Topics

VNA Cal Kit

g3cwi <richard@...>
 

Dear all

I have two calibration kits for my VNWA. However I don't know if either is "good". Does anyone here who has access to a proper calibrated VNA be willing to assess my calibration kit to (say) 2GHz?

My two kits are a homemade one and a wimo one and both seem identical in performance.

Thanks.

73

Richard
G3CWI

Mark GM4ISM
 

Richard
If you are up this way I can probably help.
I don't have a VNA but I have access to a calibrated modern R&S spec analyser & tracking generator, a Wiltron VSWR bridge with 46dB directivity & a cal kit with precision loads (50dB ret loss), precision open and shorts all rated to 2GHz and beyond.
The bridge and load are no longer formally calibrated, but I am able to check them regularly, against equipment that is.
Regards
Mark GM4ISM

From: g3cwi
Sent: Monday, March 14, 2011 8:27 PM
Subject: [ukmicrowaves] VNA Cal Kit

 

Dear all

I have two calibration kits for my VNWA. However I don't know if either is "good". Does anyone here who has access to a proper calibrated VNA be willing to assess my calibration kit to (say) 2GHz?

My two kits are a homemade one and a wimo one and both seem identical in performance.

Thanks.

73

Richard
G3CWI

Roger Ray
 

Richard,
I would be happy to measure them on a calibrated VNA to 6GHz, if you could send them.
73
Roger

--- In ukmicrowaves@..., "g3cwi" <richard@...> wrote:

Dear all

I have two calibration kits for my VNWA. However I don't know if either is "good". Does anyone here who has access to a proper calibrated VNA be willing to assess my calibration kit to (say) 2GHz?

My two kits are a homemade one and a wimo one and both seem identical in performance.

Thanks.

73

Richard
G3CWI

g3cwi <richard@...>
 

Thanks for the kind offers (Mark, Darren and Roger). I will pop them in the post to Roger in the first instance! Are you QTHR?

I tried to get them tested today in the RF lab at a particle accelerator but they were too busy.

73

Richard
G3CWI

Roger Ray
 

--- In ukmicrowaves@..., "g3cwi" <richard@...> wrote:

Thanks for the kind offers (Mark, Darren and Roger). I will pop them in the post to Roger in the first instance! Are you QTHR?

I tried to get them tested today in the RF lab at a particle accelerator but they were too busy.

73

Richard
G3CWI

mike.willis@...
 

I would hope so! Was it Alice?

I have a full N and SMA cal kit here, but only a VNWA to compare with.

Mike

--- In ukmicrowaves@..., "Roger" <g8cub@...> wrote:



--- In ukmicrowaves@..., "g3cwi" <richard@> wrote:

Thanks for the kind offers (Mark, Darren and Roger). I will pop them in the post to Roger in the first instance! Are you QTHR?

I tried to get them tested today in the RF lab at a particle accelerator but they were too busy.

73

Richard
G3CWI

Andy G4JNT
 

I don't think there is really too much need to go overboard on cal kits.   For frequencies up to 18GHz, where many of the commercial ones are designed for, yes, they do become quite critical, BUT.... 
 
... at 1.3GHz maximum,  a short circuit made from an SMA with a disc of copper/timplate soldered across the centre pin/body makes a more-than-adequate short.  The SMA oon its own an adequate open, and a 50 ohm termination is usually a more than OK load.   Any uncertaintires in length on the first two are only going to be of the order of 4mm max (, and the return loss of every SMA 50 ohm load I've ever tested, has always been better than 30dB in the sub-GHz range.   Maybe not satisfactory for teh perfectionists, but who is interested in, or even believes, a return loss of  more than 30dB
 
Shout me down if you don't accept this, but I've always belived in the pragmatic enginering approach and getting results accurate enough, rather that trying for absolute perfection and continuously worrying about it.   And as for worrying about whether out-of-cal cal kits need to be tested, please... :-)
 
I've had to make up some odd cal kits for other purposes. Like BNC ones.   And how about one for testing a switchable dipole / Tee antenna.  The cal kit (only needed for 50kHz to 30MHz) had to be based around a 'chocolate-block' screw connecter, isolated from the VNWA with a stack of assorted Ur ferrite rings to make a good wideband common mode choke.    The cal kit itself was a loop of thick copper wire (for the short) , the obvius open, and 3 150 ohms wire ended resistors for the load.   Most of the proving was about making sure the common mode choke was doing its job, by connecting either side of the load back to the casing of the VNWA and making sure the trace didn't shift unacceptably.

