Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

How to calibrate an open end of semi-rigid coax to be used as a test probe?

Status
Not open for further replies.

myRF

Junior Member level 1
Joined
Oct 10, 2013
Messages
19
Helped
0
Reputation
0
Reaction score
0
Trophy points
1
Location
UK
Activity points
168
Hi,

I need to do some return loss measurements on a PCB and was planning to use a piece of open-ended semi-rigid (034 size) with an SMA on the one end. This will allow me to solder the outer to a nearby ground pad and the inner conductor to the test point in question.

To calibrate to the end of the "probe" I was planning on just doing an open (nothing connected), short (solder blob between outer and inner), and load (51R 0402 chip resistor) between outer and inner.

I realise this is not an ideal calibration but the best method I can think of.

Does anyone have any better suggestions/tips?

Thanks.
 

FvM

Super Moderator
Staff member
Joined
Jan 22, 2008
Messages
48,418
Helped
14,252
Reputation
28,765
Reaction score
12,941
Trophy points
1,393
Location
Bochum, Germany
Activity points
280,188
Semirigid is near to ideal transmission line, you might get off with a calibration at the SMA connector reference plane + port extension.
 
  • Like
Reactions: myRF

    myRF

    Points: 2
    Helpful Answer Positive Rating

SunnySkyguy

Advanced Member level 5
Joined
Sep 26, 2007
Messages
6,744
Helped
1,675
Reputation
3,348
Reaction score
1,644
Trophy points
1,413
Location
Richmond Hill, ON, Canada
Activity points
50,737
You just want to measure S11 on a PCB fixture as a 1 port device with a Network Analyzer?

To what accuracy ?
< 15 dB RL? ( 3%)
< 20 dB (1%)
< 30 dB (0.1%)

Normally Netwk.An. calibration kits include , 0R0, open cct, 50R0 termination to fix to end of jig cable.

Crude test can be a soldered coax,
better test uses onboard test connector like OSX trype
If you dont use, then solder SMA receptacle to board.

For production work on Tx/Rx antenna port to PA/LNA/filter switch , used 0.1dB accuracy for s12 calibration and <-40dB s11, s22 up to 6GHz.
 

myRF

Junior Member level 1
Joined
Oct 10, 2013
Messages
19
Helped
0
Reputation
0
Reaction score
0
Trophy points
1
Location
UK
Activity points
168
I would like better than 20dB accuracy. Unfortunately my cal kit is APC 3.5mm only! I'm not sure I have ever come across a kit with standards for the end of a open ended cable - how do they attach to the cable?

Yeah if I had an on-board connector I would be using it! IMO I think a semi-rigid soldered directly to the pad would give a better launch then an SMA as the SMA pin-to-ground distance will be significant (depending on what type of SMA connector is used) o -15 dB RL.
 

FvM

Super Moderator
Staff member
Joined
Jan 22, 2008
Messages
48,418
Helped
14,252
Reputation
28,765
Reaction score
12,941
Trophy points
1,393
Location
Bochum, Germany
Activity points
280,188
UFL connectors and cables might be an alternative for a pluggable solution up to 6 GHz. I have used it recently for PCB antenna measurements. You can make your own calibration kit to be plugged to the cable.
 

vfone

Advanced Member level 5
Joined
Oct 10, 2001
Messages
5,295
Helped
1,549
Reputation
3,099
Reaction score
1,156
Trophy points
1,393
Activity points
33,559
To calibrate to the end of the "probe" I was planning on just doing an open (nothing connected), short (solder blob between outer and inner), and load (51R 0402 chip resistor) between outer and inner

Yes, you can do the OPEN, SHORT, and LOAD calibration as you mentioned. Even the semi-rigid coax cable have good properties, this kind of calibration will take care for the length of the semi-rigid cable.

To increase the upper frequency limit of the LOAD calibration I recommend to use 3 x 150 ohms 0402 resistors in parallel (placed in star configuration). It's a little bit hard for soldering, but in this way you decrease the parasitic inductance of the load resistors.

The semi-rigid cables that you are going to use for the measurement should be the same type, and must have the same length as the semi-rigid cables used for calibration.
 
  • Like
Reactions: ferdem

    ferdem

    Points: 2
    Helpful Answer Positive Rating

SunnySkyguy

Advanced Member level 5
Joined
Sep 26, 2007
Messages
6,744
Helped
1,675
Reputation
3,348
Reaction score
1,644
Trophy points
1,413
Location
Richmond Hill, ON, Canada
Activity points
50,737
I would like better than 20dB accuracy. Unfortuantely my cal kit is APC 3.5mm only! I'm not sure I have ever come across a kit with standards for the end of a open ended cable - how do they attach to the cable?

Yeah if I had an on-board connector I would be using it! IMO I think a semi-rigid soldered directly to the pad would give a better launch then an SMA as the SMA pin-to-ground distance will be significant (depending on what type of SMA connector is used).

Get an onboard connector. Pick any.
https://www.pasternack.com/sma-female-right-angle-connectors-category.aspx

The reason for the cost is precision machining tolerances for impedance control.
Beware of cheap clones.

Otherwise, expect calibration errors for non-ideal open circuit due to stray leakage more than 1% and hope for accuracy to -15 dB RL.
 

E Kafeman

Advanced Member level 2
Joined
Jan 4, 2007
Messages
621
Helped
274
Reputation
548
Reaction score
268
Trophy points
1,343
Location
Sweden
Activity points
8,077
Calibration at PCB is often only way to reach any accuracy if working above few 100 MHz and want good accuracy but there are many problems related to do this calibration.
Ideally is it enough to do port forwarding, compensating for the added cable and maybe a short length of PCB trace and maybe empty component pads, to reach actual measuring point, typical a active circuit port or location of a passive impedance matching circuit.
Result of this port forwarding is often poor, so next step is to do a calibration directly at measurement location at PCB.
Most professional calibration kits, have internal parameters loaded in VNA as these kits includes an offset, either in length or as added fringing capacitance or both. It makes them less useful to be used with home-brewed calibration components.
A possible solution is to temporary switch to user defined calibration parameters in VNA, it will give better result but it is not ideal either, to calibrate with none-offset parameters.
A somewhat better way is to do normal calibration but also include same type and length of semi rigid coax as was intended to be soldered at PCB, but now including connector fitting you calibration kit and finally do a port extension.
It is not the best to include a extra connector in calibration process, and then exclude it in actual measurement and at PCB, even very short PCB traces and pads can create problems if you not want to include them in measurement, and these will not be compensated with this method.
If working above 2GHz can resulting error be huge.

To solve these problems, I recommend my own Win software, AnTune. It is free to try and is fully functional in trial mode. AnTune require a GPIB, USB or LAN connection to the VNA.
AnTune have automatic port forwarding settings, not only as a time-delay but do also calculate a complex network to linearize for stray capacitance or inductive losses in PCB traces or attenuation in measurement cables.
It is very simple to use, just push a button while shortcutting measurement point and AnTune will then show live calibrated S11 result.
 

biff44

Advanced Member level 5
Joined
Dec 24, 2004
Messages
4,903
Helped
1,360
Reputation
2,716
Reaction score
1,039
Trophy points
1,393
Location
New England, USA
Activity points
36,892
an open circuited semirigid line is NOT the same as a theoretical open circuit. there is a fringing capacitance at the end that needs to be approximated for the de-embedding calculations to be accurate. Now if you are at 100 MHz, you can ignore that effect. but at 5 GHz, it will be significant.
 

Status
Not open for further replies.

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Top