Is it possible to use single PXIe-5673 for IIP3 (Two-Tone Third-Order Intermodulation Distortion Measurement)?

[Ivan.Chen]

Hi,

In my test system, I need to use PXIe-5673 for the GSM/EDGE/WCDMA modulation application. I also need to test the IIP3(Two-Tone Third-Order Intermodulation Distortion Measurement) performance. As I can see from this link:http://zone.ni.com/devzone/cda/tut/p/id/4384, It states that two signal generator(PXI565X) are needed for the IIP3 measurement. As I already have a PXIe-5673, Is it possible to test IIP3 using one single PXIe-5673 ? or I need to buy one more PXI565X, that means PXI565X+PXIe-5673 for the IIP3 measurement? I am really new guys to the RF field application. Highly appreciate to any kindly reply!

[bettyyu]

If you want to do Two-Tone Third-Order Intermodulation, you need at least two signal sources, f1 and f2. So only one NI-567x is not enough. You need buy at least another signal source, can be NI-565x or 567x, depends on your detail requirements and your frequency range.

[Ivan.Chen]

The reason I raise this question is that: the NI PXIe-5673 is consist of three parts:

1. NI PXIe-5611RF upconverter

2. NI PXI-5652 RF continuous wave (CW) source (carrier wave)
3. NI PXIe-5450 dual-channel arbitrary waveform generator (AWG).

Is it possible that we just generate two-tone signal by the NI PXIe-5450 ,then upconverter these two tone signal by the upconverter PXIe-5611RF ?

[JerryAL]

Hi Ivan

The answer is maybe.  The NI 5673 can generate two tones.  The question is to what level of performance you need in the NI 5673 for IMD3 for the measurement you need to make.  If your DUT has a better IMD3 performance than the NI 5673, then you will be measuring the NI 5673’s IMD3 and not your DUTs.

The NI 5673’s IMD3 specifications are in the module’s specification document.

Having said that, if you do need two CW sources, you can get away with just adding a NI 5652 to the system.

Jerry

[Ivan.Chen]

Hi Jerry,

Thank you very much for your reply.We are trying to introduce PXI solustion for semiconductor test for the production test. If the PXIe-5673 is capable of testing the IIP3, we would not willing to add one more signal generator.Because,you know,the test cost is very sensitive for the production test. The following is the test specification of DUT. Would you help to have a look?

Parameter	Condition	Frequency	Min.	Typ.	Max.	Unit
Intermodulation intercept point (IP3)	Two-tones 900 MHz, 5 MHz spacing +10 dBm each tone	0.5–3 GHz		46 dBm

As I can see from the datasheet of 5673:", NI PXIe-5450 is capable of producing more than 100 MHz of RF bandwidth". For the frequecy,I think it should be able to generate two-tones 900MHz with 5MHz spacing.Is it? But I am not quite sure about the Intermodulation Distortion (IMD3) of the PXIe-5673's signal if we generate it in this case. Do you have any information I can refer to?

It will be highly appreciated if you can give some suggestions on it.

[Ivan.Chen]

Hi Jerry_L,

It's me again. I do lot of research these days. Would you mind spend some time to see if my understanding is correct or not?

As I see from the user manual of PXIe-5673, the worst IMD3 is –48 dbc while the two tone spacing 300KHz.

But base on the IP3 calculation formula: "IP3=Pout+|IMD3/ 2| "and the product specification(see my reply above) of my DUT, the DUT's IMD3 performance should as below:

IMD3(DUT)=-2*(IP3-Pout)=-2*(46dbm-10dbm)=-72dbc.

That means the IMD3 performance of PXIe-5673(-48dbc) is worst than the DUT's (-72dbc). So PXIe-5673 is not suitable for my application.

But I still have concern is that:

These IMD3 are measured under different condition.The PXIe-5673 is measured with two tones, 300 kHz apart, at –6 dBm per tone; And the DUT specification is measured with Two-tones 900 MHz, 5 MHz spacing +10 dBm each tone

Does the IMD3 performance vary a lot when changing the spacing width?Do you have the IMD3 performance of PXIe-5673 with 5MHz spacing? Appreciate it if you can give some suggestions