Safely getting the electrical length of a connected antenna feedline
Earlier this year, with some help, I pulled the dish off of an old DishTV antenna on the roof and put a 5V bullet antenna on the mast. I also pulled a new cable through by attaching it to the old one. The problem is that I was not able to measure the new cable. So, the question is, without going back on the roof in this heat, how can I measure the electrical length of the line I pulled?
I was thinking of using my 8640B signal generator and sending some RF back up the line to get a quarter wavelength at the null. But that assumes a lot, including that the other end is open at 3MHz, or whatever the frequency works out to be, as well as that the high voltage on the antenna end won't be high enough to blow the LNA.
So, how much RF I can safely send up the line? I've got an 8558B spectrum analyzer, but it's not on the bench, and it would be easier to use my scope, which sadly is a 70s vintage Tek 455. Do I put this all together with a lead from the generator to a tee at the measuring device and tune for a null? My experience at getting precise measurements on anything longer than a few inches is effectively none, but I'd guess that I want less than 0.5V at the LNA during this test. Oh, and I do have an 8444A tracking generator that can output -10 dbm as well as a 10 db attenuator within easy access. That could get a quick spot on the null point.
Most importantly, of course is the question of whether this will even work.
Bob - AE6RV -----------------------------------------------------------------
AE6RV.com
GFS GPSDO list:
groups.yahoo.com/neo/groups/GFS-GPSDOs/info
I host a group called something like HF Antennas. There I posted a link to an article on how to measure coaxial cable. The easiest way is with a spectrum analyzer and a tracking generator.
You connect the generator to the analyzer through a Tee that goes to the unknown coax. You will see a group of peaks and nulls over the spectrum. The spacing is a half wave of the cable. The match needs to not be good to see the nulls best, and you will need to know the propagation constant of the cable. Chances are, the match won't be good over the entire range so you are okay with that. Propagation constant of most coaxial cable runs around 66%.
You can also use a TDR setup but you'll have to make one, with a pulser and a 'scope. I downloaded a circuit for a pulser that uses one IC. I have the parts but haven't built it yet, as I am stalled by the problem of connecting to a 14 pin SMD part. The IC uses one part as an oscillator and the other 5 in parallel to drive 50 Ohms. Again, you use a Tee and measure the time for a reflection, bearing in mind that the trip is two ways over the same cable and the time shown will be double the time for the calculation.
Bob
On Monday, August 8, 2016 12:00 PM, Bob Stewart <bob@evoria.net> wrote:
Earlier this year, with some help, I pulled the dish off of an old DishTV antenna on the roof and put a 5V bullet antenna on the mast. I also pulled a new cable through by attaching it to the old one. The problem is that I was not able to measure the new cable. So, the question is, without going back on the roof in this heat, how can I measure the electrical length of the line I pulled?
I was thinking of using my 8640B signal generator and sending some RF back up the line to get a quarter wavelength at the null. But that assumes a lot, including that the other end is open at 3MHz, or whatever the frequency works out to be, as well as that the high voltage on the antenna end won't be high enough to blow the LNA.
So, how much RF I can safely send up the line? I've got an 8558B spectrum analyzer, but it's not on the bench, and it would be easier to use my scope, which sadly is a 70s vintage Tek 455. Do I put this all together with a lead from the generator to a tee at the measuring device and tune for a null? My experience at getting precise measurements on anything longer than a few inches is effectively none, but I'd guess that I want less than 0.5V at the LNA during this test. Oh, and I do have an 8444A tracking generator that can output -10 dbm as well as a 10 db attenuator within easy access. That could get a quick spot on the null point.
Most importantly, of course is the question of whether this will even work.
Bob - AE6RV -----------------------------------------------------------------
AE6RV.com
GFS GPSDO list:
groups.yahoo.com/neo/groups/GFS-GPSDOs/info
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
Hi
An even more significant question:
Is it worth doing?
More or less:
Do you know the delay numbers for your antenna?
Do you know the delay numbers for your GPS module?
How close can you guess the length of the cable?
Knowing absolutely nothing at all about your setup, I’ll guess the cable is 50 feet long. Maybe I’m off by 20 or 30 feet.
Call that +/-40 to 60 ns. I’d hope you can guess closer than that. Your antenna and module could easily have delays
in the 40 ns range. It has no impact on a “frequency” GPSDO. It is one of a number of static offsets in a time transfer system.
Even the NIST level outfits seem to have issues coming up with a purely mathematical answer to “what is the offset”.
The standard answer is to bring in a calibrated receiver and see how it all measures out.
