Auto range / Voltage limit

[appleSwell]

Hello,

I've been using an example LabVIEW program such as Sweep and Acquirement to attain a simple I-V measurement. I am aware that 4-wire setup has a hard-limit on voltage of 21V without an interlock, and I only try to sweep up to 20V. However, somehow the voltage gets capped at the level less than 20V--somewhere 16V to 17V depending on a sample (I source voltage and measure current).

I instead tried the manual setting on the touchscreen. I changed the range from auto to 200V, and when I source 20V, the sourcemeter correctly puts 20V to the sample and displays a reasonable amount of current reading. Therefore, I assume the premature voltage cap that happens on LabVIEW implementation comes from the voltage range set as auto. However, I am still very confused as to why this is so. Why does it get capped below the hard voltage limit when the range is in auto mode? And, what kind of modification do I need in LabVIEW program to avoid this issue?

[cstorey]

Before anyone can help you need to specify which make and model of DMM/SMU/Voltmeter/Electrometer/Microohmmeter you are using.   You should probably also attach the code to which you refer.  People who answer questions here are very knowledgeable, but not necessarily psychic.  Detailed descriptions and questions get the best answers.

If I had to guess, I suspect the answer lies in the default settings of your meter.  If you use Reset=True when initializing communications with the instruments, then you are accepting some set of default configurations - which the manual will explain in detail.  There maybe a power compliance limit for the default 20V range, which you hit at 16-17V in your measurement.  Going to the next scale might set that power limit higher.  If using an SMU your measurement might be in a range compliance.  All just guesses without spedific details on the instrument.

BTW, what are you measuring?  Why not source current and measure voltage?  In nearly all applications except high resistance measurements it is more accurate to source current and measure a voltage. 

[appleSwell]

I apologize for not having made a full explanation. I am using Keithley 2450 Sourcemeter to attain an I-V characteristic of a thin insulating film grown on a hard substrate. I attached the LabVIEW block diagram. The program is based off of an example code "output list and acquire.vi." I want the program to perform a hysteresis-type measurement, in which the source bias goes up to positive, comes down to negative, and finally back to zero.

 

After reading your reply, I specified the source voltage range to be 200V, and the voltage goes up to 20V. However, I do not understand how the power limit works. From the power envelope of the 2450 sourcemeter datasheet, for 20V range, current can reach up to 1A, whereas for 200V range, current can reach only up to 200V. I also attached the screenshot of the IV graph to demonstrate my confusion. This is when I let voltage range to be auto. At 16V, the current is only at 1mA, far from the current limit.

 

In addition, take a look at the behavior from 0 to 1V range. Why is there some kind of a jump? And this "jump" is also observed at larger voltage. I attached an additional screenshot of I-V curve when I set the voltage range as 200V.

 

I'm currently confused by a lot of things that are happening with my I-V characteristic. When I do a sweep in a small voltage range, in the order of 0.1V, it results in a clean linear relationship. I want to think that this weird "jumping" behavior is intrinsic to the sample that I'm measuring, not from a faulty probe setup/program issue.

Thank you so much for reading through my long reply. 

[appleSwell]

I apologize for not having made a full explanation. I am using Keithley 2450 Sourcemeter to attain an I-V characteristic of a thin insulating film grown on a hard substrate. I attached the LabVIEW block diagram. The program is based off of an example code "output list and acquire.vi." I want the program to perform a hysteresis-type measurement, in which the source bias goes up to positive, comes down to negative, and finally back to zero.

 

After reading your reply, I specified the source voltage range to be 200V, and the voltage goes up to 20V. However, I do not understand how the power limit works. From the power envelope of the 2450 sourcemeter datasheet, for 20V range, current can reach up to 1A, whereas for 200V range, current can reach only up to 200V. I also attached the screenshot of the IV graph to demonstrate my confusion. This is when I let voltage range to be auto. At 16V, the current is only at 1mA, far from the current limit.

