wrote in message
...
Yeah but in the real world you don't usually need to match a commercial
TDR to find a fault.

Back in the days of ethernet over RG-58 I was able to find lots of
poorly
attached connectors and cables crushed by the electricians that
installed
them (there were no network engineers and installers in those days)
with
a 100 MHz 'scope and a pulse generator made from a 555 timer and a 50
Ohm
line driver.

Can you detect where RG-58 was kinked and then straightened out with the
damaged area being less than 1/2 inch? Likewise can you detect an
abraded shield leaving a 1/4 inch hole in the shielding?

An LM555 has a rise and fall time of 100 nanoseconds. I have no idea of
how much your line driver sharpens the edges of the pulse, but a good
line driver should provide rise and fall times of 5 nanoseconds. The
original TEK 1502 at 140 picoseconds can resolve to about 2 centimeters.
With a risetime of 5 nS, your resolution will be about 70 times worse
(140 cm). Add to this the 100 MHz limitation of the scope (an additional
10 nS), and it becomes far worse.

Yes, you can easily detect shorts and open circuits with your setup, but
you will not be able to detect the 1 to 2 inches or so where the center
conductor has migrated in the foamed dielectric from hanging the cable
over a building edge.

The accuracy of the measurement for fault finding doesn't need to be
much
better than a few feet to be able to find the fault visually once you
know
about where it is and faults at the end are immediately obvious.

If the task is to find {gross} faults, a 'scope and simple pulse
generator works
just fine.

If the task is to certify 6 inches of hard line to GHz to some
mil-spec, you
probably want something more sophisticated.

The kind of fault Spamhog originally described, a 75 ohm cable dropping
to less than 50 ohms over two inches and going back to 75 ohms, will not
be detected by your setup.

73, Barry WA4VZQ

Barry wrote:
wrote in message
...
Yeah but in the real world you don't usually need to match a commercial
TDR to find a fault.

Back in the days of ethernet over RG-58 I was able to find lots of
poorly
attached connectors and cables crushed by the electricians that
installed
them (there were no network engineers and installers in those days)
with
a 100 MHz 'scope and a pulse generator made from a 555 timer and a 50
Ohm
line driver.

Can you detect where RG-58 was kinked and then straightened out with the
damaged area being less than 1/2 inch? Likewise can you detect an
abraded shield leaving a 1/4 inch hole in the shielding?

I was able to detect where cables were squashed by overly tight bundle lacing
and bending over a short radius.

It got me close enough that a visual inspection of a few feet found the
"bad" place.

An LM555 has a rise and fall time of 100 nanoseconds. I have no idea of
how much your line driver sharpens the edges of the pulse, but a good
line driver should provide rise and fall times of 5 nanoseconds. The
original TEK 1502 at 140 picoseconds can resolve to about 2 centimeters.
With a risetime of 5 nS, your resolution will be about 70 times worse
(140 cm). Add to this the 100 MHz limitation of the scope (an additional
10 nS), and it becomes far worse.

Like I said, if you have a pair of eyes, all you have to do is get close.

And in any case, the solution for a cable run that is hosed somewhere in
the middle is to replace the entire section. Cutting a section out and
putting in connectors to splice the cable is just asking for more problems.

Yes, you can easily detect shorts and open circuits with your setup, but
you will not be able to detect the 1 to 2 inches or so where the center
conductor has migrated in the foamed dielectric from hanging the cable
over a building edge.

Yet I could, so either I'm a lier or you are overestimating how hard it
is in the real world.

The accuracy of the measurement for fault finding doesn't need to be
much
better than a few feet to be able to find the fault visually once you
know
about where it is and faults at the end are immediately obvious.

If the task is to find {gross} faults, a 'scope and simple pulse
generator works
just fine.

If the task is to certify 6 inches of hard line to GHz to some
mil-spec, you
probably want something more sophisticated.

The kind of fault Spamhog originally described, a 75 ohm cable dropping
to less than 50 ohms over two inches and going back to 75 ohms, will not
be detected by your setup.

