Tektronix Technical Forums are maintained by community involvement. Feel free to post questions or respond to questions by other members. Should you require a time-sensitive answer, please contact your local Tektronix support center here.

Step-like behaviour of 6221 DC sweep for high ranges

6220, 6221 Current Sources
Post Reply
vskhromov
Posts: 2
Joined: January 11th, 2019, 7:48 am
Country: Russian Federation

Step-like behaviour of 6221 DC sweep for high ranges

Post by vskhromov » January 11th, 2019, 8:03 am

Dear colleagues,

I use 6221 to perform DC current (I) sweep through 2kOhm resistor. Voltage drop (U) across
resistor is measured with 2182a nanovoltmeter. Since it is expected to get linear (I,U) curve,
to control the accuracy of the experiment I consider quantity "U - Ufit_line", where "Ufit_line"
is linear approximation of experimental data.

For ranges 2e-9, 2e-8, 2e-7 of 6221 it works great and I get randomly placed points near
zero (Fig.1a) as expected. But for higher ranges these points are no more random, but
shaped like a ramp, for both small (Fig.1b) and large (Fig.1c) current steps.
Similar result (Fig.1d) is observed using data from topic on this forum: viewtopic.php?t=1721

It means that for range 2e-6 and higher experimental (I,U) curve is not linear, but splits into
series of steps that cross straight line (I,"Ufit_line").

My question is how to avoid this step-like behaviour, or where does it come from.

Regards, Vyacheslav.

P.S. My 6221 setup is:
Inner Shield = Guard;
Output low = Floating;
Current filter ON;
Output response SLOW;
Attachments
Graph1.jpg
Graph1.jpg (550.33 KiB) Viewed 2118 times

Andrea C
Keithley Applications
Keithley Applications
Posts: 1280
Joined: October 15th, 2010, 10:35 am
Country: United States
Contact:

Re: Step-like behaviour of 6221 DC sweep for high ranges

Post by Andrea C » January 16th, 2019, 9:09 am

So you are using fixed source range based upon the max value that will be sourced?
If you have appx 41 steps in a sweep between 0 and some full scale value, then for the higher source ranges and the values closer to zero, you will of course have more sourcing error relative to allowing for change of source range used for these smaller values of current.

The pattern in your error vs. source value plots may be partly a consequence of how the fit is performed and the attempt to "drive down the middle" of the samples.

vskhromov
Posts: 2
Joined: January 11th, 2019, 7:48 am
Country: Russian Federation

Re: Step-like behaviour of 6221 DC sweep for high ranges

Post by vskhromov » January 17th, 2019, 3:27 pm

Hello Andrea C, and thank you for reply!
I get the point, that for higher ranges it is more sourcing error. I did some research and got
interesting result, that for range 2e-6 current step value 47nA gives better results than other
values: quantity "U-Ufit_line" has almost two times less interquartile range (IQR) and, more
importantly, has randomly placed points around zero, not repeating ramp-like shape "two points
up, one down" in case of current step value 50nA (see Fig.2.).
The question is how to determine this optimal current step values for every range. I suppose
that current accuracy (from section A of the 6221 manual) and relation between resistance of sample
and 6221 internal resistance for each range play a key role here. I haven't found values of latter in
the manual, and if you will provide them, it would be really helpful.

Regards, Vyacheslav.
Attachments
Fig.2_.jpg
Fig.2_.jpg (110.17 KiB) Viewed 2090 times

Andrea C
Keithley Applications
Keithley Applications
Posts: 1280
Joined: October 15th, 2010, 10:35 am
Country: United States
Contact:

Re: Step-like behaviour of 6221 DC sweep for high ranges

Post by Andrea C » January 20th, 2019, 1:25 pm

Interesting observations.

Even so, the type of information you request is not something we have. Surely the 2182A could play a role, or even the DUT itself. Humidity and temperature will cause systemic shift too.

You are right to consider the sourcing accuracy specs for the 6221 and consider how that will impact the data.

Post Reply

Return to “Current Sources”

Who is online

Users browsing this forum: No registered users and 2 guests