Andrea C

For others who might follow along:
2450 has USB, LAN and GPIB for remote command control.
Newer 4200A-SCS have VISA library available from KULT, so could use any of those command interfaces if we code at VISA  API.  Newer 4200A also have the GPIB interface API.

Older 4200-SCS do not have the VISA library.  They have GPIB API for external instruments accessible from KULT and UTMs.

This sample code is written using GPIB API to do a voltage sweep on a model 2450 from a UTM in KITE or Clarius.
The 2450 requires the TSP command set and connected on GPIB to the 4200x.

In the attached files:
the *.c file is contain in a zip so that this forum tool permits it.
The PDF has instructions for installing it.

Thank you for posting the screenshots.

So clearly there are some crest factor issues as you have a significant leading edge spike in the waveform.

In the digitizer mode, if you up range the voltage, what is the max voltage of the spike?  On the full screen graph, you can zoom in and place cursors too.

I suspect that when using the Freq or VAC modes and low ranges, the spike is saturating the range.

But higher ranges that tolerate the spike, don’t have sensitivity for the much lower amplitude remaining signal.

Not sure the origin of the spike.  Calm it down if you can (reduce inductance if the spike corresponds to current load changes).

With TSP and Digitize, you could extract the freq info from timestamps between two peaks for example.

Can your use case support idea of not scanning conventionally, but closing a channel and then calling a TSP function?

My question about using DCV was to get some confidence of the relay function and the wiring to it.

To make it simpler, do select front terminals and directly feed your signal to dmm.

For the digitize V, same place you select ACV or DCV, pick Digitize V.

Swipe screen to see the digitize settings for sample rate and number of samples.

For your ~20Hz signal, I’d say slow down the sampling rate and set buffer size.

1KHz and 1000 samples = 1 second of data

Swipe to graph.

What range do you have to use to reliably see the waveform?

What does waveform look like?  Hoping it gives some insight into why VAC is having trouble.

For freq measurements, what aperture setting did you use?  A 2msec aperture is not long enough to measure freq or period of a 20Hz signal.

Let’s summarize:  you have two older 4200-SCS/F running either WinXP or Win7.
The 4200 have some SMU and you use KITE software.

You now have some 2450, 2461 and 2601B and looking for software to operate them.
Is allocating a Win10 or newer computer not an option?
If yes, you could use ACS-Basic or KickStart.

But if constrained to using the 4200 as the computer for the new SMUs, you need to use WinXP vintage tools.

Does older 4200 have the Microsoft compiler installed and is KULT functional?
Perhaps walk through the KULT tutorial in the manual to verify it is functional.

If functional, it has GPIB read and write and you could connect to external SMUs.
Existing sample libraries on the 4200 for LCR meter would be an example of GPIB to external instrument.
(newer 4200A supports usb, lan and gpib to external instruments)

With KULT you can interact with internal 4200 related SMUs and external SMUs over GPIB.
Just depends on the use case you are trying to support.
 

if you put known DC on the channel, does DCV read it correctly?

For the AC signal, can you close the channel and use the Digitize V function on the signal?
You can slow the digitizer to 1KHz sample rate and auto aperture to get boxcar average from 1MHz digitizer.
What does that think you signal looks like?

Can you set break points and/or single step your code to identify which line causes an error?
And what is the error?
Queue Full means you have a lot of errors.
Try errorqueue.clear() at start of code.

If it worked well with older Python, then probably just a syntax issue/mistake in the conversion.
The SMU has no awareness of what is sending the commands.

Have you evaluated the 6517B response to the motion/vibration?  I suggest to remove any cable at triax connector and put metal cap on triax input to shield it.  Use foil if you do not have metal triax cap.  Run your test.  Is 6517B unaware of motion?


But pertaining to your question, in amps mode, the 6517B is essentially a dead short between HI and LO.  For static situation, if charge is stored on a capacitor and a voltage results, when you connect amps measure….big whoosh of charge, and voltage depleted.
I would not expect appreciable voltage potential to develop while amp meter is attached.  The voltage burden spec tells you the max expected value.


is the setup well shielded?

 

I disagree about the part availability.
As a repair part, item TF-245 is available component.  Contact your local Tektronix sales office.
It is a resistor network in SOIC-16.
 

In contrast to our SourceMeter products which can directly be current source or voltage source, a power supply is always a voltage source.

But even for voltage source, there is also the parameter setting for the current compliance limit.
A power supply can achieve current source (CC mode) behavior by deliberately operating it in current limit.

Suppose you have a device that is approximately 10Ω and you want to source 100mA through it.
So you’d expect a 1V potential to develop.
Configure the power supply for the 100mA limit and ask it to source more than 1V…..let’s say 10V.
It will not be able to achieve the 10V because it will first encounter the 100mA limit.
The power supply read back functions will show the appx 1V of potential and the 100mA of current.


 

Can you post a bit more information about your coding approach?
I am not fully understanding the situation.

Not sure this applies, but keep in mind that 2450 will process commands more quickly than 2400.
So a loop running on PC certainly can blast through more quickly.