Limitations of Cascading two PXie-4163 SMUs in Parallel for Higher Current

[ksalve]

Hello,

I have cascaded two 4163 in parallel for higher current.

My reference for this are the NI articles:

Cascading the Outputs of the NI PXIe-4140/4141/4142/4143/4144/4145 - NI

Connecting Multiple NI Source Measure Unit Channels in Parallel - NI

My circuit looks like this:

Where the Lo-sense are near the DUT, and LO-force are at AGND.

Aside from the limitation of voltage accuracies, is there any limitation on the current and current range that I should be aware of?

Or any limitation in general.

Thank you!

Ken

[santo_13]

 1+1 is not 2 in this case of parallel connection due to circulating current between channels due to differences in voltages.

Do you occasionally need higher than 50mA or always need more than 50mA for your DUT supply?

[ksalve]

Hi Santhosh,

Thank you for the reply!

Yes, I occasionally need higher than 50mA for my DUT, 100mA in fact, thus I cascaded these two.

The mode where I normally use the two SMU is when I force current (100mA) for a specific test.

Any other test, when I force voltage, I disconnect the second SMU using the output disconnect and output disabled function since I don't need (expect the 100mA).

I asked about the limitations because I notice when I'm forcing voltage on the cascaded SMU's above a certain level (~16.5V), the current is rising (around uA). While on a normal 4162, on the same type of pin, this behavior is non-existent (measuring nA or even pA).

I cannot rule out that this is my setup, but since there isn't enough documentation, I also can't say if it's because of the cascaded SMU's situation that the current rises when a certain forced voltage is reached.

Let me know if I wasn't clear enough, maybe I can provide more details.

Thanks!

Ken

[santo_13]

If you're using both channels in force current mode, then, you will be able to achieve 100mA.

How are you merging in software? are you using the merged channel feature of the driver?

Important information: The output disconnected feature uses solid-state relays; they are not electrically isolated. This means that when one channel is output disconnected and the other channel is sourcing, a small leakage current flows into the disconnected channel. For true disconnection, use an electromechanical relay on the loadboard.

Please read the merging topic here -https://www.ni.com/docs/en-US/bundle/pxie-4163/page/merged-channels.html

[ksalve]

Hi Santhosh,

I have read up on the merging topic you provided.

I have not done this part on my program.

Thank you for this. I will try it and see this might be one of the reason.

I also was not aware that the relay was solid state.

I will take note of it.

I'll get back to this thread again once I confirm these suggestions.

Thank you!

Regards,

Ken