LabVIEW

cancel
Showing results for 
Search instead for 
Did you mean: 

Keithley 24XX single read timing

Solved!
Go to solution

Hello,

I am very new to Labview and am newly working with a Keithley 2400 source-meter in hopes of recording single resistance measurements of a product. I was able to use the Keithley 24XX Read Single example provided as a basis for creating this program. I'm curious specifically now in the timing of acquiring a single reading (as I want a resistance value after charging the very small capacitor involved where the source-meter is connected as well).

 

When and how is a single measurement acquired and recorded in the Keithley 24XX Read Single Labview example program?

Is it recorded for a period of time or number of counts after voltage is sourced? 

Are a certain number of measurements taken and filtered/averaged into a single reading? Or is it over a specific time period (for instance 1 second) that measurements are averaged? 

 

From the Keithley 2400 manual (attached), I've deduced that speed (specifically the number of power line cycles (NPLC)) and perhaps filters (specifically using a repeating filter with a designated filter count) could affect the timing  and value of the measurement. In Appendix A: Specifications  it then lists "Single Reading Operation Reading Rates (rdg./second)" and looking at the Source-Measure to GPIB column specifies for different speeds the number of reading per second that can be taken for each speed.   I'm just unsure what conclusions I can make from these parameters and how the Labview program uses them to determine what a single acquired measurement value is.

 

0 Kudos
Message 1 of 7
(5,100 Views)

You need to see what command is being sent by the VI you are using.

Then read the programmer manual to see what that command does.

 

 

0 Kudos
Message 2 of 7
(5,069 Views)

My understanding (from memory) is that the example uses 1 NPLC (power line cycles) to make each resistance measurement.  It also defaults to AutoZero ON so, two measurements are taken (total 1/30th of a second on a 60hz grid)


"Should be" isn't "Is" -Jay
0 Kudos
Message 3 of 7
(5,060 Views)

Just running Keithley 24XX Read Single.vi at 20 and connecting the output to an oscilloscope, I saw an approximately 200ms pulse (see Source Meter Peak Pulse attachment)

 When are these two measurements taken during this pulse?

And then which measurement or mathematical combination is displayed in the front panel?

In which sub VI is the timing of them controlled in?

As my display is impacted by capacitance, timing is essential. (see attached waveforms when display driven by Read Single.vi)

0 Kudos
Message 4 of 7
(5,024 Views)

Before getting into the specifics of how the meter works, and there are lots of setting you can tweak, lets start with a few basic questions.

 

1) What is it you really want to do with it? 

2) Why is the timing of the measurement important?  Are you trying to trigger or synchronize with another instrument? 

3) Do you want to measure blazingly fast or with high accuracy? 

 

If you can answer those I can steer you toward answers you'll like! 😉 

 

SMUs are versatile but have limitations and you must understand those before writing a lot of code.  Your code should essentially leave timing up to the hardware, if timing is important, and should just mind its own business until the measurement is over and there's data to read back.

0 Kudos
Message 5 of 7
(5,016 Views)

1) I want a resistance measurement taken of a display once the internal capacitance has been charged and the desired voltage is seen across the display. In addition, I want to ideally know what current is drawn to drive the display once the internal capacitance has been charged as well. (basically I want the measurement taken sometime between the cursors pictured in Display peak pulse.png and not before when there isn't actually 20V seen across the display).

2) The specifics of when the meter takes the measurement is important because based on the size of the display the charge time of the capacitance varies and I want to ensure I'll be getting the appropriate current value (as I'm expecting a fairly low current draw even when the capacitance is charging I can't determine when the measurement is being taken) I'm not trying to trigger or synchronize with another instrument.

3) I don't want to measure blazingly fast or which high accuracy.

 

My initial inquiry was simply trying to understand when and how a single measurement is acquired and recorded in the Keithley 24XX Read Single.vi

Basically trying to understand when along the pulse a measurement is taken or if its an average of measurements.

And if there are any changeable parameters and if so which specifically within the VI or in a Sub VI that can alter when/how that occurs.

 

Hopefully that makes sense. I'm trying to gain an understanding of the data I'm getting.

 

0 Kudos
Message 6 of 7
(5,004 Views)
Solution
Accepted by topic author kOneill215

Ok, now I understand what you want to do and why your asking about timing. The explanation and answer aren't simply its X ms unfortunately, and I didn't want to go into detail if you didn't need them all.

 

The Keithley 2400 uses a source, delay, measure procedure.  First it sources current/voltage, then applies the desired settling delay (0-10k seconds) and then measures 1 time, averaging for NPLC (number of power line cycles from 0.01 - 16, at 60Hz that min of 167us).  It can repeats delay/measure N times if you want multiple measurements at the same bias point.  See all the gory details of the SDM timing on p152, and remote trigger SDM timing on p264 of the manual - https://mm.ece.ubc.ca/mediawiki/images/2/26/K2400Manual.pdf  

 

 

Now, as to the "Keithley 24XX Read Single.vi", I don't think you'll find any information about the timing of when the LabVIEW command is sent until the measurement is actually executed. There are just too many variables that affect that; Windows clock,  GPIB board, GPIB/USB cable length, GPIB latency, instrument latency, etc..  If you asked Keithley they would probably say it happens within ~1s.   BUS triggered measurements aren't accurate!  

 

If I'm really concerned about timing a measurement to occur before/after/during something else I use the trigger.  The hardware is faster, more accurate and more repeatable than software.

 

So,I think you have 2 options to get good measurements:

 

1 - Use triggered measurements.  A Schmitt trigger (http://hyperphysics.phy-astr.gsu.edu/hbase/Electronic/schmitt.html#c1) could be used to create a trigger pulse when voltage >20V across the display. The K2400 measurement takes place within a few us of the trigger ( and you could play with trigger delay, NPLC and auto-delay until you get the delay you desire.

 

2 - Essentially reproduce your scope trace with the Keithley 2400.  Get it to make N measurements at equal intervals using its internal timer.  Start the measurements when the charging starts and watch it go for what ever time you need.  The limit is 2500pts to the internal buffer, and just tailor the measurement interval and delay before measurements start as desired to get the best data possible.   Once you have acquired a "trace", use software to then decide at what conditions you extract your resistance measurement.  Is the Keithley sourcing the current to charge the display?  You can measure I, V and R at the same time so all that data is available to use.

 

 

 

 

Hope that helps.  Feel tree to ask more.

Message 7 of 7
(4,992 Views)