On 16 March 2011 21:41, Roger <g8cub@...> wrote:
 



--- In ukmicrowaves@..., "g3cwi" wrote:
>
> Thanks for the kind offers (Mark, Darren and Roger). I will pop them in the post to Roger in the first instance! Are you QTHR?
>
> I tried to get them tested today in the RF lab at a particle accelerator but they were too busy.
>
> 73
>
> Richard
> G3CWI
>


mike.willis@...
 

Andy is quite correct.

There is no need to have a full professionally calibrated kit for what we tend to need to do. My F-type cal kit (don't laugh) is just as he has described. Open is a PCB connector with no pin, short is a PCB connector with a bit of copper shorting it all out and the load is a connector with a 75 ohm chip resistor pin to ground. It is more than good enough up to VHF.

This is no longer the case for metrology or for frequencies greater than a few GHz, but that is not a normal amateur requirement.

Mike

--- In ukmicrowaves@..., Andy Talbot <andy.g4jnt@...> wrote:

I don't think there is really too much need to go overboard on cal kits.
For frequencies up to 18GHz, where many of the commercial ones are designed
for, yes, they do become quite critical, BUT....

... at 1.3GHz maximum, a short circuit made from an SMA with a disc of
copper/timplate soldered across the centre pin/body makes
a more-than-adequate short. The SMA oon its own an adequate open, and a 50
ohm termination is usually a more than OK load. Any uncertaintires in
length on the first two are only going to be of the order of 4mm max (, and
the return loss of every SMA 50 ohm load I've ever tested, has always been
better than 30dB in the sub-GHz range. Maybe not satisfactory for teh
perfectionists, but who is interested in, or even believes, a return loss
of more than 30dB

Shout me down if you don't accept this, but I've always belived in the
pragmatic enginering approach and getting results accurate enough, rather
that trying for absolute perfection and continuously worrying about it.
And as for worrying about whether out-of-cal cal kits need to be tested,
please... :-)

I've had to make up some odd cal kits for other purposes. Like BNC
ones. And how about one for testing a switchable dipole / Tee antenna.
The cal kit (only needed for 50kHz to 30MHz) had to be based around a
'chocolate-block' screw connecter, isolated from the VNWA with a stack of
assorted Ur ferrite rings to make a good wideband common mode choke. The
cal kit itself was a loop of thick copper wire (for the short) , the obvius
open, and 3 150 ohms wire ended resistors for the load. Most of the
proving was about making sure the common mode choke was doing its job, by
connecting either side of the load back to the casing of the VNWA and making
sure the trace didn't shift unacceptably.

On 16 March 2011 21:41, Roger <g8cub@...> wrote:





--- In ukmicrowaves@..., "g3cwi" <richard@> wrote:

Thanks for the kind offers (Mark, Darren and Roger). I will pop them in
the post to Roger in the first instance! Are you QTHR?

I tried to get them tested today in the RF lab at a particle accelerator
but they were too busy.

73

Richard
G3CWI

Christopher Bartram <cbartram@...>
 

I find myself - perhaps surprisingly - in complete agreement with Mike and
Andy. Having been through the 'shall I spend a large amount on a VNA cal kit'
dilemma in a work context, I decided that it wasn't money well spent ... Don't
forget that this is my money, not that of a well funded multinational or a
cosseted governmental research lab! (No slight intended to anyone who works in
those environments! I've been there.)

Most of the time, for everyday measurements in the <2GHz area I use a set of
'standards' made with conventional 0603 resistors, and good quality SMA
sockets mounted on a piece of FR4 PCB material.

My 'good' standards consist of the best SMA-f 50ohm load that I could find on
the surplus market, combined with home-made open and short-circuit loads. As I
have a calibrated air-line, I can check the phase of these, and at the 6GHz
frequency limit of my 'good' net. an., with care I was able to adjust the open
and short circuits to a 180degree, +/-1degree phase difference. The results at
<2GHz agree with my everyday 'standard' to very acceptable limits.

My homemade 'standards' are more than good enough to evaluate my work designs
to 6GHz using my 8753 net. an., and I've never had a conflict with a client.
For my own use I've used the same 'standards' to calibrate my wonderful old
clunker 8410 VNA system and used it very successfully, albeit with quite a lot
of manual calculation to about 15GHz.

