AnsweredAssumed Answered

Calibration of 10-port MCX device with E5061A ENA + 87075C

Question asked by ljcorsa on Jul 24, 2008
Latest reply on Aug 5, 2008 by ljcorsa
I need some help using an ENA with multiport switch connected to a 10-port mass-interconnect that uses MCX connections.  The current system uses an 8712ES, and they are migrating these test systems to the ENA. 

The system calibration uses a mass interconnect to a custom handmade "cal box" consisting of a row each of opens, shorts, loads, and thrus.  Each of these standards was hand-made using PCB-mounted connectors, then carefully measured and adjusted using a PNA.

During system cal, the mass interconnect is mated manually to the individual rows of standards.  All the error correction is done on the controller PC in software using the SDATA array.  The 8712 is put into default 2-port mode, and the 87075's testset cal is completely bypassed.  The system has seemingly worked well for several years.

Neglecting the error correction, the actual measurement software is pretty simple.  Here are the commands used to capture a sweep (either cal or DUT measurements).  I only show S11, but all four S-params are measured.

Existing commands for the 8712:
:SOUR:POW 10DBM
:DISP:ANN:FREQ1:MODE SSTOP
:CONT:MULT:STATE ON
:CAL:SELF OFF
:SENS:FUNC 'XFR:S 1,1'
:SENS1:CORR:CLASS DEF2
:SENS:FREQ:STAR 5000000
:SENS:FREQ:STOP 70000000
:SENS:SWE:POIN 201
:DISP:WIND1:TRAC:Y:PDIV 10
:DISP:WIND1:TRAC:Y:RLEV 0
:DISP:WIND1:TRAC:Y:RPOS 10
:OUTP ON
:ROUT:PATH:DEF:PORT  5,6
:SENS:FUNC 'XFR:S 1,1'
:ABORT; :INIT1:IMM; *OPC?
:TRACE:DATA? CH1SDATA


Modified commands for the ENA:
:SOUR:POW 10
:DISP:ANN:FREQ ON
:SENS1:CORR:STAT OFF
:SENS1:CORR:MULT OFF
:SENS1:MULT ON; :SENS1:MULT:PROP ON
:CALC1:PAR:COUNT 1; :CALC1:PAR1:DEF S11; :CALC1:PAR1:SEL
:SENS:FREQ:STAR 5000000
:SENS:FREQ:STOP 70000000
:SENS:SWE:POIN 201
:DISP:WIND1:TRAC:Y:PDIV 10
:DISP:WIND1:TRAC:Y:RLEV 0
:DISP:WIND1:TRAC:Y:RPOS 10
:OUTP ON
:SENS1:MULT:PORT1:SEL 5; :SENS1:MULT:PORT2:SEL 6
:TRIG:SOUR BUS; :ABORT; :INIT1:CONT ON;
:TRIG:SING;*OPC?
:CALC1:DATA:SDAT?


I think I have migrated the SCPI commands faithfully, though I'm a bit unsure about the ENA sweep triggering.

Unfortunately, I do not get good agreement on results.  Most of the results are within tenths of a dB, but others are off by several dB.  The direct solution probably requires getting an in-depth understanding of all the error correction software, which I am most reluctant to attempt (I am not an EE).

Are there ways of calibrating this MCX mass interconnect using the built-in features of the ENA and the 87075C?  I am sure there will be a lot of manual connections to make, but if the resulting calibration is stable for weeks or more (as previous experience suggests), it might be better than relying on handmade cal standards and custom error-network solutions.  All suggestions greatly appreciated!  

Outcomes