5
\$\begingroup\$

I am in need of testing a GPS antenna and cable assembly after installation at a reasonable cost.

I have a GPS antenna installed on an airplane, then the coax cable from the antenna goes down to the receiver. The system is already assembled and validated technically and in production, but we find that after some shipments around the country something goes wrong in the cable assembly.

We are testing out a complete system inside of a metal hanger so there is no available sky view for real satellite reception. Besides we are trying to control the test and provide a go/no-go simple automated test.

We really want to know that we have not pinched the coax cable and confirm that all the connections are still good from outside the antenna through the aircraft and all the way down the cable. I am thinking it should not be necessary to pass a GPS capture and location measurements. I am thinking that I can find a simple method to transmit an RF signal in the GPS range and then receive it on the other end of the cable to confirm the antenna and all the connections and the cable are still good.

I am thinking that I could attach an instrument on the receive end of the coax cable to detect what is transmitted from a non-contact transmitting antenna held a few inches above the GPS antenna.

Any ideas out there? Any simple equipment that does this already that can be connected to a Windows or Linux PC to initiate the test and read back the response signal if not to say go or no-go?

----- edit of March 25 ----- At this point, I believe the best way to go on this test effort is to provide a signal generator driving an antenna, then provide a receiver / detector at the "bottom" end. Now I need to find hopefully a single PCIe card that will give me an L-Band range signal generator and a receiver with an API to control and read from s/w. I think the SDR version is more complex than it needs to be but that still may work. And since the quantities of those are high, that may get those prices down to a good range anyway.

Any ideas out there? I am looking at Pentek, Agilent, NI and any others that may be appropriate. I haven't yet figured out what to call it to do my search for what I need, other than a Signal Generator and Receiver.

\$\endgroup\$
  • 2
    \$\begingroup\$ Sounds like a job for an RF network analyser. Sadly, neither simple nor cheap. \$\endgroup\$ – Brian Drummond Mar 13 '18 at 17:57
  • 1
    \$\begingroup\$ Wouldn't making actual GPS measurements and position fixes be the right kind of idea? \$\endgroup\$ – Nick Alexeev Mar 13 '18 at 18:37
  • \$\begingroup\$ @NickAlexeev - No need to get GPS fixes, Only want to confirm RF signal passing through antenna and through cable. Any valid RF signal - is my thought. \$\endgroup\$ – Bill Tilman Mar 13 '18 at 19:35
  • \$\begingroup\$ @BrianDrummond - My hope is not an RF network analyzer, but simpler since I only need to confirm basically between 1-2GHz bandwidth signal pass through from Antenna through the Coax. \$\endgroup\$ – Bill Tilman Mar 13 '18 at 19:38
  • \$\begingroup\$ Using SDR as a crude network analyser may be an alternative for a go/no-go indication. My only worry would be that it might be more of a research project than I wanted, but Marcus Muller's tutorial may be good start. \$\endgroup\$ – Brian Drummond Mar 13 '18 at 19:42
2
\$\begingroup\$

A TDR (Time Domain Reflectometer) would seem to be the answer here - we used these in the Navy to ring out long RF cable runs/transmission lines. It will tell you if there are any shorts or opens (and where they are) and whether the end is terminated in a proper load (antenna connected).

Its a relatively pricy piece of equipment at about $1500-$2000, but its a huge labor saver.

\$\endgroup\$
  • \$\begingroup\$ Norm - I think the TDR would be good for the cable only. I am after a signal injection from external - RF wise - into the antenna and then to measure at the far end of the coax cable. \$\endgroup\$ – Bill Tilman Mar 13 '18 at 21:28
1
\$\begingroup\$

My intuition would be that you could use a network analyzer for this.

MiniVNA Tiny: http://miniradiosolutions.com/54-2/

Software to use with MiniVNA: http://vnaj.dl2sba.com

I've used a very cheap USB network analyzer ($550) for simple impedance matching. This seems like the cheapest and most direct approach. You can even get one of these for $330 from China on AliExpress (not sure if they ripped off the Italian guys who originally created this or not...).

I've never worked with SDR radio, but there seems like a bit of a learning curve compared to just using a VNA and getting the measurement right away. I am with Brian Drummond using an SDR will be a learning project. If you guys have a GPS system on an airplane I'm guessing $550 is not going to a deal breaker?

A tracking generator is not crazy either (check the manufacturer Rigol for "cheap" spectrum analyzer -- I'm guessing it is $1800 for a spectrum analzyer + tracking generator).

