|
Hello all, I am wondering whether I could use two TinySAs, one as a signal generator and the other as a spectrum analyzer, to verify the amount of signal loss across various lengths of 75 ohm RG-6QS cable.? The frequencies of interest would be 200, 500, 600 and 700 MHz.? The system I am working on would be used for OTA television distribution.? I have tested the cables for return loss with a 75 ohm terminator at the far end, and the return losses are all between 36.8 and 40 dB, which to my understanding is pretty good.? The cables vary in length from 15 feet to 60 feet.
?
For each frequency, this would be done first by generating a known signal level with the TinySA that will be the signal generator, with the signal generator directly connected to the other TinySA (which will measure the signal) with an interconnect adapter (for the minimum loss possible).? I would then compare the directly connected signal level to the signal level shown on the SA when the cable under test is inserted between the signal generator and the SA.? I would also like to perform a similar process to verify losses through some high quality splitters that are rated to 1GHz (return loss at the input on these splitters are all between 35 and 40 dB when all output ports are terminated to 75 ohms).
?
The 75 ohm cable(s) that I am testing have the following loss specifications per 100 ft from the manufacturer:
200 MHz? 2.0db
400 MHz? 3.0dB
700 MHz? 5.9dB
?
My questions are:
- will the performance of the TinySA, both as a signal generator at these frequencies and as a SA, be sufficient to provide me with useful data?? I am concerned whether the readings I obtain at the SA will be stable and accurate enough for the data to be of any value in accurately assessing cable losses on each leg of the system.
- Will the impedance mismatch of the 50 ohm TinySA to the 75 ohm cable be a serious concern to me here, either as a signal generator or a SA?? I am considering using 50 ohm to 75 ohm resistive minimum loss pads (5.7dB loss).? If that is recommended, should I use a pad on both the signal generator as well as the SA?
- Is there a better way to accomplish what I am trying to do above that won't be prohibitively expensive?? I have access to some other test gear, including analog and digital oscilloscopes, as well as an analog Instek RF signal generator (can reach 450 MHz via harmonics) and a freq counter.? I tried to do this using OTA over-the-air signals directly and measuring them with an ATSC analyzer, however, the OTA signal levels constantly change and are not steady enough for me to discern how much of the variation is due to cable losses.
I appreciate any input or advice anyone might be willing to offer.? Thanks in advance, Tony.
|
Tony,
Do you actually NEED to know the exact figure or is this an exercise in "can it be done this way" experiment? If all you need is a rough idea, tape measure, data sheet and calculator.
John VE7KKQ
On Mon, Dec 14, 2020 at 1:05 PM Tony Dudieron via <tdudieron= [email protected]> wrote: Hello all, I am wondering whether I could use two TinySAs, one as a signal generator and the other as a spectrum analyzer, to verify the amount of signal loss across various lengths of 75 ohm RG-6QS cable.? The frequencies of interest would be 200, 500, 600 and 700 MHz.? The system I am working on would be used for OTA television distribution.? I have tested the cables for return loss with a 75 ohm terminator at the far end, and the return losses are all between 36.8 and 40 dB, which to my understanding is pretty good.? The cables vary in length from 15 feet to 60 feet.
?
For each frequency, this would be done first by generating a known signal level with the TinySA that will be the signal generator, with the signal generator directly connected to the other TinySA (which will measure the signal) with an interconnect adapter (for the minimum loss possible).? I would then compare the directly connected signal level to the signal level shown on the SA when the cable under test is inserted between the signal generator and the SA.? I would also like to perform a similar process to verify losses through some high quality splitters that are rated to 1GHz (return loss at the input on these splitters are all between 35 and 40 dB when all output ports are terminated to 75 ohms).
?
The 75 ohm cable(s) that I am testing have the following loss specifications per 100 ft from the manufacturer:
200 MHz? 2.0db
400 MHz? 3.0dB
700 MHz? 5.9dB
?
