PNA-X has a function in IMD class to calculate composite CATV products like

Composite Triple Beat (CTB)

Composite Second-Order (CSO)

Cross-Modulation Distortion (XMOD)

I see an issue with the setting of all 3 of the parameters.

The formula used to calculate the parameters from reading or intermodulation is correct however, the implementation is not.

The formulas for each parameter are as follow:

Mid-Band CTB(dB)=-2(Pi-Ps)+6+10Log(3N2/8)+CTB Offset

Band Edge CTBE(dB)=-2(Pi-Ps)+6+10Log(N2/4)+CTB Offset

CSO(dB)=(Pi-Ps)+10Log(N)+CSO Offset

XMOD=-2(Pi-Ps)+6dB+20Log(N)

We can see common variable Pi (power of intermodulation) , Ps ( power of the carrier).

Now why I think the implementation is wrong.

Let say we have amplifier that have P1dB at 20dBm and IP3 is 40dBm. Now what output power per tone I will set to get best performance form this amplifier and in the same time to get best dynamic range from PNA-X?

We know that PNA-X IM3 around 90-95dBc. We know that we need to set the tone output power to be at list 10dB or better yet 16dB below P1dB to avoided additional distortion from DUT.

In our case the output power per tone should be:

Tone power = P1dB -16dBm

20dBm -16db=4dBm

Knowing the DUT IP3 value, we can calculate power of IM3 and Pi:

IM3= (IP3 –Pout) *(-2)

IM3= ((40dBm-(+4dBm))*(-2)=-72dBc

Pi = IM3+Pout

Pi= -72dBc+4dBm=-68dBm

So -68dBm will be used for calculation of CTB and XMOD. Here is a problem. That represents only 2 carrier. Let me point out that IM3 is-72dBc is well in PNA-X dynamic range.

For realistic test, we need 100 carriers (or channels). 100 carries generate 20dB higher total power in the DUT. In such case, the results of let say CTB will represent intermodulation tone power of 24dBm what is above compression power of the DUT.

What to do in such case … reduce intermodulation tone power by 20dB to -16dBm.

Let calculate:

IM3= (IP3 –Pout) *(-2)

IM3= ((40dBm-(-16dBm))*(-2)=-112dBc

Pi = IM3+Pout

Pi= -112dBc+(-16dBm)=-128dBm

We can see that the IM3 of -112dBc is way outside PNA-X capability.

In the first case we getting too optimistic reading for the set power of -4dBm

In the second case, we are getting way to pessimistic reading due to PNA-X dynamic range limitation.

So what to do?

We can assume that intermodulation below certain power is predictable (reduce the power by 10dB the IM3 drop by 20dB …). What we need is mathematical conversion.

We can measure at optimum power (4dBm per tone), and use the IP3 value to calculate Pi power base on operator power requirement.

So what we need is to add to IMD setting, power value at which the operator like to calculate the CTB, CSO and XMOD .

Regards.

I've forwarded this request to the project team and it is being evaluated. We will post a response here when the evaluation is done.