Hi,

can I use standard type N calibration set to calibtate all the way up to 18 GHz? Typically how far would I be off on S11 measurement, which is only what I am concerned with.

Thanks.

can I use standard type N calibration set to calibtate all the way up to 18 GHz? Typically how far would I be off on S11 measurement, which is only what I am concerned with.

Thanks.

> Hi,

> can I use standard type N calibration set to calibtate all the way up to 18 GHz? Typically how far would I be off on S11 measurement, which is only what I am concerned with.

> Thanks.

What do you mean by a "standard type N calibration set" ? There are numerous different models, and the Agilent 85054B (standard) and 85054D (economy) are both rated to 18 GHz. The 85032B is rated to 6 GHz, the 85032F to 9 GHz.. The 85054B is pretty damm expensive.

You would need to state the model of your cal kit, and preferably your VNA too.

From my understanding, the two most significant errors would be:

*1) The return loss of the loads.* For optimal performance you want sliding loads, but economy cal kits don't have them, so I assume that for less accurate measurements you don't need them.

Using Agilent 909F loads if you don't already have them would be sensible.

*2) The fringing capacitance of the open.*

If you take a kit like the 85032B that I have, I *suspect* that the polynomial fit would be alful at 18 GHz.

*IF* I was going to chance my luck at 18 GHz with my 6 GHz kit, *personally* I'd be tempted to calculate the fringing capacitance of the two opens at 6 GHz by using the polynomial coefficients and assume they are constant above 6 GHz. In other words, lets assume the polynomials indicate at 6 GHz the fringing capacitance are 120 fF for the female open and 64 fF for the male open, then I'd be tempted to define a user cal kit with:

C0=120, C1=0, C2=0, C3=0. OPEN(M)

C0=64, C1=0, C2=0, C3=0. OPEN(F)

That might go horribly wrong, but my guess is it would be better than relying on the polynomial coefficents of the open being OK at 18 GHz.

As for what accuracy you could get, I really don't know.

I don't think the shorts will make too much difference.

You might want to consider buying the two opens from the 85054B cal kit (parts 85054-60027 and 85054-60028), using the shorts you already have, and defining a cal kit which is a mix of the two. That would be my preferred approach. If your kit comes with loads rated to 18 GHz, as are those in the 6 GHz 85032B cal kit, I would think that would be OK.

Dave

PS, the "Network Analyzers - Calibration/Error Correction" is where such questions should be posted, but I'm sure the much the same group of people read the different VNA forums.