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.

Does the KXCI command "DE CH" REALLY disable a SMU?

Post Reply
Der_Iltis
Posts: 11
Joined: November 9th, 2017, 2:09 am
Country: Germany

Does the KXCI command "DE CH" REALLY disable a SMU?

Post by Der_Iltis » October 5th, 2018, 8:28 am

Dear all,

While programming a 4200A for stress measurements, I noticed something interesting and probably dangerous: Disabling a channel has no effect on a SMU which standby mode was set!

I want to perform a stress measurement and thus keep, say, SMU1 active after the measurement. I can do so by using the following commands with the appropriate parameters (in particular, with those for SMU1)
  • "DE CH" (channel definition),
  • "SS VC" (output configuration),
  • "ST" (disable standby, used on SMU 1)
  • "MD ME" (perform measurement)
The measurement works as expected and the output of SMU1 remains also active.

Problems start when I want to disable (physically disconnect) SMU 1 after some time. The reference manual states that this can be done with using the "DE CH" command. However, after doing so the constant voltage on SMU1 remains! Even worse, if I now configure and perform a different measurement on different SMUs, not even having SMU1 in the channel definitions anymore, the constant voltage on SMU1 STILL remains! This is dangerous both for devices and humans.

Can you reproduce this issue?


I also played around a little bit to understand what is happening and found another issue: In a subsequent test, I explicitly enabled standby on SMU1 before performing the first measurement (stated above, involving SMU1 again). As expected, SMU1 now goes into standby mode after the measurement, which means it forces 0V. BUT if I disable the channel now, the output seems to keep actively forcing 0V. This was difficult to measure using a simple volt meter. I used the capability of my voltmeter to measure resistance, the result I got was "0 Ohms", but I expected a very high resistance, or the voltmeter status "Disconnected".

Can someone from Tek reproduce this? And would someone please comment on this behavior?

How can I be sure that “DE CH” physically disconnects a SMU after having used it?
Is it actually intended that I cannot physically disconnect an output, even if I only explicitly enabled standby mode, which is the default setting anyway?
Is the standby mode affecting this behavior at all, or does it also occur when standby mode is untouched?

Andrea C
Keithley Applications
Keithley Applications
Posts: 1244
Joined: October 15th, 2010, 10:35 am
Country: United States
Contact:

Re: Does the KXCI command "DE CH" REALLY disable a SMU?

Post by Andrea C » October 25th, 2018, 3:53 am

For SMU1, are you sending "DE CH1" or are you leaving off the "1"?
To reset multiple SMU: "DE;CH1;CH2;CH3;CH4"

This should be putting the SMU back into standby mode.
Standby is not the same as disconnecting the SMU. In fact, it is forcing 0 volts with a current limit influenced by the standby current range. The standby current range is selected in KCON.
4200_SMU_Standby.PNG
4200 SMU Standby Behavior
4200_SMU_Standby.PNG (35.91 KiB) Viewed 860 times
I'm not aware of any commands to open relays to literally disconnect the SMU. Instead, can you achieve a high impedance state by commanding SMU to force 0 Amps with a voltage compliance higher than any voltage that might be applied externally. For example, if connected to a battery, you do not want a compliance condition to occur.

Der_Iltis
Posts: 11
Joined: November 9th, 2017, 2:09 am
Country: Germany

[Update] Correspondance with Tektronix' Technical Support, and the reply by Andrea C

Post by Der_Iltis » October 26th, 2018, 1:36 am

UPDATE after correspondance with Tektronix' technical support and the reply by Andrea C:

I do not omit the channel number, I am sending "DE CHn", with n being the channel number. Indeed, the issue was reproduced at Tektronix, it probably is a bug or at least a flaw in the reference manual. The following snippets should give more details on this issue. A sample project is given at the end of this reply.

Here is the first response that I received from Tek:
(...) Here are some information you may know.
I tried this short sequence, and it works as it should and as you see.

Code: Select all

DE                                                                                -----------------> access to the DE page command
CH1, ‘V1’, ‘I1’, 1, 3                                                    -----------------> setup SMU1 : Voltage source mode and constant source function
SS                                                                                 -----------------> access to the SS page command
VC1, 10.00, 0.01                                                        -----------------> setup output configuration of SMU1: force 10V and current limit of 0.01A.
ST 1, 1                                                                          -----------------> setup auto standby mode to 1: SMU1 turns off at end of measurement.
MD                                                                               -----------------> access to the MD page command
ME1                                                                              -----------------> simple trigger: trigger a single measurement
At end of the test, SMU1 is turned off.
if ST 1, 1 command is replaced with ST 1, 0 -----------------------------> SMU1 stays on even if the test is completed.