It would detect that there was an impedance lump.

It would not tell you it was a 50 Ohm lump.


--
Jim Pennino

Remove .spam.sux to reply.

Jim,

I think you need to go back and read Spamhog's original question. He was
trying to determine whether the center conductor of a piece of coax had
migrated away from center. He knew where this might have happened - 10
feet from the end, and the migration would have occurred over less than
two inches. So the question of locating where the problem might be is
moot. What is needed is a measurement of the cable impedance in this
region.

First, let us get an estimate of what the impedance of the damaged
section might be. Spamhog was using RG-6 cable with a foamed
polyethylene dielectric. Its velocity factor is 0.85 making its relative
permittivity 1.384. The center conductor is 1 mm, and the normal
diameter of the center insulator is 4.7 mm. The thickness of the
insulator is 1.85 mm. We need to know the impedance if the center
conductor had migrated 0.925 mm toward the jacket.

For a quick estimate, use the formula for off-center coax
(http://www.microwaves101.com/encyclo...offcenter.cfm). This
gives an impedance of 69.8 ohms in this section compared to 78.9 ohms in
the non-distorted coax. A TDR displays the reflection coefficient
from -1 (short) to +1 (open). Here the reflection coefficient is -0.06.
So the TDR trace will drop from the center line by 6% for 200
picoseconds.

If your 100 MHz scope has a typical Gaussian response, its rise time is
at least 3.5 nanoseconds. Do you really think that your oscilloscope
trace will clearly show the 200 picosecond dip? Even with the
wide-screen magnifier that KB7QHC suggested, I think you will have great
difficulty seeing this.

73, Barry WA4VZQ

Barry wrote:
Jim,

I think you need to go back and read Spamhog's original question. He was
trying to determine whether the center conductor of a piece of coax had
migrated away from center. He knew where this might have happened - 10
feet from the end, and the migration would have occurred over less than
two inches. So the question of locating where the problem might be is
moot. What is needed is a measurement of the cable impedance in this
region.

First, let us get an estimate of what the impedance of the damaged
section might be. Spamhog was using RG-6 cable with a foamed
polyethylene dielectric. Its velocity factor is 0.85 making its relative
permittivity 1.384. The center conductor is 1 mm, and the normal
diameter of the center insulator is 4.7 mm. The thickness of the
insulator is 1.85 mm. We need to know the impedance if the center
conductor had migrated 0.925 mm toward the jacket.

For a quick estimate, use the formula for off-center coax
(http://www.microwaves101.com/encyclo...offcenter.cfm). This
gives an impedance of 69.8 ohms in this section compared to 78.9 ohms in
the non-distorted coax. A TDR displays the reflection coefficient
from -1 (short) to +1 (open). Here the reflection coefficient is -0.06.
So the TDR trace will drop from the center line by 6% for 200
picoseconds.

If your 100 MHz scope has a typical Gaussian response, its rise time is
at least 3.5 nanoseconds. Do you really think that your oscilloscope
trace will clearly show the 200 picosecond dip? Even with the
wide-screen magnifier that KB7QHC suggested, I think you will have great
difficulty seeing this.

73, Barry WA4VZQ

The bandwidth of the 'scope will make the trace have a ripple intead of
the nice, sharp bump you would get from a faster 'scope.

Spamhog's original statement was that the cable was crushed to half diameter
for about 2 inches as I recall.

I was able to see ripples in the display for cables with less crush then
that which were on the order of 1/4 wide.

The bottom line is the faster the 'scope you use and the faster the rise
time of the applied pulse, the better the meaurement.

And since this is a hobby and not building man rated space craft, I would
say try whatever you can get your hands on for free and see what happens.

Or spend eternity arguing whether or not it is possible to do.

--
Jim Pennino

Remove .spam.sux to reply.