My questions are:
- will the performance of the TinySA, both as a signal generator at these frequencies and as a SA, be sufficient to provide me with useful data?? I am concerned whether the readings I obtain at the SA will be stable and accurate enough for the data to be of any value in accurately assessing cable losses on each leg of the system.
- Will the impedance mismatch of the 50 ohm TinySA to the 75 ohm cable be a serious concern to me here, either as a signal generator or a SA?? I am considering using 50 ohm to 75 ohm resistive minimum loss pads (5.7dB loss).? If that is recommended, should I use a pad on both the signal generator as well as the SA?
- Is there a better way to accomplish what I am trying to do above that won't be prohibitively expensive?? I have access to some other test gear, including analog and digital oscilloscopes, as well as an analog Instek RF signal generator (can reach 450 MHz via harmonics) and a freq counter.? I tried to do this using OTA over-the-air signals directly and measuring them with an ATSC analyzer, however, the OTA signal levels constantly change and are not steady enough for me to discern how much of the variation is due to cable losses.
I appreciate any input or advice anyone might be willing to offer.? Thanks in advance, Tony.
|
Tony:
Let me suggest
you use a Network Analyzer (maybe the NanoVNA?) and take a look
at some of the work by AC6LA.? Start with
which describes how to measure TL characteristics.? Second is to
look at .?
If you determine the TL parameters according to the first link,
you can input that data into the second link and then compare
your info with Dan's coax database.?
I've found the work by AC6LA to be excellent and have
characterized a specific manufacture's 75-ohm coax for use in a
160 meter phased array.
I think this is a much easier approach that trying to worry
about the uncertainties of using 2 TinySAs.
Regards
charlie, k1xx
On 12/14/2020 3:33 PM, Tony Dudieron
via groups.io wrote:
toggle quoted message
Show quoted text
Hello all, I am wondering whether I could use two TinySAs,
one as a signal generator and the other as a spectrum analyzer,
to verify the amount of signal loss across various lengths of 75
ohm RG-6QS cable.? The frequencies of interest would be 200,
500, 600 and 700 MHz.? The system I am working on would be used
for OTA television distribution.? I have tested the cables for
return loss with a 75 ohm terminator at the far end, and the
return losses are all between 36.8 and 40 dB, which to my
understanding is pretty good.? The cables vary in length from 15
feet to 60 feet.
?
For each frequency, this would be done first by generating a
known signal level with the TinySA that will be the signal
generator, with the signal generator directly connected to the
other TinySA (which will measure the signal) with an
interconnect adapter (for the minimum loss possible).? I would
then compare the directly connected signal level to the signal
level shown on the SA when the cable under test is inserted
between the signal generator and the SA.? I would also like to
perform a similar process to verify losses through some high
quality splitters that are rated to 1GHz (return loss at the
input on these splitters are all between 35 and 40 dB when all
output ports are terminated to 75 ohms).
?
The 75 ohm cable(s) that I am testing have the following loss
specifications per 100 ft from the manufacturer:
200 MHz? 2.0db
400 MHz? 3.0dB
700 MHz? 5.9dB
?
My questions are:
- will the performance of the TinySA, both as a signal
generator at these frequencies and as a SA, be sufficient to
provide me with useful data?? I am concerned whether the
readings I obtain at the SA will be stable and accurate enough
for the data to be of any value in accurately assessing cable
losses on each leg of the system.
- Will the impedance mismatch of the 50 ohm TinySA to the 75
ohm cable be a serious concern to me here, either as a signal
generator or a SA?? I am considering using 50 ohm to 75 ohm
resistive minimum loss pads (5.7dB loss).? If that is
recommended, should I use a pad on both the signal generator
as well as the SA?