There's a lot of b******s talked about VNA cal. kits. Unless you actually
_need_ National Standards traceable equipment, which by definition for amateur
radio, you don't, make your own! If there is sufficient interest, I'll write
something up for Scatterpoint.

Vy 73

Chris
GW4DGU

Geoff Blake <geoff@...>
 

On 17 March 2011 00:43, Christopher Bartram <cbartram@...> wrote:

[Snip]
``
My homemade 'standards' are more than good enough to evaluate my work designs
to 6GHz using my 8753 net. an., and I've never had a conflict with a client.
For my own use I've used the same 'standards' to calibrate my wonderful old
clunker 8410 VNA system and used it very successfully, albeit with quite a lot
of manual calculation to about 15GHz.

There's a lot of b******s talked about VNA cal. kits. Unless you actually
_need_ National Standards traceable equipment, which by definition for amateur
radio, you don't, make your own! If there is sufficient interest, I'll write
something up for Scatterpoint.

Vy 73

Chris
GW4DGU

Hi Chris, I for one would be very interested in your thoughts and practice.

Personally, I love a 50 Ohm sliding load, only because I have one :-)

Geoff G8GNZ

 


------------------------------------

Yahoo! Groups Links

<*> To visit your group on the web, go to:
   http://groups.yahoo.com/group/ukmicrowaves/

<*> Your email settings:
   Individual Email | Traditional

<*> To change settings online go to:
   http://groups.yahoo.com/group/ukmicrowaves/join
   (Yahoo! ID required)

<*> To change settings via email:
   ukmicrowaves-digest@...
   ukmicrowaves-fullfeatured@...

<*> To unsubscribe from this group, send an email to:
   ukmicrowaves-unsubscribe@...

<*> Your use of Yahoo! Groups is subject to:
   http://docs.yahoo.com/info/terms/


Roger Ray
 

I agree with most of the comments. All you really need is good 50 ohm load, and a short that is a good short. The main thing is to know that load is good, and a check that it has not changed. The important bit is to put it in a box. Dosn't have to be a fancy wooden box, just so long as you keep it separate. Mainly to stop yourself using it for that quick check on that 5W TX!
I can't quite see the built a load with resistors, when you can pick up a good sma or N load for a couple of quid on ebay or a radio rally

Some good response from a 'blank' posting. Must do that again ..not.
Roger

--- In ukmicrowaves@..., Geoff Blake <geoff@...> wrote:

On 17 March 2011 00:43, Christopher Bartram <cbartram@...> wrote:

[Snip]
``

My homemade 'standards' are more than good enough to evaluate my work
designs
to 6GHz using my 8753 net. an., and I've never had a conflict with a
client.
For my own use I've used the same 'standards' to calibrate my wonderful old
clunker 8410 VNA system and used it very successfully, albeit with quite a
lot
of manual calculation to about 15GHz.

There's a lot of b******s talked about VNA cal. kits. Unless you actually
_need_ National Standards traceable equipment, which by definition for
amateur
radio, you don't, make your own! If there is sufficient interest, I'll
write
something up for Scatterpoint.

Vy 73

Chris
GW4DGU

Hi Chris, I for one would be very interested in your thoughts and practice.
Personally, I love a 50 Ohm sliding load, only because I have one :-)

Geoff G8GNZ





------------------------------------

Yahoo! Groups Links



Andy G4JNT
 

Certainly agree with keeping the bits is a box (old tobacco tin here). Its also worth putting in a few assorted coax adaptors too, so that when you need to shift from one connector type to another, you always use the same adapters for measurement and remove one of the unknowns from the measurements.
 