Make sure to order a set of calibrations from the vendor you choose so you don't have to hand solder them. (I ordered from Wimo in Germany).

You are just going to calibrate your unit, get an SMA to Coax adapter because the MiniVNA has SMA ports. (Technically, if you calibrate with the SMA standards from MiniVNA but then add a coax adapter you are adding some inaccuracy, but for a cable good or cable bad approach, I think you should be okay).

NXP MiniVNA Setup Document: https://www.nxp.com/docs/en/application-note/AN11535.pdf

This seems like the easiest path to get your measurement. Check if MiniVNA also has a bluetooth version of sorts, you may be able to setup the VNA at the antenna end and then remotely scan if the antenna is hard to get at.

If you decide to go with this approach, just comment in the answer if you need help later on. I can step you through some stuff if you need help.

\$\endgroup\$
0
\$\begingroup\$

Well, if you just need to test the antenna + coax, then sure: Unplug your GPS receiver, plug in (if necessary, but probably is) a bias-tee to power the LNA in the active antenna, and connect something that can receive at least L1 frequencies.

You'll need something very weak transmitting in the GPS range – not only because otherwise you'd risk saturating the amplifier in the antenna, but also because you don't want to get a visit from the friendly neighbors of your radio regulation agency (FCC if this is the US).

Then, you'll need your receiver to look for that very weak signal in the receiver noise - that's exactly what GPS does, correlating until it finds signal, only for GPS signals.

Both ends, Transmit and Receive, can relatively cheaply be implemented using an SDR (Software Defined Radio). A system that transmits a known sequence on one end of your hall, and another one that looks for it would probably be two separate cheap SDR devices.

Transmitting the known signal would probably be easiest with GNU Radio as signal-generating software (darn easy, tutorial) and a COTS SDR like the Ettus USRP B2xx family (or other manufacturer's products, but no experience with those) with... a paperclip instead of a proper antenna and low transmit gain.

Reception: again, GNU Radio, correlating with the same sequence, and as receiver another one of these SDRs (could've used the same one, but you really want good TX/RX isolation), or, as stupid as that sounds, a $30 RTL-SDR dongle, as your quantization and noise don't matter, due to the immense correlation gain you'll be getting.

\$\endgroup\$
  • \$\begingroup\$ disclosure: I'm involved with both the GNU Radio Project and Ettus; I still think this very likely a very satisfactory solution, and price-wise competitive (I'm not being paid by anyone for this, gentlemen's promise). Bonus is you'll have an SDR lying around, should you need to test anything else (you can easily generate anything from AM voice comms to telemetry and analyze your ADS-B / Mode-S transponder's data with the same device without any further software investment. There's also free GNSS/GPS simulators, but I don't know whether you want to confuse your GPS). \$\endgroup\$ – Marcus Müller Mar 13 '18 at 19:22
  • \$\begingroup\$ I like this idea. This is worth looking at further. Time for some good bit of research. I have seen the Ettus stuff prior, but only as receiver and I thought lower frequencies only. \$\endgroup\$ – Bill Tilman Mar 14 '18 at 16:07
0
\$\begingroup\$

I recommend a signal generator with tracking generator. The tracking generator produces a signal at the same frequency that the spectrum analyser is looking, making it into a simple one-path scalar network analyser.

With a length of cable, a bias T, an attenuator and antenna, you will be able to generate traceable records of performance of the GPS antenna system. You can define a pass/fail limit line and test against that, store and print the result.

You'll need to make a small antenna jig that sits atop the GPS antenna on the aircraft roof. I'd use a plywood frame or a solid block of polystyrene foam, some padding for the aircraft, supporting a transmitting antenna about 400 mm above the GPS antenna.

I mention the attenuator because a) you don't want to transmit a whole milliwatt of RF at GPS frequencies, and b) you'll saturate the GPS antenna if you transmit much more than -30 dBm.

Used, a R&S FSH-3 or similar from Agilent/Keysight, Anritsu site master, etc, will cost only a few thousand dollars. The nice thing about GPS is that it fits into the budget price range for test equipment, it's between the cellular bands. It's battery powered and fairly rugged. A spectrum analyser is a very handy tool for other radio purposes too. Check that it has the tracking generator enabled, sometimes its a software option you pay extra for.

There are also a range of low cost USB network analysers, copper mountain make some for about $10,000.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.