Keithley 2400 Repair - Large Source Output Impedance

[jl02]

Hello everyone, I've recently repaired two separate Keithley 2400 SMUs, and I'm seeing some strange issues while performing my verification testing. As a preface, one unit needed repairs to the analog output board, but the other only needed a front panel repair. The second unit, of which I did not touch the analog board, was calibrated August of 2023, so it should still be within calibration. I have tested both 2400s with their self test programs and a 4-wire short, and both passed with no issues.

On both the units I've repaired, there seems to be a (relatively) large output impedance of roughly 0.27 Ohms, as observed by measuring the effects of loading the units to large current ranges. I've been using an Itech IT8512A+ to verify the maximum power output, and measuring the effect of loading using a Keithley 2000. By measuring the terminal of either 2400 in voltage source mode with the E-load disabled/disconnected, the voltage the 2400 displays is spot on. However, with a 1A load, the voltage measured at their output terminals drops by about 0.27 V (therefore has an output impedance of 0.27 Ohms), yet they still display near the programmed voltage. See attached images for demonstration.

I've contacted Tektronix regarding this issue, and they state that the output impedance should be magnitudes lower, and just claim that the units need calibration.

This impedance is observed while using a lower current range and using a purely resistive load, and furthermore the issue is solved by switching to a 4 wire measurement. I've additionally verified that the impedance of my cabling is not the source of the problem.

Is anyone able to replicate this with a functioning unit? I want to double check that it's actually a calibration issue before I drop the cash to calibrate them   

[BennoG]

I have also the 0.3V drop at 1A load on my 2400.

Also noticed a bug in the firmware, if you turn it on with no load attached and set it to 20V 1A. Then turn it to Voltage measurement (as in your photo).
The compliance part starts blinking even when you load it with 100mA and the voltage drops to 0v.
Only when you turn it it measure current it corrects and can deliver the 1A.

This happens also when you have it turned on unloaded in current measurement and switch to voltage measurement and start loading it.

Benno

[TheDefpom]

Maybe try swapping the stack order of the cables when plugging into the SMU, right now you have the load plugged in first, then the DMM, swap them around so the DMM is plugged in first, this will at least eliminate the plugs on the leads you are using to the load, if they are not performing well it could be a drop within the plugs themselves.

[voltsandjolts]

What happens if you connect 2400 sense lines to your load?

[BennoG]

For me the 0.3V drop despairs to less than 1mV. (after configuring it for actual 4 wire sourcing)

Benno

[jl02]

Maybe try swapping the stack order of the cables when plugging into the SMU, right now you have the load plugged in first, then the DMM, swap them around so the DMM is plugged in first, this will at least eliminate the plugs on the leads you are using to the load, if they are not performing well it could be a drop within the plugs themselves.

Swapping the order of the banana jacks made no difference.

What happens if you connect 2400 sense lines to your load?

Just connecting the sense lines without enabling 4 wire measurement results with the same issue. Enabling the 4-wire measurement fixes the problem.

[voltsandjolts]

Congrats on the repairs, seems you have 100% functional units.