Need help about output of PXIe-5413

[whdxlad]

Hello, I encountered a problem using PXIe-5413. When I changed the output amplitude, I found that the amplitude changed from the original value to 0V, and then from 0V to the voltage value I need to change. How to change directly from the original value to the changed value, that is, when transmitting the command, 5413 outputs the original voltage value instead of disconnecting the output 0V, is it using wait behavior?
怎么实现，谢谢
How to make it happen? thank you.

[balexand]

Hi whdxlad,

The behavior you are seeing is by design.  I cannot see what the horizontal scale is on the screenshot you attached, but I would assume that the time that this 0V level is active is on the order of 10 - 100 ms.

The 5413 driver is designed is to provide optimal analog performance from the output.  As such, there is attenuation circuitry on the analog front-end that is switched when the user changes to a different amplitude.

What is happening here is that the driver software is:

1. Zeroing out the output of the DAC.
2. Writing registers on the device to change attenuation circuitry of the analog front-end of the instrument.
3. Waiting an appropriate amount of settling time for the attenuation circuitry to settle.
4. Resuming normal output from the DAC.

By design, you will see this type of "glitch" when you change the FGEN amplitude attribute.

I assume you are using the Standard Function mode of the device.  My recommendation, if you want to get rid of these glitches, would be to try changing output level using the Digital Gain attribute:

C: http://zone.ni.com/reference/en-XX/help/370524V-01/fgencref/nifgen_attr_digital_gain/

LabVIEW: http://zone.ni.com/reference/en-XX/help/370524V-01/nifgenpropref/pnifgen_digitalgain/

Set your amplitude to the max value that you'll need, and then use the Digital Gain attribute to attenuate your output signal.  This should change your output level without any glitch.  When using a digital gain less than 1.0, you will not have the same spectral quality as a sine wave that is produced with a digital gain of 1.0 with the front-end circuitry switched to a different amplitude setting.  I'm assuming that this likely won't matter for your application.

I haven't double-checked this solution with actual hardware, but based on my knowledge of this driver, I think this should be a workable solution.  Let me know how it goes.

Sincerely,

Brian Alexander

[whdxlad]

Hello Brain,

Thank you very much for your warm reply, your answer let me know more about the operation steps of the PXIe-5413 signal source card, and you provided me with another new way to drive PXIe-5413.

1. In fact, I use Pxie-5413 to drive a mechanical device, 5413 output different voltage values will lead to different mechanical states. I change the output voltage of 5413 to realize the automatic control of the state of mechanical device. Sometimes the output voltage of 5413 is from small to large, sometimes it is from large to small, and the output voltage of 5413 is changed by small steps. because the change of the voltage value by a large step may lead to damage of the mechanical equipment.

2. I learned the method you provided to me about the digitally control the output voltage of 5413. This method requires 5413 to output a large voltage, such as 12V, by changing the coefficient less than 1 to control the output voltage of 5413. Because the sudden output of a large voltage of 5413 is disadvantageous to the mechanical equipment, and I have a requirement for the resolution of the output voltage of 5413 in the control process, sometimes the voltage amplitude that I need to change is 0.001V, so it is difficult for the digital method to achieve high-resolution voltage output. so this method is not very ideal for my use.
3. The analog voltage output mode may be more suitable for me. I have used Agilent's signal source before, and it’s driver is also provided by NI. There are two kinds of driver output results, one is the same as the current 5413 output result, each voltage change needs to be cleared, and the other is to directly realize the voltage conversion, such as 0.5V-1V, instead of 0.5V-0V-1V. Can the 5413 signal source card also achieve this function? When giving the command to change the voltage to the hardware, maintain the original voltage output, do not carry out zero operation, and when commands are completed, directly output the new voltage value, so as to jump directly from the original voltage value to the new voltage value. The behavior command is placed in the test program I provided, but I feel that it did not work. I don't know what went wrong.
Can the PXIe-5413 signal source card realize the direct voltage conversion? Thank you!

Sincerely,

whdxlad

[Fanley]

Hi, Brain:

   I am TSE Fan Wang in Shanghai, China. Thanks for your instruction on this problem. It seems that using the Digital Gain attribute is not a ideal way to solve it in this special application. I use another product PXIe-5442 to run the same program and such a 'glitch' is not seen. 

   Is there any difference in hardware design between 5413 and 5442? In other words, what type of product should be recommended to customer in his application? 

regards

Fan

[balexand]

Fan,

Please check more amplitude changes on the 5442.  You may be getting lucky that the specific change you are making does not require a lot of analog front-end circuitry to be changed.

If I'm not mistaken, most of our arbitrary waveform generators will have some degree of 'glitching' when changing amplitude in certain cases, but not all cases (e.g. you would expect to see less of a glitch when changing from 1V to 1.1V as opposed to changing from 1V to 12V.

The block diagram of the 5413 Analog Output Path is documented here: https://www.ni.com/documentation/en/pxi-waveform-generator/latest/pxie-5413/analog-output-path/

It says:

8. The signal continues through either the low-gain amplifier path or high-gain amplifier path.

• The low-gain amplifier path is used for waveforms that have all output voltages equal to or less than ±2.75 V into 50 Ω loads. The low-gain amplifier has a fixed gain.
• The high-gain amplifier path is used for waveform output voltages greater than ±2.75 V into 50 Ω loads. The high-gain amplifier has a fixed gain and enables maximum peak-to-peak signal amplitude (Vpk-pk).

On the 5413, there is a big transition between 2.8 V and 2.7 V.

2.8 V will switch the High/Low-Gain Path Relay into high-gain mode, and the Medium / Coarse attenuation stages will apply a large amount of attenuation.

2.7 V will put the High/Low-Gain Path Relay into low-gain mode, and the Medium / Coarse attenuation stages will apply almost no attenuation.

As this analog output circuitry is switched and attenuation stages are switched, the signal could have some intermediate output levels that are undesirable, especially at high frequencies (try looking at this behavior on a 20 MHz sine output for example, on both the 5413 and 5442).  Our decision was that it was most simple, clean, and consistent to zero-out the DAC while the circuitry was being changed.

Regarding whdxlad's suggestion:

I have used Agilent's signal source before, and it’s driver is also provided by NI. There are two kinds of driver output results, one is the same as the current 5413 output result, each voltage change needs to be cleared, and the other is to directly realize the voltage conversion, such as 0.5V-1V, instead of 0.5V-0V-1V. Can the 5413 signal source card also achieve this function?

There may actually be a way that we can configure the device so that the DAC output holds the last value while we do the transition, and we take care to always program things in a way that enables the smoothest transitions.  That would take some R&D time so we would probably need a strong business case to do this work, and this type of decision would have to go through our product planning staff.  I will let them know about this feature gap in the product and we can consider further investigation into a potential solution here.

Sincerely,

Brian Alexander

[whdxlad]

Hi Brain,.

I am glad to receive your reply again, our newly purchased 5413 cannot be used because of this zero clearing action. I used to think that not all NI signal source cards had this problem, but I thought that I could replace other types of cards with NI to solve this problem, such as 5402, etc., but after reading your explanation, it is possible that all NI signal source cards have this problem, so the solution of replacing the signal source card cannot solve the problem.

We still want to make perfect use of the newly purchased 5413 instead of wasting this card. I am glad to see that you have a solution to solve this problem and enable 5413 to achieve a smooth amplitude conversion. if so, I would like to ask how long it will take to implement this scheme, and whether it can be solved in a short period of time.

We look forward to your solving this problem as soon as possible.

Thank you.

Sincerely,

whdxlad