None of that is to say you should not work out the line length. It’s just to say that there is only so much value to the
measurement.
Bob
On Aug 8, 2016, at 2:18 PM, Bob Stewart bob@evoria.net wrote:
Earlier this year, with some help, I pulled the dish off of an old DishTV antenna on the roof and put a 5V bullet antenna on the mast. I also pulled a new cable through by attaching it to the old one. The problem is that I was not able to measure the new cable. So, the question is, without going back on the roof in this heat, how can I measure the electrical length of the line I pulled?
I was thinking of using my 8640B signal generator and sending some RF back up the line to get a quarter wavelength at the null. But that assumes a lot, including that the other end is open at 3MHz, or whatever the frequency works out to be, as well as that the high voltage on the antenna end won't be high enough to blow the LNA.
So, how much RF I can safely send up the line? I've got an 8558B spectrum analyzer, but it's not on the bench, and it would be easier to use my scope, which sadly is a 70s vintage Tek 455. Do I put this all together with a lead from the generator to a tee at the measuring device and tune for a null? My experience at getting precise measurements on anything longer than a few inches is effectively none, but I'd guess that I want less than 0.5V at the LNA during this test. Oh, and I do have an 8444A tracking generator that can output -10 dbm as well as a 10 db attenuator within easy access. That could get a quick spot on the null point.
Most importantly, of course is the question of whether this will even work.
Bob - AE6RV -----------------------------------------------------------------
AE6RV.com
GFS GPSDO list:
groups.yahoo.com/neo/groups/GFS-GPSDOs/info
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
Hi Bob,
Correct me if I'm wrong, but for timing purposes, I think that what I want is the electrical length in nanoseconds, not the physical length. I hadn't considered the fact that the distance between peaks would give me a halfwave. But, that's a halfwave of what, exactly? Is it the distance from the far end to the tee, or to the generator or to the measurement device (SA/scope/meter)? And since nanoseconds are important, is it actually to some point within the measuring device? Or, since it's only nanoseconds, maybe the last few inches inside the scope aren't that important? It's starting to sound like what I really want to do the final measurement with is my old Voltohmyst with the RF probe at the tee if I really want to be accurate. That would be using the 8640 and tuning for either null or peak, whichever is sharper.
As far as a TDR is concerned, I think that the 1PPS output from my GPSDO should be sufficient. In this case, I think I'd send the 1PPS to a tee at the start input on my 5370 and lock start and stop together with the phase switches opposite for start and stop. For this test, I could actually control the length of time the 1PPS is high so that it doesn't spoil the measurement. But, once again, will I see any sort of spike on the other end with a 3.3V 1PPS pulse?
Bob
-----------------------------------------------------------------
AE6RV.com
GFS GPSDO list:
groups.yahoo.com/neo/groups/GFS-GPSDOs/info
From: Bob Albert <bob91343@yahoo.com>
To: Bob Stewart bob@evoria.net; Discussion of Precise Time and Frequency Measurement time-nuts@febo.com; Discussion of precise time and frequency measurement time-nuts@febo.com
Sent: Monday, August 8, 2016 2:18 PM
Subject: Re: [time-nuts] Safely getting the electrical length of a connected antenna feedline
I host a group called something like HF Antennas. There I posted a link to an article on how to measure coaxial cable. The easiest way is with a spectrum analyzer and a tracking generator.
You connect the generator to the analyzer through a Tee that goes to the unknown coax. You will see a group of peaks and nulls over the spectrum. The spacing is a half wave of the cable. The match needs to not be good to see the nulls best, and you will need to know the propagation constant of the cable. Chances are, the match won't be good over the entire range so you are okay with that. Propagation constant of most coaxial cable runs around 66%.
You can also use a TDR setup but you'll have to make one, with a pulser and a 'scope. I downloaded a circuit for a pulser that uses one IC. I have the parts but haven't built it yet, as I am stalled by the problem of connecting to a 14 pin SMD part. The IC uses one part as an oscillator and the other 5 in parallel to drive 50 Ohms. Again, you use a Tee and measure the time for a reflection, bearing in mind that the trip is two ways over the same cable and the time shown will be double the time for the calculation.
Bob
On Monday, August 8, 2016 12:00 PM, Bob Stewart <bob@evoria.net> wrote:
Earlier this year, with some help, I pulled the dish off of an old DishTV antenna on the roof and put a 5V bullet antenna on the mast. I also pulled a new cable through by attaching it to the old one. The problem is that I was not able to measure the new cable. So, the question is, without going back on the roof in this heat, how can I measure the electrical length of the line I pulled?
