Purpose : To provide examples of alternate methods of entering Calibration Data to the Device Interface.
The Dut Interface is last part of the RI7100A cascaded calibration approach. The calibration path is comprised of the tester cal data > fixture cal data > dutinterface cal data. In practice the dutinterface contains two types of data. True calibration data used by the tester in setting stimulus and reporting data, and offset data, used for correlation. This note will deal with the true calibration data.
Typically the dut interface calibration data deals with dutboard components that are hard to include in s-parameter calibration either due to physical connectivity issues or non-symmetry. Examples are baluns, 50 to 75 ohm transitions, circulators, etc.,.
The Dut Interface can be calibrated across frequency just as a fixture is for s-parameters, mag phase or loss/gain, using an automatic calibration plan. But in some cases it is easier to insert a loss value that is appropriate for the frequency of interest or import S-parameter data taken / generated elsewhere. For data generated outside the system, the RI Device Interface can import industry standard Touchtone xxxx.S2P file data. This data is frequency dependant therefore it can do a better job of compensating for corrections needed that vary vs frequency.
Entering a Value for Balun Loss
First lets consider a dutboard that includes a balun, that either by datasheet or measurement has 2.5dB loss. The RI7100A's native measurements are voltage and so all the calibration schemes are based on voltage. Therefore the loss of -2.5dB needs to be converted to voltage ratio:
Recall the transform:
In this case -2.5dB in voltage ratio = 0.74989
1.) First ascertain what DutRf path is desired to be corrected. Remember the fixture path names. For example RF3DutRf8 is the path between The testhead port RF3 to the dutboard DutRf8. This is the path we will use.
2.) Edit the dutinterface file. Select DutRf8. Choose the calibration type Loss. Then Inspect the DutInterface file.
3.) Highlight the DutRf8 entry. The data (assuming it was newly established) should be nil.
4.) Type 0.74989 over the nil value (the system requires the zero ahead of the decimal point.)
5.) Right click and select save.
Now all measurements that use DutRf8 will be corrected for -2.5dB loss.
To Import Touchstone format S2P calibration files first start as we did with single point data:
1) Click on the DutRf8 entry. The data (assuming it was newly established) should be nil.
2) Click the Button 2 (RMB) and select the Reset menu pick. This should reset the data item to a RiFvs2pS data type.
3) Now Click the Button 2 (RMB) and select the Import menu pick. The system will ask you to chose the data type, chose Touchone format
4) Select the file to be imported from the file selection dialog box.
5) Next inspect the Cal Data by double clicking on the DutRf8 entry again. The system will present a dialog box listing each set of S-parameter data vs Frequency. Check to make sure the data is what you expect.
The purpose of the Calibration of this type is to remove data ripple which is a systematic error and as long as it is stable we should be able to remove it. To use cal data in the files generated elsewhere we need to convert the data into an industry standard touchtone S-parameter file format. The data type of your Device Interface or Test Fixture cal variable must be in the 2 port s-paramter form ( RiFvs2pS) for the importer to work. You can reset you cal variables back to RiFvs2pS form so then the importer with allow you to read in a file in Touchtone format.
To see an example of the form the data must be in ( Touchstone S2P ) you can try the exporter and write out a touchtone format file which will show you what to mimic. The data is space separated and in S-parameter grouping - you will need to convert your data to Linear if it is in Log format and add it in both the S12 and S21 spots. Here is what an xxxx.S2P file looks like:
The Touchstone format file is ascii and has a file extension .S2P. (lines that start with "!" are comment lines)
The first line is a header that tells the units of the frequency( #MHz), followed by the data format (S) for S-parameter, followed by the data type (MA) for Magnitude and Angle, followed by R and the Characteristic Impendence ( 50 ) for 50 Ohms.
! 2 port thru data taken at 10 points
#MHz S MA R 50
! Freq, S11 Mag, S11 Angle, S12 Mag, S12 Angle, S21 Mag, S21 Angle, S22 Mag, S22 Angle
200 0 0 1.005545901 -31.996 1.005545901 -31.996 0 0
400 0 0 1.001188725 -66.150 1.001188725 -66.150 0 0
600 0 0 1.040934733 -98.037 1.040934733 -98.037 0 0
800 0 0 0.972613963 -128.62 0.972613963 -128.62 0 0
1000 0 0 0.966660561 -165.69 0.966660561 -165.69 0 0
1200 0 0 1.066248664 164.51 1.066248664 164.51 0 0
1400 0 0 0.982206944 135.54 0.982206944 135.54 0 0
1600 0 0 0.972372124 96.944 0.972372124 96.944 0 0
1800 0 0 1.085877914 66.803 1.085877914 66.803 0 0
2000 0 0 1.010357571 35.261 1.010357571 35.261 0 0
Dut Interface Data with Variation vs Frequency:
In some cases a device interface may need multiple values for loss across frequency. Such as in the case of a PA or Transceiver that has multiple frequency bands that use the same port. In that case we need a different type of dut interface. We will still use the concept of loss in linear terms (Voltage ratio). But the type of dut interface cal must change from loss (which does not support frequency information) to RF. RF is the general case and consists of S-parameters vs frequency.
1.) To create the dut interface cal value, as an S-parameters vs frequency type, Highlight the fixture connection port and select Calibration Type RF
2.) Once the interface has been created, Select Inspect and you will get the dut interface cal data. In this case, the string you see is the s-parameters vs frequency.
Then Double Click on DutRf1 in the left box.
3.) You will now see the data in the cal. In this case there is two points. Selecting the 1 in the left box will show its values in the right box. In this case the frequency is 100 kHz, S21 and S12 are unity, with S11 and S22 at zero.
4.) We will be adding points. In addition we will want to keep points at the beginning and end of the range. This avoids warning messages during compile.
To do the select RiData>Add. You will then get a dialog box asking how many?. Choose the number of frequencies required + 2. The two extras are for 100kHz and 20 GHz.
5.) Now we have 5 points. The added points are at the beginning. We will need to create a new low end point. Select the freq RiFreqD(0.0).
6.) A small dialog comes up. In this case type in the frequency, be sure to use a leading zero before any decimal.
7.) Once the frequency is in we need to modify the loss data. In this case we are filling in the end point. A good default is a lossless data. In the lossless condition, S21 and S12 are 1 and S11 and S22 are 0. Input these values. Select S21 and an additional dialog appears (see #8)
8.) The numbers can be put in real and imaginary or mag and phase by selecting the re/im goggle.
9.) Here is the completed data for the low end (0.1 MHz) frequency point. In addition Point #5 should still be at 20000 (20 GHz) in the lossless condition.
10.) Now it is time to add the data. As an example we want to input a data value of -3.5dB loss at 2.44 GHz. Convert -3.5dB to voltage ratio. (0.668, see beginning of note) and input the value for frequency, s21 and s12. In addition if any phase offsets need to be added they can be added here as well.
11.) After adding all the points, close the data editor. Then in the inspect window for the port, RMBC> Save. Then close the dut interface inspect window.
12.) At the container window RMBC the Dut interface and select Save Calibration.