SMU delay time

[frosturi]

I am using a PXI-4132 SMU with NI-DC Power version 15.0.  I am unfamiliar with this module and to get started I have been experimenting with one of the shipping examples.  NI-DCPower Hardware-Timed Voltage Sweep.vi.  This VI configures the source delay to 5ms.  However, when I measure the delay time with an oscilloscope I get consistent delay times of 39.5ms between steps.  I do not have a load on the SMU, am not trying to remote sense, and have the scope connected to the LO and HI outputs of the 4132.  So the correct voltages are set, but not at the correct interval.  The only modifications that I have done to the example VI consists of selecting the correct Resource Name.

The example VI does not specify any necessary hardware setup.  I'm not sure if there is something elese that I need to set up software or hardware-wise.

[Libbin_la_vida_loca]

Hey frosturi,

In general, you should expect the true delay to be the Aperture Time + Source Delay. In this case that would be around 16 + 5 = 21 ms.

https://knowledge.ni.com/KnowledgeArticleDetails?id=kA00Z000000P6CsSAK&l=en-US

I'm not quite sure why you are getting nearly double that, but I'll be setting up a PXI 4132 soon to test this out.

[frosturi]

Hi Libby,

I appreciate you taking the time to look into this.  Please let me know what you find out from your testing.  FYI, the rest of my system consists of a PXIe-1078 chassis, PXIe-8135 controller, the SMU resides in slot 7, and I am using LV2015.

Thanks

[Libbin_la_vida_loca]

Hey frosturi,

I was able to reproduce this issue and saw delays of around 40 ms between steps. After a bit of tinkering, I found that setting the Autozero property to Off allowed me to achieve delay times closer to 22ms.

Another forum post details a similar issue:

https://forums.ni.com/t5/Digital-Multimeters-DMMs-and/4132-SMU-measurement-time-is-too-high/m-p/1909...

[frosturi]

Hi Libby,

I was able to turn off the Auto Zero property and verify your results with an oscilloscope.  Why is the fastest time around 22ms?  The specification for the PXI-4132 states that the device has an update rate of 4200 updates/s.  Does this mean that the device should be able to support updating the voltage every ~ 238us?  If so how could I configure that?  Is there a way to turn measuring off, if I only care to set the voltage levels?  (I'm assuming that something like that could help speed things up).  I will eventually be switching to a PXI-4137, which has an update rate of 100,000 updates/s.  If I cannot acheive the 4200 updates/s on a 4132, then I'm concerned that I will not be able to acheive the 100,000 updates/s with the PXI-4137.

Thanks

[Libbin_la_vida_loca]

Hey frosturi,

When the Apertume time is 1 PLC (1/60Hz = 16.6 ms) and the delay is 5 ms, this results in a time of 21.2 ms. However the Aperture time can be lowered. The lowest Aperture time that you can set for the NI 4132 is 1/64 PLC (1 / 3850Hz = 260 ms).

By changing this property I was able to achieve results in the ~300 ms range.

[frosturi]

Hi Libby,

I changed the aperture time to 259.74us (3850Hz) and was able to get updates every ~ 5.3ms.  This is certainly better than the ~22ms that I had before.  However, it is not in the ~ 300us range that you were able to achieve.  I am assuming that the ms units you referenced were a typo and meant us.  In any case, I've attached a screen shot of the captured SMU output on the oscilloscope to demonstrate the 5.3ms updates.  I've also attached the VI that I am using.  It is taken from the NI-DCPower Hardware-Timed Voltage Sweep.vi.  I've made modifications such as setting Auto Zero to off.  Is there any other modificationt that you have done software or hardware-wise?

Thanks,

Jeff

[Libbin_la_vida_loca]

Is your source delay still set to 5 ms? Then that would give 5 ms + 300 us = 5.3 ms. Try setting your delay time to 0.

[frosturi]

Hi Libby,

You were right, I had my source delay setting at 5ms.  I reduced it to 0ms.  Now I am able to get the SMU to update much faster, and I am in the same range as you.  350us.  This puts my update rate at ~2850Hz, which is still below the specification's update rate of 4200 updates/sec.  Have you been able to acheive this performance?  Are there some other settings that must be adjusted to achieve this?

[menay]

Jeff,

The 4,200 samples/sec refers to how quickly the SMU can begin trying to reach the next setpoint, and is not actually an indicator of how quickly the output of the device will be able to reach the next setpoint. It does not take into consideration the settling time. If you take a look at page 4 of the PXI 4132 specification sheet, the maximum measure to host measurement speed at 1/64 PLC aperture time is 3490 Hz when operating at 60 Hz and 2900 Hz when operating at 50 Hz. I have included a link to the specification sheet for reference.

https://www.ni.com/docs/en-US/bundle/pxi-4132-specs/page/specs.html