On Mon, 10 May 2004 18:13:00 GMT, Richard Clark
wrote:
Sagging between two plates would seem to me to be a non-issue.

On further reflection (no pun), I would retract this because line sag
would necessarily induce a voltage variation between the probe and the
line that did not actually exist (probe tracking is very important, of
course). Sag may not introduce residual SWR, but it may appear to.

Also, yes, you could use a less than one wavelength long slotted line
- provided you had a sliding load to make up for the remaining length.
In other words, instead of moving the probe to measure the
crests/troughs, you move the load (or both very carefully). When you
introduce limitations into design and wish to maintain precision, it
necessarily follows that you need more standards to compare against.

73's
Richard Clark, KB7QHC

"Richard Clark" wrote in message
...
On Mon, 10 May 2004 06:22:29 GMT, "Jerry Martes"
wrote:
I was hoping to get a response from someone who either had built a line
or
knew of a publication on the pitfalls associated with making a slotted
line.

Hi Jerry,

In response to this, and your email, your wishes are most likely to
evince a deafening silence. If you are trying to survey those who
have built one from a population of those who used them, I think
you've already hit a saturation of 2 possible users. Your own
construction effort would qualify you to expand the population by 50%.

As for why a slab style over a cylinder style. That for me is taken
on faith that greater minds, in the person of HP Metrologists, figured
that one out a long time ago. Myself, I can infer their rationale
that perhaps demands more precision and accuracy than you care to
pursue; but as any path demands a handcrafted solution, why put in 80%
of the effort for a 20% design when 90% effort would double or triple
your return?

Harkening back to my days at Metrology school, I can imagine that the
slab method was chosen because of the inevitable inner line sag that
would inject residual SWR into the system. Sagging between two plates
would seem to me to be a non-issue. Sagging within a cylinder may not
bring enough residual SWR to cause you grief either, but you have to
build and test one to discover the error of your presumption if it
disappoints you.

As for using your scope to eke out the voltage measurements. That is
a tantalizing thought, but the big boys accomplish more with less.
Simplicity is the keyword, with thinking outside of the box. You are
focusing on the literal, absolute voltage measurement when SWR is all
a matter of relativity that affords orders of magnitude more
sensitivity and resolution (and hence accuracy).

Research the Agilent archives for the Metrology papers of the 1960s.
The discussion is very accessible (with only the math necessary to
perform a real measurement) and focused to the concept and the theory.
The writing of that era is a hallmark of clarity.

73's
Richard Clark, KB7QHC

Richard

I used to think the HP slab configuration was choosen so as to put the
probe in a place of minimum fiels within the line. That might be needed so
that the probe introduces a minimum disturbance, or influence. The truth
is; I dont really know why slab is prefered. I thpought I could attach some
light dielectric supports along the bottom of the center conductor to
eliminate the serious sag. I think the diam. ratio for the condoctors is
close to 2.3 to 1 for 50 ohms without dielectric loading. Thats something I
could evaluate after I decide how to make a slotted line.

All thoughts from 'readers' will be appreciated, I'm way out of touch with
antenna design these days.

Jerry

"Richard Clark" wrote in message
...
On Mon, 10 May 2004 18:13:00 GMT, Richard Clark
wrote:
Sagging between two plates would seem to me to be a non-issue.

On further reflection (no pun), I would retract this because line sag
would necessarily induce a voltage variation between the probe and the
line that did not actually exist (probe tracking is very important, of
course). Sag may not introduce residual SWR, but it may appear to.

Also, yes, you could use a less than one wavelength long slotted line
- provided you had a sliding load to make up for the remaining length.
In other words, instead of moving the probe to measure the
crests/troughs, you move the load (or both very carefully). When you
introduce limitations into design and wish to maintain precision, it
necessarily follows that you need more standards to compare against.

73's
Richard Clark, KB7QHC

Richard

This is a good example of how out of touch I am. I thought I could get
all the impedance measuring data from a line only slightly over 1/4 wave
long. Even after thinking about it for several minutes, I cant convince
myself that the slotted line has to be more than 1/2 wave long.

Jerry

On Mon, 10 May 2004 06:22:29 GMT, "Jerry Martes"
wrote:
I was hoping to get a response from someone who either had built a line or
knew of a publication on the pitfalls associated with making a slotted line.


This month's QEX has an article with some information about slotted lines
(used at 2.4 GHz in that article if I remember).

Torsten Clay
N4OGW

How 'bout this...'