I was thinking of using my 8640B signal generator and sending some RF back up the line to get a quarter wavelength at the null. But that assumes a lot, including that the other end is open at 3MHz, or whatever the frequency works out to be, as well as that the high voltage on the antenna end won't be high enough to blow the LNA.
So, how much RF I can safely send up the line? I've got an 8558B spectrum analyzer, but it's not on the bench, and it would be easier to use my scope, which sadly is a 70s vintage Tek 455. Do I put this all together with a lead from the generator to a tee at the measuring device and tune for a null? My experience at getting precise measurements on anything longer than a few inches is effectively none, but I'd guess that I want less than 0.5V at the LNA during this test. Oh, and I do have an 8444A tracking generator that can output -10 dbm as well as a 10 db attenuator within easy access. That could get a quick spot on the null point.
Most importantly, of course is the question of whether this will even work.
Bob - AE6RV -----------------------------------------------------------------
AE6RV.com
GFS GPSDO list:
groups.yahoo.com/neo/groups/GFS-GPSDOs/info
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
If you have a function generator with a sync output, you can use that pulse
into a T connector with your scope to do a TDR measurement of the cable. If
you have a scrap of the new cable, you can use that to calibrate the setup.
Regards,
Tom
----- Original Message -----
From: "Bob Stewart" bob@evoria.net
To: "Discussion of Precise Time and Frequency Measurement"
time-nuts@febo.com
Sent: Monday, August 08, 2016 2:18 PM
Subject: [time-nuts] Safely getting the electrical length of a connected
antenna feedline
Earlier this year, with some help, I pulled the dish off of an old DishTV
antenna on the roof and put a 5V bullet antenna on the mast. I also pulled
a new cable through by attaching it to the old one. The problem is that I
was not able to measure the new cable. So, the question is, without going
back on the roof in this heat, how can I measure the electrical length of
the line I pulled?
I was thinking of using my 8640B signal generator and sending some RF back
up the line to get a quarter wavelength at the null. But that assumes a
lot, including that the other end is open at 3MHz, or whatever the
frequency works out to be, as well as that the high voltage on the antenna
end won't be high enough to blow the LNA.
So, how much RF I can safely send up the line? I've got an 8558B spectrum
analyzer, but it's not on the bench, and it would be easier to use my
scope, which sadly is a 70s vintage Tek 455. Do I put this all together
with a lead from the generator to a tee at the measuring device and tune
for a null? My experience at getting precise measurements on anything
longer than a few inches is effectively none, but I'd guess that I want
less than 0.5V at the LNA during this test. Oh, and I do have an 8444A
tracking generator that can output -10 dbm as well as a 10 db attenuator
within easy access. That could get a quick spot on the null point.
Most importantly, of course is the question of whether this will even
work.
Bob -
AE6RV -----------------------------------------------------------------
AE6RV.com
GFS GPSDO list:
groups.yahoo.com/neo/groups/GFS-GPSDOs/info
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to
https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
Hi Bob,
I believe I know the delay for the antenna and Ublox specifies the delay in the receiver. I can't guess within 20ft of the physical length of the antenna. But, this is time-nuts, so yeah I want to be able to work this out. And besides...
I figured out how to output an accurate 1PPS with my unit. So, it's no longer just a frequency standard. Briefly: I take advantage of the fact that the sawtooth on the Ublox units is less 22ns wide (about +/- 10.5ns according to my data). I send the 1PPS from the receiver through a 30ns delay line to trigger a d-flop in a 7474, which arms the other d-flop in the 7474. The 30ns delay ensures that the 1PPS output of the Ublox, including the sawtooth, is always within the dead time between the pulses from the OCXO. So, the next 10MHz pulse triggers the other d-flop which goes through a 125 gate to the 1PPS output. 1/16th of a second later (a convenient time for me), the PIC resets the two d-flops and so it goes. So, yeah, that's 100ns late, but the Ublox has a userDelay field, which I set to 100ns. That moves the PPS back to the proper time. Yes, there is a delay of the 10MHz through the d-flop and the gate, but that's knowable. And there's about a 5ns difference between the 1PPS output and the 10MHz output. So, given that the delays are all knowable, I think it's usable as a timing unit now; given the limitations of a GPSDO for that purpose.
Any discussion about this method of getting a 1PPS should probably be in a new thread. I'd be happy to link to a schematic if wanted.