 

In addition, take a look at the behavior from 0 to 1V range. Why is there some kind of a jump? And this "jump" is also observed at larger voltage. I attached an additional screenshot of I-V curve when I set the voltage range as 200V.

 

I'm currently confused by a lot of things that are happening with my I-V characteristic. When I do a sweep in a small voltage range, in the order of 0.1V, it results in a clean linear relationship. I want to think that this weird "jumping" behavior is intrinsic to the sample that I'm measuring, not from a faulty probe setup/program issue.

Thank you so much for reading through my long reply. 

[appleSwell]

I apologize for not having made a full explanation. I am using Keithley 2450 Sourcemeter to attain an I-V characteristic of a thin insulating film grown on a hard substrate. I attached the LabVIEW block diagram. The program is based off of an example code "output list and acquire.vi." I want the program to perform a hysteresis-type measurement, in which the source bias goes up to positive, comes down to negative, and finally back to zero.

 

After reading your reply, I specified the source voltage range to be 200V, and the voltage goes up to 20V. However, I do not understand how the power limit works. From the power envelope of the 2450 sourcemeter datasheet, for 20V range, current can reach up to 1A, whereas for 200V range, current can reach only up to 200V. I also attached the screenshot of the IV graph to demonstrate my confusion. This is when I let voltage range to be auto. At 16V, the current is only at 1mA, far from the current limit.

 

In addition, take a look at the behavior from 0 to 1V range. Why is there some kind of a jump? And this "jump" is also observed at larger voltage. I attached an additional screenshot of I-V curve when I set the voltage range as 200V.

 

I'm currently confused by a lot of things that are happening with my I-V characteristic. When I do a sweep in a small voltage range, in the order of 0.1V, it results in a clean linear relationship. I want to think that this weird "jumping" behavior is intrinsic to the sample that I'm measuring, not from a faulty probe setup/program issue.

Thank you so much for reading through my long reply. 

[cstorey]

It would be better if you posted your actual code, rather than a picture of it.  It shows the case for sweeping current..bt not the settings for a voltage sweep.  At the least post a VI snippet with the values you are using saved as the default settings.  

---

@appleSwell

In addition, take a look at the behavior from 0 to 1V range. Why is there some kind of a jump? And this "jump" is also observed at larger voltage. I attached an additional screenshot of I-V curve when I set the voltage

I'm currently confused by a lot of things that are happening with my I-V characteristic. When I do a sweep in a small voltage range, in the order of 0.1V, it results in a clean linear relationship. I want to think that this weird "jumping" behavior is intrinsic to the sample that I'm measuring, not from a faulty probe setup/program issue.

As for the measurement shown, a "thin insulating film" is essentially a capacitor.  Ideal capacitors have no leakage, so if you swept voltage from 0-200V you would measure no current.  Real capacitors have leakage paths**. You represent these parasitics as a capacitor in series or parallel with some resistors and inductors (usually a C||R + Rseries + Lseries).   From 0 to 1V you are essentially charging the a capacitor while very little current flows and then you reach the voltage where the film becomes a leaky and current flows through the resistor in parrallel and series.  What were you expecting to see?  What are you really trying to measure!  

** Leakage has all sorts of flavours - defects in the films, film materials characteristics, parallell paths in the device due to poor design or processing issues,  irreversible damage above some threshold, etc..  

Your sweeping out from 0V and back is the best way to look at things.  I suggest sweeping current and measuring voltage.  Go from -1mA to +1mA, use the same out and back strategy as before.   I also suggest starting very low current (+-10uA, +-100uA, etc..) and saving the measurements as you go.  Compare them to see if you are observing damage or breakdown of the dielectric.

[Minions]

cstorey is correct.  Additionally, you should brush up on what typical V-I curve traces are for different components.  You would get a better understanding of charging/discharging causing hysteresis loops in your traces and why.