know if you need additional information.

Thanks.

[Kevin_H.]

I believe I understand what you're trying to do. You cards maximum sample speed for single channel analog output is 1MS/s. This means that 1/1MS/s = 0.000001s/S is the minimum interval between samples that you can write. This is precisely 1 microsecond minimum dt. As a result, trying to write a glitch at 1microsecond means that it will be instantaneous and not ramped (there is no room to ramp because it would have to happen in the interval between 2 samples). Also, are you trying to make the clitch 2microseconds wide (1 down and 1 up)? That is reasonable as you can have 1 sample of down time at .5V and the rest up at 1.2V. Otherwise you're working on too small of a time scale (.5 microseconds down and .5 up is too small of an interval for the maximum sample speed of the card).

Also, if you're trying to do regenerative output (meaning you load a set number of samples to the card), you're going to have to make sure that the FIFO size (which is somewhere around 16k samples) is big enough to give you the appropriate interval upon which you want to re-trigger the glitch. In the case of that cards FIFO size this is a very small (16ms) interval. Is that a small enough interval for retriggering? You can also do it in non-regenerative mode which would allow you more control, however you'll need to be very careful about underflow issues. See the example in labview about continuous non-regenerative internal clock analog generation for an example of how to do this (should be in Hardware Input and Output->DAQmx->Analog Generation). If your "configurable time period" is smaller than around 16ms, you'll want to do things nonregeneratively. There may be a way to do it with some sort of complex triggering system, but I don't think the card is going to be capable of that, I think non-regenerative software methods are the only way to accomplish it on a larger time interval.

You may also (if this seems to daunting) wish to be put in touch with an alliance member who might be able to help you get the code written. It will cost a little, but if this will take a sizable chunk of your time it may be worth it.