PXI-2570 switch capacitance

[Murphyky]

I am generating a 5V peak, 400 kHz signal from an ARB (NI 5401) and inputting it into an amplfier with a gain of about 2, current draw around 400mA and an input voltage of 13V. I am taking the output of that amplifier and going straight into the PXI-2570 relay switches. From NO-1 and NO-2 I run the cable directly to DUT and get a ~2.5V peak-peak signal in which i should be getting a 10V peak-peak signal. If I take the output of the amplifier and without putting it through the relays I run it directly to the DUT i get the 10V peak-peak.

My Question is why when i generate the 5V peak, 400 kHz signal and run it through the PXI-2570 and attach to the DUT do I see the large decrease in peak to peak voltage? Is there some internal capacitance of the relays that would be causing such a drop? The spec for PXI-2570 seems to be able to handle a 400 kHz signal?

[JaredRo]

Murphyky,

The signal you described is within the spec for the switch module, so it shouldn't be attenuated that much.  I have some questions that may help us to diagnose the problem.

Do you know the input impedance of your DUT?

How are you measuring the peak-to-peak voltage at the DUT (10 Vpp vs. 2.5 Vpp)?

Have you tried using different switches to pass the signal through?

You mentioned that you're routing your signal from NO-1 and NO-2 to your DUT.  Can you describe all of the connections you have made to the switch module so we can verify that it is set up correctly?

We can continue troubleshooting once I have this information.

[Murphyky]

Hi Jared,

1. There is an input impedance to the DUT of 23.066 ohms.

2. I have been measuring the peak-to-peak voltage at the DUT with a bench top oscilloscope.

3. Yes i have tried using different switches. My first attempt i miss typed, but i am usually using Relays 2 and 3 (not 1 and 2 like i first mentioned). I have also tried using relays 0 and 1.

4. So i have the ARB center conductor connecting directly to COM-0 and the sheild connecting to COM-1. From there i have a wire connecting from NC-0 connecting to COM-2 and NC-1 connecting to COM-3. From there I have the signal connecting to the DUT. I have on relays 4 and 5 a +/- 13V signal (NO-4 and NO-5). Lastly when I input the signal I also have relays 8 and 9 used for the 10 ohm load (NO-8 and NO-9).

Please let me know if you need any additional information.

Thanks.

[JaredRo]

Murphyky,

I'd like to see if we can simplify your setup a bit to pinpoint the source of the problem.  Connect the Arb to Com0 (+) and Com1 (-) and connect your Scope to NC0 (+) and NC1 (-).  Click the "Reset Device" button in Measurement & Automation Explorer (MAX) to ensure all of the Com ports are set to the NC (normally closed) terminals.  At this point, verify that the Scope is reading in the 5 V 400 kHz signal.  

If the signal isn't attenuated at this point, add in the amplifier stage and test the signal at the output.  Hopefully these troubleshooting steps will make it apparent where the attenuation in the signal is coming from.

What is the nature of the signal (sine, square, etc.)?

Switching out any cables you have is another good troubleshooting step.  A damaged cable can also introduced unwanted attenuation to your signal.  

Let me know how it goes.  Good luck!

[Murphyky]

Hi Jared,

So as per your instruction i went step by step. These were the steps i took. (Note: The Amplifier named below has a gain of about 2.2)

Applied the 5V, 400 kHz signal from the ARB into our amplifier and then measured the output across a 10 ohm load with the scope and measured 12.7V peak to peak.

Next I connected the output directly into COM0 (+) and COM1 (-) and put the scope with the 10 ohm load across NC0 (+) and NC1 (-) and measured 10.2V peak to peak. Therefore a drop of 2.5V just from putting it through the relay.

Next I connected from those same NC0 (+) and NC1 (-) about a 2 ft cable instead of just about a 2 inch wire from the relay. Again measured with the 10 ohm load and got 9.84V. Therefore a drop of .4V with the longer cable.