"Jerry Martes" wrote in message
news
"Richard Clark" wrote in message
...
On Mon, 10 May 2004 06:22:29 GMT, "Jerry Martes"
wrote:
making a slotted line.

Hi Jerry,
[...] As for why a slab style over a cylinder style. That for me is
taken
on faith that greater minds, in the person of HP Metrologists, figured
that one out a long time ago. ...
73's
Richard Clark, KB7QHC

Richard

I used to think the HP slab configuration was choosen so as to put the
probe in a place of minimum fiels within the line.
Jerry


Gents.
What you are calling "slab line" is more properly called "Stripline".
It is a very well defined type of transmission line...that is, it has exact
equations (unlike the one-sided Micro-strip). I suspect that Jerry's
conclusion is right-on. At least theoretically, there is zero field along a
line through the center conductor, parallel to the ground planes. So it
appears that you would indeed disturb things the least...as long as your
'probe' exited parallel to the ground planes.
Ignore everything but the lower right figure on page 3 of the stripline E
field:
http://www.altera.com/literature/wp/lvdsboardwp.pdf

It is all sort of 'sucked' into the ground planes...
The best pix I could find... I think as long as the probe has a small cross
section when viewed from the line you will disturb the field the least.
Seems to me in the old days, the probe consisted of a 1N21 diode...
--
Steve N, K,9;d, c. i My email has no u's.

On Mon, 10 May 2004 17:14:06 -0500, "Steve Nosko"
wrote:

|How 'bout this...'
|
|"Jerry Martes" wrote in message
|news | "Richard Clark" wrote in message
| ...
| On Mon, 10 May 2004 06:22:29 GMT, "Jerry Martes"
| wrote:
| making a slotted line.
|
| Hi Jerry,
| [...] As for why a slab style over a cylinder style. That for me is
|taken
| on faith that greater minds, in the person of HP Metrologists, figured
| that one out a long time ago. ...
| 73's
| Richard Clark, KB7QHC
|
| Richard
|
| I used to think the HP slab configuration was choosen so as to put the
| probe in a place of minimum fiels within the line.
| Jerry
|
|
|Gents.
| What you are calling "slab line" is more properly called "Stripline".
|It is a very well defined type of transmission line...that is, it has exact
|equations (unlike the one-sided Micro-strip). I suspect that Jerry's
|conclusion is right-on. At least theoretically, there is zero field along a
|line through the center conductor, parallel to the ground planes. So it
|appears that you would indeed disturb things the least...as long as your
|'probe' exited parallel to the ground planes.
|Ignore everything but the lower right figure on page 3 of the stripline E
|field:
|http://www.altera.com/literature/wp/lvdsboardwp.pdf
|
|It is all sort of 'sucked' into the ground planes...
|The best pix I could find... I think as long as the probe has a small cross
|section when viewed from the line you will disturb the field the least.
|Seems to me in the old days, the probe consisted of a 1N21 diode...

Steve Adam in "Microwave Theory and Applications," pp. 384-385
discusses the H-P slab line slotted section. (No surprise, Adam
worked for H-P) He references, "A New Type of Slotted Line Section",
Proceedings of the I.R.E., Vol. 38, No. 3 (March 1950).

The slab line as shown is a round center conductor between parallel
plates. Equations for this type of line are given in the ITT
"Reference Data For Radio Engineers" and are different from those for
stripline, which uses a flat center conductor.

General Radio manufactured a true coaxial slotted line section
(Discussed in their "Handbook of Coaxial Microwave Measurements"), but
it is much more complicated mechanically than the H-P slab line.

As to disturbing the fields with the probe, of course the fields are
disturbed. Just like every other electrical measurement I can think
of, the device or system under test is *always* disturbed. The
question is just a matter of degree.

As far a building a coaxial slotted line long enough for the frequency
of interest, let me just ask how would you propose to machine a
straight, uniform width slot down the length of a round piece of soft
copper. And then, how would you support the center conductor without
the supports getting in the way of the moving probe. Answer these
questions (to start with) and then you'll understand why H-P used slab
line.

