PXIe 4138 current mode in different ranges

[mexaviesta]

Hallo everyone,

i've been using a SMU (PXIe 4138) to output a constant current for two inputs (I0 and I1) of a analog module, and simultaneously i'm using a DMM (PXI 4065) to measure the current flows the inputs (see the image attached below).

But i noticed that the measurement from the DMM depending on the current level limits selected in SMU, (please see the table below), i was wondering why the measured current increased with the increase of the current limits in SMU, and why the current measured even larger than outputed? Any idea what's causing that? Thx a lot.

Current setting and output  in SMU	Current level limits in SMU	DMM measurement
4 mA	10 mA	3.987 mA
4 mA	1 A	4.007 mA
4 mA	3 A	4.120 mA

[santo_13]

That is called accuracy specifications. 

Each current range has an accuracy specification as specified in the datasheet of 4138,

You can calculate the accuracy of 4mA sourced in each range by using the formula given in the accuracy column of the table

[mexaviesta]

Thanks a lot for your replay. I've had calculated the accuracy in each range accoding to the accuracy specification, and the measurement with 10mA range should in principle have the best accuracy. However, as we can see in the following table, the measurement in 1A range actually has the best result (0.007mA difference), and i came to the same conclusion in another experiment. This got me confused. 

Current setting and output  in SMU	Current level limits in SMU	DMM measurement	Diff.
4 mA	10 mA	3.987 mA	- 0.013 mA
4 mA	1 A	4.007 mA	0.007 mA
4 mA	3 A	4.120 mA	0.120 mA

[santo_13]

First, you should not compare between ranges to get the best accuracy as each range would have a different calibration factor and accuracy specifications.

Next, if you're using DMM to verify the source accuracy, your DMM must have better accuracy than the SMU you're trying to verify.

What is the accuracy of your DMM as per the range?

As long as the 4138 has a proper valid calibration, it will definitely operate within the specifications in the datasheet.

[mexaviesta]

"First, you should not compare between ranges to get the best accuracy as each range would have a different calibration factor and accuracy specifications."

According to the datasheet i think the range 10mA should be the best option to output a 4mA current, is it correct? Or should i consider anything else? Is there any possibility that the accuracy of the SMU output depending on the load in the circuit? We assume that the SMU with a low range may not have enough "power" to generate a accurate output, when the circuit has a high load in it. Is it possible?

[santo_13]

You'll get the best accuracy (worst case +/- %) in the closest range but it does not mean that other ranges might not have better realistic accuracy.

No, accuracy should not depend on the load as you're sourcing CC as long as the load is stable, your accuracy will be within specifications.

Now, in reality, it is definitely "possible" that the 10mA range may be performing at close to the worst +/-% accuracy (but still within datasheet spec) whereas some other higher range might be operating in an ideal +/-0% accuracy.

Each range will have a different actual +/- accuracy percentage and need not be exactly the same. With this being said, in your case, for a 4mA source 1A range is operating in a more ideal lesser +/- accuracy whereas the 10mA range is quite far from ideal compared to the 10mA range.

For example, the realistic accuracy measurements are as below,
10mA range -0.01%

1A range +0.005%

3A range -0.008%

All the above range accuracy is valid since they are within the 0.03% accuracy specified in the datasheet but in terms of actual accuracy, 1A range is the best.

All this has to do with how each range was calibrated, temperature drift, self-calibration state, circuit noise, and range-related noises. But in any case, the instrument datasheet will be valid if the instrument is calibrated properly

[mexaviesta]

thanks a lot for your explanation, it makes sense. But when we in our case unter the same condition measured the 4mA source of SMU directly by using the DMM, without connecting to our module(load), we could verify that the measurement in range 10mA was much more accurate  than 1 A, that's also why we suspect that the accuracy could possibly depend on the load. How can we explain that?

[santo_13]

Could you please elaborate on your device (share datasheet and model) and how it is actually connected? and other 4138 settings such as Voltage clamp, Transient response etc. Maybe your load is causing instability or oscillations to the SMU?

[mexaviesta]

Thanks, the device is an analog module with 2 analog inputs, as shown in the screenshot below: 

What we wanna do is to adjust the input current, and the adjustment needs a constant current for the inputs. So we used a SMU (pxie 4138) as currentsource for both inputs, and the wiring is as shown in following:

Actually i haven't used any other specific Settings for 4138 such as Voltage clamp, Transient response etc.