A DC check of a load is worth doing.  If an SMA termination measures 50.0 at DC, there's a pretty good chance it will be 50+j not-a-lot at 1GHz since most are designed for use up to 12GHz
 
'JNT


 

On 17 March 2011 10:19, Roger <g8cub@...> wrote:
 

I agree with most of the comments. All you really need is good 50 ohm load, and a short that is a good short. The main thing is to know that load is good, and a check that it has not changed. The important bit is to put it in a box. Dosn't have to be a fancy wooden box, just so long as you keep it separate. Mainly to stop yourself using it for that quick check on that 5W TX!
I can't quite see the built a load with resistors, when you can pick up a good sma or N load for a couple of quid on ebay or a radio rally

Some good response from a 'blank' posting. Must do that again ..not.
Roger

--- In ukmicrowaves@..., Geoff Blake wrote:


>
> On 17 March 2011 00:43, Christopher Bartram wrote:
>
> [Snip]
> ``
>
> > My homemade 'standards' are more than good enough to evaluate my work
> > designs
> > to 6GHz using my 8753 net. an., and I've never had a conflict with a
> > client.
> > For my own use I've used the same 'standards' to calibrate my wonderful old
> > clunker 8410 VNA system and used it very successfully, albeit with quite a
> > lot
> > of manual calculation to about 15GHz.
> >
> > There's a lot of b******s talked about VNA cal. kits. Unless you actually
> > _need_ National Standards traceable equipment, which by definition for
> > amateur
> > radio, you don't, make your own! If there is sufficient interest, I'll
> > write
> > something up for Scatterpoint.
> >
> > Vy 73
> >
> > Chris
> > GW4DGU
> >
> > Hi Chris, I for one would be very interested in your thoughts and practice.
>
> Personally, I love a 50 Ohm sliding load, only because I have one :-)
>
> Geoff G8GNZ
>
>
>
> >
> >
> > ------------------------------------
> >
> > Yahoo! Groups Links
> >
> >
> >
> >
>


Paul Reeves <paul@...>
 

Is another confirmation of Andy's opinion needed? If so, here it is -
although I have access to (inordinately expensive) HP and Anritsu cal
kits (in nice wooden boxes...) I usually end up (at home and, for
prototyping, at work) using just the sort of standards that Andy
describes - although 2x100 ohm sm resistors may be arguably better that
a 50 ohm for the termination. Any measurable difference is below the
error margins (below 2.5GHz) using HP4510, 4753x & similar network
analysers.
But don't tell the QA dept ..... :-)

73s
Paul G8GJA

On Thu, 2011-03-17 at 12:22 +0000, Andy Talbot wrote:

Certainly agree with keeping the bits is a box (old tobacco tin
here). Its also worth putting in a few assorted coax adaptors too, so
that when you need to shift from one connector type to another, you
always use the same adapters for measurement and remove one of the
unknowns from the measurements.

A DC check of a load is worth doing. If an SMA termination measures
50.0 at DC, there's a pretty good chance it will be 50+j not-a-lot at
1GHz since most are designed for use up to 12GHz

'JNT



On 17 March 2011 10:19, Roger <g8cub@...> wrote:

I agree with most of the comments. All you really need is good
50 ohm load, and a short that is a good short. The main thing
is to know that load is good, and a check that it has not
changed. The important bit is to put it in a box. Dosn't have
to be a fancy wooden box, just so long as you keep it
separate. Mainly to stop yourself using it for that quick
check on that 5W TX!
I can't quite see the built a load with resistors, when you
can pick up a good sma or N load for a couple of quid on ebay
or a radio rally

Some good response from a 'blank' posting. Must do that
again ..not.
Roger

--- In ukmicrowaves@..., Geoff Blake <geoff@...>
wrote:


>
> On 17 March 2011 00:43, Christopher Bartram <cbartram@...>
wrote:
>
> [Snip]
> ``
>
> > My homemade 'standards' are more than good enough to
evaluate my work
> > designs
> > to 6GHz using my 8753 net. an., and I've never had a
conflict with a
> > client.
> > For my own use I've used the same 'standards' to calibrate
my wonderful old
> > clunker 8410 VNA system and used it very successfully,
albeit with quite a
> > lot
> > of manual calculation to about 15GHz.
> >
> > There's a lot of b******s talked about VNA cal. kits.
Unless you actually
> > _need_ National Standards traceable equipment, which by
definition for
> > amateur
> > radio, you don't, make your own! If there is sufficient
interest, I'll
> > write
> > something up for Scatterpoint.
> >
> > Vy 73
> >
> > Chris
> > GW4DGU
> >
> > Hi Chris, I for one would be very interested in your
thoughts and practice.
>
> Personally, I love a 50 Ohm sliding load, only because I
have one :-)
>
> Geoff G8GNZ
>
>
>
> >
> >
> > ------------------------------------
> >
> > Yahoo! Groups Links
> >
> >
> >
> >
>