Lastly I connected that longer cable to the DUT. Which puts it through a resonant inductor of 2.8uH in series with a resonant capacitor of 66nF in series with a 10 ohm resistor. I measured with a scope at that input location and got 3.6V. Without going through the relay and making that same input connection I measure about 9.2V

The nature of the signal is a sine wave, and i did attempt to switch out the cable to see if that affected anything and a slight decrease but nothing significant.

Please let me know if i can supply any other information to help out.

Thanks,

[JaredRo]

Murphyky,

How are you connecting your scope to the 10 Ohm load?

Because your DUT has a low input impedance, it sounds like the few Ohms of series resistance from your cabling and your switch are creating a voltage divider with your DUT.  Try connecting the signal through the switch with the longer cable to a load larger than 10 Ohms (1 kOhm or larger will be fine).  I expect that the signal will be appear to be near 12.7 Vpp.  This will verify that the switch is working properly. 

You may be able to design an amplifier with a higher gain so the switch/cabling voltage drop doesn't render the signal at the DUT unusable. 

[Murphyky]

Jared,

Before i was connecting the signal to the DUT i would hook up the scope probe clip to one resistor end and the ARB center conductor. The other end of the resistor I would hook to the ARB shield and the scope ground clip.

After I connected the signal to the DUT i would hook up the scope directly to the input tabs on the board. The scope probe clip and ARB center conductor together and the scope ground clip and ARB shield together on the DUT. Now at this point I hook the 10 Ohm resistor up to its location on the DUT at the of the Resonant inductor and resonant capacitor. Now that I look more closely, the 10 Ohm resistor is connected in parallel at the end of the caps, and inductors. Sorry for the confusion.

I did attempt to try different load resistors but basically everything after 10 Ohms gave me the same value of about 3.67V peak-to-peak. I tried, 100 ohm, 1.8K, and 4.99K ohms. Now adversly, I tried a 8.2 ohm, a 5.9 ohm and a 1.2 ohm resistor and the value got steady better ending with a 9.2V peak-to-peak signal with the 1.2 ohm resistor.

Lastly I noticed when i removed the scope ground clip from the return of the DUT the signal was much better, 3.8V peak-to-peak to 9.2V peak-to-peak with still a 10 ohm load.

Please let me know if I can give you any more information that can help.

Thanks

[Tommy_G]

Hi Murphyky,

Could you provide us with a circuit diagram of how you are connecting everything for the different situations so we can get a clearer picture of where everything is connected and be able to better support you with this issue?

Regards,

[Murphyky]

Hi Tommy,

So I have attached the circuit diagram for where I am performing connections and scope measurements. I am connecting the ARB signal to J12 (+) which is coming from NO_0 and J11 (-) which is coming from NO_1. I am also connecting the scope probe to J12 (+) and the ground clip to J11 (-). The 10 ohm load is being put across J2 and J10.

Please let me know if i can explain anything else further.

Thanks.

[JaredRo]

Murphyky,

There’s one more test I’d like you to do with your setup.  Connect the signal from your amplifier, through the switch (modeled as 1 Ohm resistors), and then to a large load (not connected to your DUT).  If the signal shows up without attenuation, then we can confirm that the switch itself is working correctly. 

Switches have a resistance listed in the specifications manual.    It is small, but it should still be considered in your design.  You’ve basically built a voltage divider with your circuit.

A colleague of mine simulated your DUT in our Multisim circuit design software to see if we could determine the exact source of the signal attenuation in your DUT.  What sort of load are the two transformers driving?  We found that we could get a wide range of voltages across the DUT just by altering the load attached to the transformers.  See the images I’ve attached for reference:

With the transformers driving 2 Megaohm loads:

With one load reduced to 100 Ohms:

Do you know the specifications of the amplifier that you're using?  Is it possible that it can't source the power required to drive the DUT.

Let me know how the test with the switch goes.  If it passes, there is some other limitation in the circuit that is causing the signal attenuation.

Good luck with everything.