- Is there a better way to accomplish what I am trying to do
above that won't be prohibitively expensive?? I have access to
some other test gear, including analog and digital
oscilloscopes, as well as an analog Instek RF signal generator
(can reach 450 MHz via harmonics) and a freq counter.? I tried
to do this using OTA over-the-air signals directly and
measuring them with an ATSC analyzer, however, the OTA signal
levels constantly change and are not steady enough for me to
discern how much of the variation is due to cable losses.
I appreciate any input or advice anyone might be willing to
offer.? Thanks in advance, Tony.
|
Charlie,
??? Thank you for the valuable advice regarding the use of a VNA.? I will research the further tonight, which looks very interesting and reasonably priced (I like the use of N connectors instead of SMA too). Zplots looks excellent as well, thank you for bringing that gem to my attention!
??? With regard to the mismatch between the NanoVNA's 50 ohm impedance and the 75 ohm cable impedance, I'm thinking that a couple of 50 to 75 ohm minimum loss pads might be needed in order to get decent results.? Would you agree?
Tony
|
Thanks John for replying back.? Many of these runs were already installed within the walls when I bought the house, so I don't know the exact lengths of all the cables. There are a couple for which I cannot read the manufacturer's label on the cable, although I know it is RG6.? But I do have samples for which I do know the exact length, so perhaps using a VNA and Zplots will be the way to go.
So it is not just a "can it be done this way" exercise but rather I was hoping to come up with a reasonable approach to validate that my calculated losses are close to reality, before I make any design changes to improve the system's performance.? Something better than a rough idea, I guess.
Best regards, Tony
|
Something with a TDR (Time Domain Reflectometer) function would be ideal, not sure if the nanovna has that function, good luck anyway.
John VE7KKQ
On Mon, Dec 14, 2020 at 5:16 PM Tony Dudieron via <tdudieron= [email protected]> wrote: Thanks John for replying back.? Many of these runs were already installed within the walls when I bought the house, so I don't know the exact lengths of all the cables. There are a couple for which I cannot read the manufacturer's label on the cable, although I know it is RG6.? But I do have samples for which I do know the exact length, so perhaps using a VNA and Zplots will be the way to go.
So it is not just a "can it be done this way" exercise but rather I was hoping to come up with a reasonable approach to validate that my calculated losses are close to reality, before I make any design changes to improve the system's performance.? Something better than a rough idea, I guess.
Best regards, Tony
|
Tony:
You do not need 50-75 ohm matching when using the VNA approach
described by AC6LA.
It really is as simple as hooking up a piece of the RG-6 to the
50-ohm port of a VNA and then measuring the resistance/reactance
with the other end of the RG-6 shorted and then opened.? These
measurements are then entered into Zplots for computation.
Here's a small piece of my measured data for a length of C59 by
Davis RF; this was performed on my 50-ohm VNA.
Freq (MHz)??? |Zo|?? ?? Ro??????? Xo??????? VF????? Loss???
/100ft??? /100m??? |S21| (dB)??? Dist R(ft)
?? 1.000??? 73.534??? 73.477??? -2.882??? 0.6408??? 0.255???
0.342??? 1.123?? ? 0.610???????? 57.675
FWIW, I've done several project where I've had to measure 75 ohm
devices with 50 ohm test equipment.? I've used 3 different
methods:
minimum loss pads like you suggested, homewound ferrite
transformers (5:4 turns), and 75:50 ohm transformers from
Minicircuits (ADT1.5-2+ is one example).? So, you have several
options.
As John suggested, you might consider a TDR if that function's
available in something like a NanoVNA.
73 charlie, k1xx
On 12/14/2020 8:04 PM, Tony Dudieron
via groups.io wrote:
toggle quoted message
Show quoted text
Charlie,
??? Thank you for the valuable advice regarding the use of a VNA.?
I will research the further tonight, which looks very interesting and
reasonably priced (I like the use of N connectors instead of SMA
too). Zplots looks excellent as well, thank you for bringing that
gem to my attention!