About CHn command ( n = SMU channel number) : CHn command turns off capabilities to setup SMUn. When off, there is no way to setup output configuration or to trigger measurement. Chn command uses latest configuration ( default or entered by user).
It does not turn off the SMU output whatever the source mode is.


(image showing System Mode SMU Default Settings, can be found in reference manual)

To disconnect a SMU, we do recommend configuring it to Common (COM) -----------------> Output HI terminal of the SMU internally connected to common (COM) or ground unit Force (LO terminal of each card connected to a common reference)
Second, here is my reply, including the sample
Thanks for your immediate reply. (...) I can explain my setup:

We want to measure Voltage and Current over a resistor. I will do this twice, using different approaches:
First: SMU1 will force a DC bias voltage of 5V, while SMU2 sweeps linearly from 0V to 10V, such that the voltage across the resistor varies from -5V to +5V. SMU2 will NOT go into standby, but will be disabled after the test.
Second: Only SMU1 is set up. Consequently, it sweeps linearly from -5V to 5V. However, the voltage across the resistor will vary between -15V to –5V, as SMU2 is still actively forcing 10V.

Taking your example as a starting point, here is what I do:

Code: Select all

// Part 1:
DE                                          -----------------> access to the DE page command
CH1, ‘V1’, ‘I1’, 1, 3           -----------------> setup SMU1 : Voltage source mode and constant source function
CH2,  ‘V2’, ‘I2’, 1, 1          -----------------> setup SMU2 : Voltage source mode and VAR 1 source function
SS                                          -----------------> access to the SS page command
VC1, 5.00, 0.01                  -----------------> setup output configuration of SMU1: force 5V and current limit of 0.01A.
ST 1, 1                                   -----------------> setup auto standby mode to 1: SMU1 turns off at end of measurement (which ACTUALLY means, it is actively forcing 0V, right?)
VR2,0,10,0.5,0.01            -----------------> setup output configuration of SMU2:  VAR1 linear sweep from 0 to 10V in 0.5V steps, with I-compliance set to 0.01A.
ST 2, 0                                  -----------------> setup auto standby mode to 0: SMU2 stays on at end of measurement.
MD                                        -----------------> access to the MD page command
ME1                                      -----------------> simple trigger: trigger a single measurement
DE                                          -----------------> access to the DE page command
CH2                                       -----------------> “Disable” CH2 (page 763 in 4200A manual: “To disable a channel, send: CHx” – but as you say, the output stays on at 10 Volts. Isn’t the manual misleading here?)
// Part one is finished

// Part 2:
DE                                          -----------------> access to the DE page command
CH1,  ‘V2’, ‘I2’, 1, 1          -----------------> setup SMU2 : Voltage source mode and VAR 1 source function
SS                                          -----------------> access to the SS page command
VR1,-5,5,0.5,0.01             -----------------> setup output configuration of SMU1:  VAR1 linear sweep from -5 to +5V in 0.5V steps, with I-compliance set to 0.01A.
ST 1, 1                                   -----------------> setup auto standby mode to 1: SMU1 turns off at end of measurement (which ACTUALLY means, it is actively forcing 0V, right?)
MD                                        -----------------> access to the MD page command
ME1                                      -----------------> simple trigger: trigger a single measurement
DE                                          -----------------> access to the DE page command
CH1                                       -----------------> “Disable” CH1 (Although 0V are applied, it still does not physically disconnect the output waveform of the SMU from my DUT?)
// Part 2 is finished
Regarding your comment on configuring SMU1 to common: This would still not physically disconnect the output waveform of the SMU from my DUT, right? As a consequence, eventually occurring accumulation of charge would always directly flow to ground,? So is there a way to physically disconnect a SMU after it has been used? Isn’t that the initial setting after a reset?
Third, here are some notes I took about the call that I received as a response to my reply, includign Andrea C's comment.

The CHn command of the 4200 does indeed only remove the channel from the defintion. It does not change the output of the channel, even if the standby mode was disabled and thus, might differ from 0V. Both me and Tek's Application Engineer believe that this can be regarded as a bug.
The Application Engineer also stated that according to his opinion, the 4200 does not have a high impedance state such as, for example, the 26XX SMU series. In particular, all SMUs are already forcing 0V initially after the 4200 is turned on (system mode defaults). However, Andrea C's suggestion (forcing a constant current of 0A with a suffieciently high voltage compliance) sounds like a promising workaround to achieve a high imepdance state. Are there an differences between this "explicit HighZ state" and the "built in High-Z states" of the 26XX?

Post Reply

Return to “4200-SCS Semiconductor Parameter Analyzer”

Who is online

Users browsing this forum: No registered users and 1 guest