Siglent SVA1015X correction proceedure

[canyon]

Gang,

I am trying to apply a correction to a sweep of a rather poor 30db attenuator. I am not understanding the correction details in the manual. I have attached a couple of pics of my setup. At the marker, I am trying to apply a -5db correction at 125MHz as the first step in forming a correction table for this attenuator. Turning the correction on and off, I see no change in the trace. Should I?

Thank you,

Ken

[tautech]

To negate the influence that attenuator has on a through signal go back to the TG menu and perform Normalize. You will need to save the Ref trace then apply the Normalize and set the trace to the reference level that suits your needs.

[canyon]

Thank you for your quick reply and the answer to my issue. 73, Ken - wa4mnt

[tautech]

Thank you for your quick reply and the answer to my issue. 73, Ken - wa4mnt

Sorry as I could have addressed how you are attempting to characterize the attenuator instead.

So first a through TG to RF In connection at the levels and span you are to use is Normalized then the attenuator is inserted to see its linearity and cursor measurements taken at points of nonlinearity and recorded.
Applying them as a Correction is probably possible but sorry I'm not sure how. 

BTW
Capturing a sweep with the Trace View function and activating another trace to work with could be handy in this use case. Yesterday I posted some examples that should give you an idea of how this is done:
https://www.eevblog.com/forum/testgear/siglent-sva1015x-1-5ghz-spectrum-vector-network-analyzer-(coming)/msg4443871/#msg4443871

[canyon]

So the key is to determine how the normalized trace is saved and recalled so it can be applied to future measurements using that device?

[tautech]

So the key is to determine how the normalized trace is saved and recalled so it can be applied to future measurements using that device?

Unless someone that has done this chimes in I need to investigate possibilities.
I'm thinking saving a setup and reference trace but I need have a play first.

First you need determine exactly what settings to Normalize with as any changes pop up a Uncal warning as changing settings can impact on accuracy.

Bit to do here but should have time this evening to work on this.

[tautech]

canyon, there are 2 ways to get measurements on/of your attenuator.
Analyzer mode.

First is what one does with just an analyzer sweeping it with the TG over the range you want with a Through connection to RF In.
Start by setting the sweep and TG stimulus amplitude and connect between the TG port and RF In with just a through and Normalize after storing the Reference trace which you may chose to keep hidden or display.

Once you have a through and Normalized trace insert the DUT (attenuator) note its nonlinearity over the swept range where you can change/magnify the scale and drop Markers to take measurements anywhere over the sweep.
Now of course if we are to ignore the DUT's impact on measurements we would instead Normalize with it in circuit.
I've tried to capture my process using a N-N cable and a M-F N type 10dB attenuator/pad in the following screenshots which should be self explanatory.

[tautech]

VNA mode
Log Mag and a S21 measurement is appropriate for sweeping an attenuator but first we need calibrate the VNA with just a simple Response Through with a connection between Port 1 & 2. Then Enter to apply. (png11)
We can see in png12 a trace just like in SA mode Normalized to 0dB ready for the DUT and linearity tests.

With just a Response Through calibration we can adjust the scale and reference positions to display this 10dB pad near the middle of the display however doing so with more complex VNA calibrations the Correction can be lost and the VNA will throw up a Uncal message.
As in SA mode we can drop markers for accurate measurements once we have sufficient vertical resolution to do so.

Like in SA mode you can calibrate/Normalize the attenuator out to have a flat 0dB trace ready for the insertion of another DUT.