On Nov 10, 7:44*pm, "Barry" wrote:
Jim,

I think you need to go back and read Spamhog's original question. *He was
trying to determine whether the center conductor of a piece of coax had
migrated away from center. *He knew where this might have happened - 10
feet from the end, and the migration would have occurred over less than
two inches. *So the question of locating where the problem might be is
moot. *What is needed is a measurement of the cable impedance in this
region.

First, let us get an estimate of what the impedance of the damaged
section might be. *Spamhog was using RG-6 cable with a foamed
polyethylene dielectric. *Its velocity factor is 0.85 making its relative
permittivity 1.384. *The center conductor is 1 mm, and the normal
diameter of the center insulator is 4.7 mm. *The thickness of the
insulator is 1.85 mm. *We need to know the impedance if the center
conductor had migrated 0.925 mm toward the jacket.

For a quick estimate, use the formula for off-center coax
(http://www.microwaves101.com/encyclo...offcenter.cfm). *This
gives an impedance of 69.8 ohms in this section compared to 78.9 ohms in
the non-distorted coax. *A TDR displays the reflection coefficient
from -1 (short) to +1 (open). *Here the reflection coefficient is -0.06..
So the TDR trace will drop from the center line by 6% for 200
picoseconds.

If your 100 MHz scope has a typical Gaussian response, its rise time is
at least 3.5 nanoseconds. *Do you really think that your oscilloscope
trace will clearly show the 200 picosecond dip? *Even with the
wide-screen magnifier that KB7QHC suggested, I think you will have great
difficulty seeing this.

* * 73, Barry *WA4VZQ

It may be that a 100 MHz scope has better than a 3.5 nsec risetime,
given that it is sped'ed for response flatness to that limit and its
response actually extends beyond !00 MHz. In retirement, I no longer
have access to test equipment that would support my point.

On Sun, 14 Nov 2010 18:22:54 -0800 (PST), "Sal M. Onella"
wrote:

It may be that a 100 MHz scope has better than a 3.5 nsec risetime,
given that it is sped'ed for response flatness to that limit and its
response actually extends beyond !00 MHz.

The simple correlation between risetime and bandwidth is roughly:
BW = 1/(3·t)

Unfortunately, peaking bandwidth can degrade risetime, and vice-versa.
O'scopes have a lot of conflicting adjustments within them. The
TEK545 had something like an 8 to 12 hour tune-up procedure for the
average bench tech (a calibration specialist could do it in 3 to 4
hours).

73's
Richard Clark, KB7QHC

"Sal M. Onella" wrote in message
...
It may be that a 100 MHz scope has better than a 3.5 nsec risetime,
given that it is sped'ed for response flatness to that limit and its
response actually extends beyond !00 MHz. In retirement, I no longer
have access to test equipment that would support my point.

See the following article:
http://www.eetimes.com/design/microw...ight-Bandwidth

There are two paragraphs in the article of importance he

"All oscilloscopes exhibit a low-pass frequency response that
rolls-off at higher frequencies, as shown in Figure 1. Most scopes with
bandwidth specifications of 1GHz and below typically have what is called
a Gaussian response, which exhibits a slow roll-off characteristic
beginning at approximately one-third the -3dB frequency. Oscilloscopes
with bandwidth specifications greater than 1GHz typically have a
maximally-flat frequency response, as shown in Figure 2. This type of
response usually exhibits a flatter in-band response with a sharper
roll-off characteristic near the -3dB frequency.

"Closely related to an oscilloscope's bandwidth specification is its
rise time specification. Scopes with a Gaussian-type response will have
an approximate rise time of 0.35/f(sub)BW based on a 10- to 90-percent
criterion. Scopes with a maximally-flat response typically have rise time
specifications in the range of 0.4/f(sub)BW depending on the sharpness of
the frequency roll-off characteristic. But it is important to remember
that a scope's rise time is not the fastest edge speed that the
oscilloscope can accurately measure. It is the fastest edge speed the
scope can possibly produce if the input signal has a theoretical
infinitely fast rise time (0 ps). Although this theoretical specification
is impossible to test (since pulse generators don't have infinitely fast
edges) from a practical perspective, you can test your oscilloscope's
rise time by inputting a pulse that has edge speeds that are 3 to 5 times
faster than the scope's rise time specification."