"Wes Stewart" wrote in message
...
On Mon, 10 May 2004 17:14:06 -0500, "Steve Nosko"
wrote:

|How 'bout this...'
|
|"Jerry Martes" wrote in message
|news | "Richard Clark" wrote in message
| ...
| On Mon, 10 May 2004 06:22:29 GMT, "Jerry Martes"
| wrote:
| making a slotted line.
|
| Hi Jerry,
| [...] As for why a slab style over a cylinder style. That for me is
|taken
| on faith that greater minds, in the person of HP Metrologists,
figured
| that one out a long time ago. ...
| 73's
| Richard Clark, KB7QHC
|
| Richard
|
| I used to think the HP slab configuration was choosen so as to put
the
| probe in a place of minimum fiels within the line.
| Jerry
|
|
|Gents.
| What you are calling "slab line" is more properly called "Stripline".
|It is a very well defined type of transmission line...that is, it has
exact
|equations (unlike the one-sided Micro-strip). I suspect that Jerry's
|conclusion is right-on. At least theoretically, there is zero field
along a
|line through the center conductor, parallel to the ground planes. So it
|appears that you would indeed disturb things the least...as long as your
|'probe' exited parallel to the ground planes.
|Ignore everything but the lower right figure on page 3 of the stripline E
|field:
|http://www.altera.com/literature/wp/lvdsboardwp.pdf
|
|It is all sort of 'sucked' into the ground planes...
|The best pix I could find... I think as long as the probe has a small
cross
|section when viewed from the line you will disturb the field the least.
|Seems to me in the old days, the probe consisted of a 1N21 diode...

Steve Adam in "Microwave Theory and Applications," pp. 384-385
discusses the H-P slab line slotted section. (No surprise, Adam
worked for H-P) He references, "A New Type of Slotted Line Section",
Proceedings of the I.R.E., Vol. 38, No. 3 (March 1950).

The slab line as shown is a round center conductor between parallel
plates. Equations for this type of line are given in the ITT
"Reference Data For Radio Engineers" and are different from those for
stripline, which uses a flat center conductor.

General Radio manufactured a true coaxial slotted line section
(Discussed in their "Handbook of Coaxial Microwave Measurements"), but
it is much more complicated mechanically than the H-P slab line.

As to disturbing the fields with the probe, of course the fields are
disturbed. Just like every other electrical measurement I can think
of, the device or system under test is *always* disturbed. The
question is just a matter of degree.

As far a building a coaxial slotted line long enough for the frequency
of interest, let me just ask how would you propose to machine a
straight, uniform width slot down the length of a round piece of soft
copper. And then, how would you support the center conductor without
the supports getting in the way of the moving probe. Answer these
questions (to start with) and then you'll understand why H-P used slab
line.

Wes

I wonder why such a smart guy like you would ask about how a person would
machine a straight slot where a "sraight slot" isnt even needed. Why would
you propose use of "soft copper" And, why would you put any supports on the
center conductor when it isnt necessary to put them on top, where the probe
would.be. Besides, at 137 MHZ, is wouldnt be all that necessary to have
the probe moveable continuously. It is aparent to me that either you want
to diminish the value of my learning project or you are just not smart. I
havent read this antenna news group for very long. Your *help* to me is
hopefully not typical. I'm trying to learn. What are you trying to do??

Jerry

Jerry Martes wrote:

I'm pretty sure I'll try to make it coax instead of "slab" like HP does
it. I saw some 2 inch copper pipe at Home Depot today. They sell it in 5
foot lengths for about $22.00. I'm OK with using only 1/2 wavelength for
the slotted coax. I thought I might be able to 'get away with' using my
scope for detecting the line voltages. I am wide open for learning that my
approach has extremely difficult to solve problems. It all seems fairly
easy right now.
I was hoping to get a response from someone who either had built a line or
knew of a publication on the pitfalls associated with making a slotted line.

Jerry

Though you don't usually see it for sale, "coax" can be square as well
as round.

Why build a square line? Because it is a hell of a lot easier to mount
and hold on to while you are doing the metal work and easier to build
a centering probe carriage on a flat surface.

--
Jim Pennino

Remove -spam-sux to reply.

On Tue, 11 May 2004 01:27:50 GMT, "Jerry Martes"
wrote:

[snip]

Let's establish something first. I used to instruct a fourth
year/graduate level microwave measurements lab in which experiment #1
was on the use of slotted lines.

|
| I wonder why such a smart guy like you would ask about how a person would
|machine a straight slot where a "sraight slot" (sic) isnt (sic) even needed.

If you don't think the longitudinal slot needs to be straight down the
length of the line then let's stop here.

If I can convince you that it needs to be then we can proceed.

|Why would
|you propose (the) use of "soft copper"(?)

Perhaps I shouldn't have said "soft" copper; nevertheless, copper
isn't the easiest material to machine, and it was you, you should
recall, that mentioned the use of copper water pipe. If you have a
milling machine with five or six feet of table travel, then good for
you, have at it.

|And, why would you put any supports on the
|center conductor when it isnt (sic) necessary to put them on top, where the probe
|would.be.(?)