Bob
-----------------------------------------------------------------
AE6RV.com
GFS GPSDO list:
groups.yahoo.com/neo/groups/GFS-GPSDOs/info
From: Bob Camp <kb8tq@n1k.org>
To: Bob Stewart bob@evoria.net; Discussion of precise time and frequency measurement time-nuts@febo.com
Sent: Monday, August 8, 2016 3:14 PM
Subject: Re: [time-nuts] Safely getting the electrical length of a connected antenna feedline
Hi
An even more significant question:
Is it worth doing?
More or less:
Do you know the delay numbers for your antenna?
Do you know the delay numbers for your GPS module?
How close can you guess the length of the cable?
Knowing absolutely nothing at all about your setup, I’ll guess the cable is 50 feet long. Maybe I’m off by 20 or 30 feet.
Call that +/-40 to 60 ns. I’d hope you can guess closer than that. Your antenna and module could easily have delays
in the 40 ns range. It has no impact on a “frequency” GPSDO. It is one of a number of static offsets in a time transfer system.
Even the NIST level outfits seem to have issues coming up with a purely mathematical answer to “what is the offset”.
The standard answer is to bring in a calibrated receiver and see how it all measures out.
None of that is to say you should not work out the line length. It’s just to say that there is only so much value to the
measurement.
Bob
On Aug 8, 2016, at 2:18 PM, Bob Stewart bob@evoria.net wrote:
Earlier this year, with some help, I pulled the dish off of an old DishTV antenna on the roof and put a 5V bullet antenna on the mast. I also pulled a new cable through by attaching it to the old one. The problem is that I was not able to measure the new cable. So, the question is, without going back on the roof in this heat, how can I measure the electrical length of the line I pulled?
I was thinking of using my 8640B signal generator and sending some RF back up the line to get a quarter wavelength at the null. But that assumes a lot, including that the other end is open at 3MHz, or whatever the frequency works out to be, as well as that the high voltage on the antenna end won't be high enough to blow the LNA.
So, how much RF I can safely send up the line? I've got an 8558B spectrum analyzer, but it's not on the bench, and it would be easier to use my scope, which sadly is a 70s vintage Tek 455. Do I put this all together with a lead from the generator to a tee at the measuring device and tune for a null? My experience at getting precise measurements on anything longer than a few inches is effectively none, but I'd guess that I want less than 0.5V at the LNA during this test. Oh, and I do have an 8444A tracking generator that can output -10 dbm as well as a 10 db attenuator within easy access. That could get a quick spot on the null point.
Most importantly, of course is the question of whether this will even work.
Bob - AE6RV -----------------------------------------------------------------
AE6RV.com
GFS GPSDO list:
groups.yahoo.com/neo/groups/GFS-GPSDOs/info
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
First, the nulls are sharper than the peaks. What you do is measure the frequency of adjacent nulls and calculate what frequency makes both of those a half wave or multiple thereof. Then you know the number of half waves to the mismatch. It's the distance from the tee to the remote end.
It takes a little thought, or you can read the article which presents some formulas to make for less confusion. I guess. The article focuses on shorted or open lines but you can get readings from nearly any line, as long as it's not a flat line. Of course, the closer you are to a good match, the harder it is to make the measurement, which is why I suggest finding a frequency range far from the design of the termination equipment.
Your TDR should work also.
Bob
On Monday, August 8, 2016 1:23 PM, Bob Albert via time-nuts <time-nuts@febo.com> wrote:
I host a group called something like HF Antennas. There I posted a link to an article on how to measure coaxial cable. The easiest way is with a spectrum analyzer and a tracking generator.
You connect the generator to the analyzer through a Tee that goes to the unknown coax. You will see a group of peaks and nulls over the spectrum. The spacing is a half wave of the cable. The match needs to not be good to see the nulls best, and you will need to know the propagation constant of the cable. Chances are, the match won't be good over the entire range so you are okay with that. Propagation constant of most coaxial cable runs around 66%.
You can also use a TDR setup but you'll have to make one, with a pulser and a 'scope. I downloaded a circuit for a pulser that uses one IC. I have the parts but haven't built it yet, as I am stalled by the problem of connecting to a 14 pin SMD part. The IC uses one part as an oscillator and the other 5 in parallel to drive 50 Ohms. Again, you use a Tee and measure the time for a reflection, bearing in mind that the trip is two ways over the same cable and the time shown will be double the time for the calculation.
Bob
On Monday, August 8, 2016 12:00 PM, Bob Stewart bob@evoria.net wrote:
Earlier this year, with some help, I pulled the dish off of an old DishTV antenna on the roof and put a 5V bullet antenna on the mast. I also pulled a new cable through by attaching it to the old one. The problem is that I was not able to measure the new cable. So, the question is, without going back on the roof in this heat, how can I measure the electrical length of the line I pulled?