73, Dr. Barry L. Ornitz WA4VZQ

On 11/10/2010 11:56 AM, Barry wrote:
wrote in message
...
Yeah but in the real world you don't usually need to match a commercial
TDR to find a fault.

Back in the days of ethernet over RG-58 I was able to find lots of
poorly
attached connectors and cables crushed by the electricians that
installed
them (there were no network engineers and installers in those days)
with
a 100 MHz 'scope and a pulse generator made from a 555 timer and a 50
Ohm
line driver.

Can you detect where RG-58 was kinked and then straightened out with the
damaged area being less than 1/2 inch? Likewise can you detect an
abraded shield leaving a 1/4 inch hole in the shielding?

An LM555 has a rise and fall time of 100 nanoseconds. I have no idea of
how much your line driver sharpens the edges of the pulse, but a good
line driver should provide rise and fall times of 5 nanoseconds. The
original TEK 1502 at 140 picoseconds can resolve to about 2 centimeters.
With a risetime of 5 nS, your resolution will be about 70 times worse
(140 cm). Add to this the 100 MHz limitation of the scope (an additional
10 nS), and it becomes far worse.

Yes, you can easily detect shorts and open circuits with your setup, but
you will not be able to detect the 1 to 2 inches or so where the center
conductor has migrated in the foamed dielectric from hanging the cable
over a building edge.

The accuracy of the measurement for fault finding doesn't need to be
much
better than a few feet to be able to find the fault visually once you
know
about where it is and faults at the end are immediately obvious.

If the task is to find {gross} faults, a 'scope and simple pulse
generator works
just fine.

If the task is to certify 6 inches of hard line to GHz to some
mil-spec, you
probably want something more sophisticated.

The kind of fault Spamhog originally described, a 75 ohm cable dropping
to less than 50 ohms over two inches and going back to 75 ohms, will not
be detected by your setup.

73, Barry WA4VZQ



Are you telling all the rest of us that may have inferior equipment that
we shouldn't even try to measure things?

Just asking.

And we aren't all dumb enough not to understand what gives us resolution.

tom
K0TAR

"tom" wrote in message
. net...
Are you telling all the rest of us that may have inferior equipment
that we shouldn't even try to measure things?

No. In fact, playing with a scope and a pulse generator is quite
educational.

And we aren't all dumb enough not to understand what gives us
resolution.

tom
K0TAR

Trying to measure the thickness of a single sheet of paper with a ruler
graduated in eighth's of an inch is analogous. About all you can say
with certainty is that the paper is much thinner than 1/8 inch.

My last post gave a nominal value for the reflection coefficient of coax
with a migrated center conductor like N1PR postulated. At 2 GHz, the
2-inch section will be about 0.4 wavelength. Will you get detectible
reflections? Yes.

73, Barry WA4VZQ

In article ,
"Barry" wrote:

"tom" wrote in message
. net...
Are you telling all the rest of us that may have inferior equipment
that we shouldn't even try to measure things?

No. In fact, playing with a scope and a pulse generator is quite
educational.

And we aren't all dumb enough not to understand what gives us
resolution.

tom
K0TAR

Trying to measure the thickness of a single sheet of paper with a ruler
graduated in eighth's of an inch is analogous. About all you can say
with certainty is that the paper is much thinner than 1/8 inch.

My last post gave a nominal value for the reflection coefficient of coax
with a migrated center conductor like N1PR postulated. At 2 GHz, the
2-inch section will be about 0.4 wavelength. Will you get detectible
reflections? Yes.

73, Barry WA4VZQ

Well not exactly, IF, one were to take a "Stack" of said paper, that was
exactly 1/8 of an Inch thick, then count the number of sheets, and
divide the 1/8 inch by the number of sheets, One could get a VERY close
approximation of the thickness of a single sheet...... Duh...