PXIe-4138 pulse width is consistently 80ms ~ 120ms longer than configured

[Wblin531]

I'm using an NI PXIe-4138 module in a PXIe-1073 chassis, trying to generate a 50 µs current pulse. I'm measuring with a 1X passive probe connected directly across the SMU output terminals, with no load resistor in between. The scope is Tektronix DPO3014.

The pulse I get is always about  80ms ~ 120 ms longer than configured, and the extra amount is constant — it doesn't scale with Pulse On Time:

    Pulse On Time 50 µs → measured ~120 ms
    Pulse On Time 200 ms → measured ~320 ms

I checked the manual and saw a note recommending Sequence mode, but I've tried both Single Point and Sequence and get the same result. My VI is in attached — I'd appreciate anyone pointing out what I'm doing wrong.

Here is my Parameter settings:
Output Function = Pulse Current;
Pulse Current Level = 100 mA; Pulse Bias Current Level = 100 mA; Pulse Voltage Limit = 2 V;
Pulse On Time = 50 µs; Pulse Off Time = 300 µs;
Source Delay = 0 s; Aperture Time = 40 µs

LabVIEW environment: 2016

I've been stuck here for 3 days, and I'd appreciate it if someone could help me out!

[Henrik_Volkers]

A short lookup in the specs tell me that the settle time  for current is <50µs 

and depends on configuration and load. 

Never used such a device , but it seems to me that your 50µs pulse length will be a challenge for that device.

And you should try it with a 10-20 Ohm resistor as a load 😉

[Wblin531]

Thanks for the suggestion. Using a resistor as a load was a good idea.

I have several resistors available (around 500 Ω to 1 kΩ), so I tried adding them to the output. I did notice some improvement. The pulse width error was reduced from the original 80-120 ms down to around 50-60 ms.

However, the result is still far from what I am expecting. My ultimate goal is to generate a 50 µs pulse, but at this point I'm seeing a significant error even when the Pulse On Time is set to 200 ms.

This makes me think there may be something else in my configuration or measurement setup that I'm overlooking.

=====================================================================

I'll continue experimenting with different load values and configurations.

I also noticed that many measurement examples and test setups include a load resistor connected to the output. As someone who is still relatively new to this area, I'd like to understand why this is commonly done.

Is it mainly for stability and settling performance, or are there other important reasons?

Any insights would be greatly appreciated.

[Henrik_Volkers · Accepted solution]

Never used such a device, but from the spec, there are different speed configurations for this unit you can (have to) set.  RTFM should help.

Again 50µs current pulse seems hardly if not possible with this unit.

You haven't mentioned your DUT / Load you want to apply.

Another way to realise a current pulse could be to use a transistor to shortcut the current (100 mA in your case) in parallel to your DUT and use a digital output from another card to open the transistor for 50 µs, so the current choose the way across your DUT.

However you maybe have to look at additional impedances (cables, your DUT.. etc) that will have an influence due to the dI/dt.