I was thinking of using my 8640B signal generator and sending some RF back up the line to get a quarter wavelength at the null. But that assumes a lot, including that the other end is open at 3MHz, or whatever the frequency works out to be, as well as that the high voltage on the antenna end won't be high enough to blow the LNA.
So, how much RF I can safely send up the line? I've got an 8558B spectrum analyzer, but it's not on the bench, and it would be easier to use my scope, which sadly is a 70s vintage Tek 455. Do I put this all together with a lead from the generator to a tee at the measuring device and tune for a null? My experience at getting precise measurements on anything longer than a few inches is effectively none, but I'd guess that I want less than 0.5V at the LNA during this test. Oh, and I do have an 8444A tracking generator that can output -10 dbm as well as a 10 db attenuator within easy access. That could get a quick spot on the null point.
Most importantly, of course is the question of whether this will even work.
Bob - AE6RV -----------------------------------------------------------------
AE6RV.com
GFS GPSDO list:
groups.yahoo.com/neo/groups/GFS-GPSDOs/info
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
I actually tested various 74120 dual 4-input NAND drivers which
produce the sync output on my B&K function generator to find ones
which would provide the fastest and cleanest pulse. AS (advanced
schottky) and FAST (fast advanced schottky TTL) were the best for me.
Modern inexpensive discrete logic however can do a much better these
days.
One interesting thing I learned is that the dual 4-input NAND pinout
overlaps the quad 2-input NAND pinout so in a properly designed layout
where only 2 gates are used, either part can be used.
On Mon, 08 Aug 2016 16:31:44 -0400, you wrote:
If you have a function generator with a sync output, you can use that pulse
into a T connector with your scope to do a TDR measurement of the cable. If
you have a scrap of the new cable, you can use that to calibrate the setup.
Regards,
Tom
Your comments on layout made me think again of how to implement these projects. How do you use a 14 pin SMD IC? I could try to connect it with flying leads but I'd like something better. Is there some kind of socket for these devices? Or a generic board to receive such things?
Bob
On Monday, August 8, 2016 8:02 PM, David <davidwhess@gmail.com> wrote:
I actually tested various 74120 dual 4-input NAND drivers which
produce the sync output on my B&K function generator to find ones
which would provide the fastest and cleanest pulse. AS (advanced
schottky) and FAST (fast advanced schottky TTL) were the best for me.
Modern inexpensive discrete logic however can do a much better these
days.
One interesting thing I learned is that the dual 4-input NAND pinout
overlaps the quad 2-input NAND pinout so in a properly designed layout
where only 2 gates are used, either part can be used.
On Mon, 08 Aug 2016 16:31:44 -0400, you wrote:
If you have a function generator with a sync output, you can use that pulse
into a T connector with your scope to do a TDR measurement of the cable. If
you have a scrap of the new cable, you can use that to calibrate the setup.
Regards,
Tom
time-nuts mailing list -- time-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
There are a few adapters over on that "auction" site.
Do a search on 14 pin smd to 14 pin dip.
73, Dick, W1KSZ
-----Original Message-----
From: time-nuts [mailto:time-nuts-bounces@febo.com] On Behalf Of Bob Albert via time-nuts
Sent: Monday, August 08, 2016 8:37 PM
To: Discussion of precise time and frequency measurement
Subject: Re: [time-nuts] Safely getting the electrical length of a connected antenna feedline
Your comments on layout made me think again of how to implement these projects. How do you use a 14 pin SMD IC? I could try to connect it with flying leads but I'd like something better. Is there some kind of socket for these devices? Or a generic board to receive such things?
Bob
On Monday, August 8, 2016 8:02 PM, David <davidwhess@gmail.com> wrote:
I actually tested various 74120 dual 4-input NAND drivers which produce the sync output on my B&K function generator to find ones which would provide the fastest and cleanest pulse. AS (advanced
schottky) and FAST (fast advanced schottky TTL) were the best for me.
Modern inexpensive discrete logic however can do a much better these days.
One interesting thing I learned is that the dual 4-input NAND pinout overlaps the quad 2-input NAND pinout so in a properly designed layout where only 2 gates are used, either part can be used.
On Mon, 08 Aug 2016 16:31:44 -0400, you wrote:
If you have a function generator with a sync output, you can use that
pulse into a T connector with your scope to do a TDR measurement of the
cable. If you have a scrap of the new cable, you can use that to calibrate the setup.
Regards,
Tom
time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.
time-nuts mailing list -- time-nuts@febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
and follow the instructions there.