??? With regard to the mismatch between the NanoVNA's 50 ohm
impedance and the 75 ohm cable impedance, I'm thinking that a
couple of 50 to 75 ohm minimum loss pads might be needed in order
to get decent results.? Would you agree?
Tony
|
On Mon, Dec 14, 2020 at 06:15 PM, charlie carroll wrote:
you might consider a TDR if that function's available in something like a NanoVNA
It is.
|
If you need high accuracy (within 1dB) I also suggest to use a (nano)VNA If you can not get both ends of the coax within 10cm and you are happy to be +/- 1dB two tinySA would do the job, one as signal generator and one as power meter (e.g. in zero span mode) or as spectrum analyzer I've used the tinySA to check the signal levels in a coax distribution system in a house and it did the job very well.
----------------------------------------
For more info on the tinySA go to https://tinysa.org/wiki/
|
TDR is available on my NanoVNA running the 20200118 AA rev firmware. How accurate is it I am not sure, I have not really used it. It can be found under DISPLAY | TRANSFORM | TRANSFORM on/off.
|
Let me amend this by saying NanoVNA-H AA 20200118.DFU firmware.
https://github.com/hugen79/NanoVNA-H/releases/download/0.4.5-4/NanoVNA-H_AA_20200118.dfu
|
Hi there A TDR which really works in both bandpass and low pass mode is NanoVNA-App version 1.1.205 at [email protected] | Files The only negative element is that is can only be calibrated with ideal calibration kit and the SMA kit supplied with the NanoVNA is to be considered as ideal and when thus a calibration done at the female Ch0 everything is fine Kind regards Kurt ? ? TDR is available on my NanoVNA running the 20200118 AA rev firmware. How accurate is it I am not sure, I have not really used it. It can be found under DISPLAY | TRANSFORM | TRANSFORM on/off.
|
Erik, ??? Thanks for your reply.? I definitely will not be able to get both ends of the coax to within 10cm of each other, as the cables I wish to test are all installed.? I had a couple of follow-up questions on your response:
- Being that what I would be most interested in at the "power meter" TinySA is any difference in signal power between the directly connected reference state and the cable-under-test state, would the +/- 1dB accuracy limitation still apply? I ask this only because in my experience some test equipment does much better at measuring small changes in power than they do in measuring absolute power levels.
- In your opinion, will the 50 - 75 ohm impedance mismatch be an issue with the two-TinySA approach? Should I consider using impedance matching pads, in spite of their 5.7 dB loss?? And if so, should it be used at both the signal generator and the "power meter" TinySA?
Thanks, Tony.
|
On Tue, Dec 15, 2020 at 04:28 PM, Tony Dudieron wrote:
- Being that what I would be most interested in at the "power meter" TinySA is any difference in signal power between the directly connected reference state and the cable-under-test state, would the +/- 1dB accuracy limitation still apply? I ask this only because in my experience some test equipment does much better at measuring small changes in power than they do in measuring absolute power levels.
- ?
The minimum resolution of the tinySA level meter is 0.5dB, so yes, it is better than +/- 1dB. but impossible to see any delta below 0.5dB
- ?
- In your opinion, will the 50 - 75 ohm impedance mismatch be an issue with the two-TinySA approach? Should I consider using impedance matching pads, in spite of their 5.7 dB loss?? And if so, should it be used at both the signal generator and the "power meter" TinySA?
- ?
Best is to do both (with and without matching pads) and if they do not give the same cable loss? do believe those with the matching pads. The tinySA in its high output mode has sufficient output power to cater for the 10dB or so loss of the matching pads. And measure at two different output levels to be able to eliminate compression at the receiver side --
------------------------------------------
For more info on the tinySA go to https://tinysa.org/wiki/
|
Hi Tony,
May not be what you are looking for but very interesting all the same, this guy has some very interesting videos, give this one a watch...
https://www.youtube.com/watch?v=z6UJPqQYzNc
John
VE7KKQ
|
Jim Allyn - N7JA
Kurt Poulsen