The center conductor needs to be precisely located so as to not have
the probe coupling vary as it is moved. This will necessitate a fair
number of disks/beads or whatever. To be precise, the center
conductor diameter should be reduced (or the outer conductor
increased) wherever there is a dielectric support. If you intend to
slot the supports, then the missing dielectric will modify the
effective dielectric constant. Be sure and calculate that effect.


|Besides, at 137 MHZ, is wouldnt be all that necessary to have
|the probe moveable continuously.

Oh, I didn't know that. Apparently General Radio didn't either; they
had a micrometer attached to theirs to measure position to the
thousandths of an inch.

|It is aparent to me that either you want
|to diminish the value of my learning project or you are just not smart.

Why thank you!

|I havent read this antenna news group for very long.

I have, although with some of the inane stuff posted lately, it isn't
often. BTW, it's an *amateur radio* group. I didn't know we had a
band at 137 MHz.

|Your *help* to me is
|hopefully not typical.

No, I believe in this instance that the advice Richard gave you and
what I am trying to give you is atypically better that what you will
usually get.

|I'm trying to learn. What are you trying to do??

Beats the hell out of me.

"Wes Stewart" wrote in message
...
On Tue, 11 May 2004 01:27:50 GMT, "Jerry Martes"
wrote:

[snip]

Let's establish something first. I used to instruct a fourth
year/graduate level microwave measurements lab in which experiment #1
was on the use of slotted lines.

|
| I wonder why such a smart guy like you would ask about how a person
would
|machine a straight slot where a "sraight slot" (sic) isnt (sic) even
needed.

If you don't think the longitudinal slot needs to be straight down the
length of the line then let's stop here.

If I can convince you that it needs to be then we can proceed.

|Why would
|you propose (the) use of "soft copper"(?)

Perhaps I shouldn't have said "soft" copper; nevertheless, copper
isn't the easiest material to machine, and it was you, you should
recall, that mentioned the use of copper water pipe. If you have a
milling machine with five or six feet of table travel, then good for
you, have at it.

|And, why would you put any supports on the
|center conductor when it isnt (sic) necessary to put them on top, where
the probe
|would.be.(?)

The center conductor needs to be precisely located so as to not have
the probe coupling vary as it is moved. This will necessitate a fair
number of disks/beads or whatever. To be precise, the center
conductor diameter should be reduced (or the outer conductor
increased) wherever there is a dielectric support. If you intend to
slot the supports, then the missing dielectric will modify the
effective dielectric constant. Be sure and calculate that effect.


|Besides, at 137 MHZ, is wouldnt be all that necessary to have
|the probe moveable continuously.

Oh, I didn't know that. Apparently General Radio didn't either; they
had a micrometer attached to theirs to measure position to the
thousandths of an inch.

|It is aparent to me that either you want
|to diminish the value of my learning project or you are just not smart.

Why thank you!

|I havent read this antenna news group for very long.

I have, although with some of the inane stuff posted lately, it isn't
often. BTW, it's an *amateur radio* group. I didn't know we had a
band at 137 MHz.

|Your *help* to me is
|hopefully not typical.

No, I believe in this instance that the advice Richard gave you and
what I am trying to give you is atypically better that what you will
usually get.

|I'm trying to learn. What are you trying to do??

Beats the hell out of me.


Well Wes I see it *IS* time to stop this rant. You *are* much more
educated than I am. And I see you as also being smug and negative , at
least as your help applies to my post. Why would I care if the slot was
straight?? That makes no sense to me. I certainly have no reason to
dissuade anyone from making the slot straight. But, why? If you are so
smart, perhaps you can tell me _why_ something is done rather than just
stating "thats the way HP or General Radio did it". What am I missing
here?? I think that it would be nice to have the slot straight, I had
intended to do that a VERY simple way, on a garage sized Bridgeport. Have
you no concept of how to ,machine metal?? I wonder why you even bothered to
post to my question when you make wrong statements.
Please dont go away till you tell why the slot in a line must be straight
if the line is to be used for impedance measurement.
Yeah, Wes, There is an amateur radio group thats interested in monitoring
weather satellites. Does that bother you?? What advice did Richard give
that might help me build a slotted line??
Please take a few minutes to consider the validity of your advice before
offering it. You seem to get side tracked easily like the softness of
copper and the length of what a mill can slot. Try to focus on the fact
that Jerry is trying to learn how to build a slotted line at home. If you
have *ANY* helpfull information I will certainly appreciate hearing it.

Jerry