Acquiring data faster

[lucasswill]

Hi, I am changing a code to operate a  Keithley 2401 sourcemeter. I want to be able to change the voltage applied on a solar cell and record the corresponding current in different bias modes, IV curves. Also, I want to repeat measurements in defined amount of time. Finally, all data should be saved in text files automatically. I have changed the original code and I achieved what I wanted my problem is that it takes a while to acquire data. It takes 30 min to run 100 IV curves with 50 points in the bi directional mode. Is there any way to make it faster? I am very knew to labview so my code is probably full of mistakes, all suggestions are very welcomed. I am using Labview 2014.

Thanks.

[James.M]

Is there a hardware limitation to your speed? Can the Keithley keep up? You have Wait nodes in the VI which could slow down the timing, but it sounds like your instrument communication is the limiting factor because those waits aren't long enough to cause it to take 30 minutes with 5000 data points.

Is there a sample rate setting that you need to configure on your Keithley? It looks like you're reading a certain number of points and waiting until all of those points come in before moving on. If you Keithley is only giving you those points at a slow rate, then that would limit how fast you can run this applcation.

(This is the most likely fix to your slow rate, find the sample rate setting and increase it. Maybe this has something to do with the "Step Delay" input, which is 2 seconds?)

I don't know much about the VIs your using. Do you need to "Configure", "Enable", and "Read" multiple times each loop, every single loop? The only thing that looks to be changing with each iteration is the Index that you write to the "Save" VI.

[Bob_Schor]

How long does it take your instrument to do an I-V curve?  I notice that you are using a Baud Rate that was considered "fast" about three decades ago.  If you use the highest Baud Rate that your Instrument will allow, that will lessen "communication delay" as part of the problem.

What do you do inside the Acquisition Loop?  A loop runs no faster than its slowest component, and if components are serialized by data flow, the time will be the sum of the times of all of the components.  So if most of the acquisition is "fast", but the processing is "slow", your loop time, "fast" + "slow", will be "slow".  You can speed things up by using LabVIEW's inherent parallelism.  Look up the Producer-Consumer Design Pattern by opening LabVIEW, click File, New ... (not New VI), expand From Template, Design Patterns, Producer/Consumer Design Pattern (Data).

Bob Schor

P.S. (wearing my Pedantic Hat) -- if you "knew" LabVIEW, you wouldn't be "new" to LabVIEW (see your final sentence).

[James.M]

---

@Bob_Schor wrote:

You can speed things up by using LabVIEW's inherent parallelism.  

---

Bob is right about the parallelism, but be careful with this when it comes to communication with a single instrument. You can run in to errors when you try to send commands at the exact same time to the same instrument. In this case, you probably do need to keep the communication in series, so hopefully you can speed up the acquisition itself by increasing the rate of the instrument's actions.