RI Title
RI logo

Purpose : To provide an example of a Calibration methodology for Lower Frequency ( 100MHz and below ) Test Fixture paths.

In applications such as modulators and demodulators extremely precise stimulus and measurements of AC signals are required. For instance, Demodulators require IQ balance measurements to less than a percent and phase balance measurements to 0.1 degree at the IF port. Modulators require stimulus accuracy to the same levels to insure appropriate characterization. In addition, these measurements typically require fixturing hardware such as low pass filters for BW limitation on IQ stimulus, switching for "switched IQ" measurements of baseband receive, and "signal buffers" for Differential to Single ended translation and vice versa.

Let us look at a typical application:

Here we need to develop a fixture calibration for the Demodulator output and the Modulator input sections of the fixture at "AC" frequencies. The normal RF fixture calibrations will be used for the RF3, 6 and 7 ports. But the other paths in the fixture need to be cal'd at an IF frequency usally around 100kHz. The standard RF calibrations do not extend that low in frequency, and some of the ports used do not have couplers, so alternate approaches must be used.

Calibration Methodology Plan:

To do these calibrations we need to look at what resources we have that are calibrated at these frequencies. The ARBs and Voltmeters and Digitizers are the resources that provide calibrated stimulus or measurement and some or all can be used in the approach.

Let us again look at a typical application:
In addition a number of strategies can be applied depending on the fixturing environment and the known error sources: In this approach we will:

1.) Calibrate the differential receive circuit by applying a common signal to both sides
2.) Use the calibrated differential receive circuit to calibrate the differential source

Let us create a fixture path definition:

On the Receive side RF4 goes to two pins on the dutboard. DutP3 and DutP2.
On the transmit side WF2 connects to DutP22 and WF3 connects to DutP33

Remember to select a calibration type for each path. This is usually of type AC.
Also the "Mode" needs to be defined as either I or Q. There can be any number of characters
in the mode name but the last character needsto be either an upper case I or Q. As a further requirement,
the number of Q paths must equal the number of I paths defined.

For Smart Carriers the 8 switch bits XXXXXXX1 or 0 are replaced with C11 or C10. Cbit 1 can be pulled directly
from the 40 pin header. This was not always the case with earlier versions of the fixture smart carrier.

This results in four fixture path names:





You can inspect these newly created path names by inspecting the fixture calibration.

Now we need to hook up the resources to enable the calibration.

Differential Receive Calibration :

This can be created on a dutboard for ease of implementation. The key to this board is:

1.) Receive Phase: WF2 is connected to both sides of the RF4 input. By definition there will be
0 degrees phase difference between the two.
2.) Receive Amplitude: Actual amplitude can be measured at VM1 and compared to what is
measured at RF4.

Here are the significant panels from the Differential Receive Calibration Plan:

1.) In this panel we measure the actual value of voltage created at the dutboard.
There are several Key elements to consider:

a.) All the fixture path data is Reset.

b.) The Voltage measurement is done at an offset value equal to the maximum and minimum
AC swing of the signal.

c.) The AC component of the arb is "none"

2.) In this panel we take the measured voltage from the previous panel and calculate the Peak to Peak swing.
This is now our magnitude reference.

3.) Now we apply the AC signal (at the same amplitude measured for the reference) and measure it at the
Testhead cal plane for the I path.

a.) The fixture head RF4 button has us switched to the I path.

b.) The 2.828 multiplier is for the receiver's measured rms peak units converted to Vpp
c.) Now the receiver measured PP value is divided by the Actual value measured by the VM.
This is the path loss in Voltage ratio.

d.) It is done 80 times and stuffed into the LV

e.) The same process is then done for the other side of the RF4 path, Path RF4DutP2.

f) For Smart Fixtures the I and Q paths should be identical path names ending with an I or Q

4.) Now the phase difference must me measured. For this we use the Switched IQ measurement.(see link).

Since the same signal is being applied to the I and Q sides of the path any measured phase difference
is the phase component created by the fixture.

Smart Fixture Convention

Note : For Smart fixtures make sure that the I side of the fixture path definition is pointed to in the test panel.

5.) Now all the data has been taken, the I loss, Q loss and the phase difference between them, we then bring
the data back. It is combined to make a vector, and then write it to the pathname.

The following is an example of the I path.

7.) For the Q path it is slightly different. Q is considered absolute for phase so the phase component of the Q path is 0.

That is the end of the Differential Receive Calibration. The fixture must be saved if the test plan is being run out of the editor as opposed to the fixture calibration executive!

At this point a validate program can be written and run to verify the data. Please see the example in the zipped archive. The main difference is there is no data reset and no writing.

Differential Source Calibration:

Now we have a calibrated differential receiver, we can use it to measure and calibrate our differential source. Here is the circuit a cal board layout. It is basically a straight through, I to I , Q to Q.

The concept is to set an Amplitude with the ARB, Measure the amplitude at the Dut Board with the Receiver
and then the ratio is the loss. For the Phase measure switched IQ and the phase delta gets written to the I path.
1.) The following example measures the I path. Note:
a.) Reset calibration
b.) Fixture Head RF4/DutP2 selected
c.) 0.707 nomalizes the ARB's peak voltage to the Receivers RMS.
d.) The measurement is averaged and repeated

2.) The amplitude ratio is then brought in, averaged, and then combined with a phase value 0
(WF2, Q is always considered absolute). This complex data is then written to the cal variable.

3.) The WF3 Path is then measured with the same considerations as WF2.

4.) Now we need to measure the phase balance since WF3 is the relative path.

5.) Then the phase data, and amplitude data are brought back, averaged, combined as a vector and then written to the WF3 path.

That is the end of the Differential Source Calibration. The fixture must be saved if the test plan is being run out of the editor as opposed to the fixture calibration executive!

At this point a validate program can be written and run to verify the data. Please see the example in the zipped archive. The main difference is there is no data reset and no writing.

PrintEmail Link
©2001-2022 Roos Instruments, Inc. All rights reserved.