Active GPS patch test successful

Found an old Pharos PX003 active GPS ceramic patch antenna, from the iPaq days, in the junkbox, so thought to try it out on Outernet. Opened the case, removed the 1575MHz filter, bridged the trace, shortened the cable length by half, installed a SMA connector, put it all back together, connected directly to RTL-SDR, and it seems to work pretty well sitting on top of the plastic box, in the window.

GPS antenna is almost identical to the one pictured here: (How to Modify a GPS Antenna for Inmarsat Reception)

This type of antenna will set you back at least $9.

You will want to have access to a SMD rework station or fine hot air gun to remove the filter easily, leaving traces intact. You will want to shorten the lead considerably as RG-174 is terrible at 1.5GHz.

I might have to try this. What is your signal with the outernet patch antenna? Just curious about the difference in snr.

That should not matter because the antenna has a built-in preamp. Remember it was designed for the
same general frequency and in its original application (which also is weak-signal) the preamp gain is
apparently enough to overcome the cable loss.

It matters a lot if you don’t want to waste the preamp gain. The dongles usually come with several feet of cable. In this case, it was about 10 feet. That is over 3.5dB loss at 1539MHz, and a little more with the connector. That is a loss of over half the available signal after the preamp. Considering that this ceramic patch has little if any gain at all, any preamplification gain needs to be preserved for reliable decoding. As you can see with the RSSI at 117dB with the shortened lead, there is little room let over to play.

This is the one I want to try next

Compared to the Outernet patch and LNA combo, it is within a dB or so RSSI. There are a couple of points to consider here:

  1. I removed any filtering the GPS antenna has. That makes it susceptible to more interference
  2. The GPS patch works best when positioned slightly above a metal backplane.
  3. The GPS patch is “aimed” in the general direction of the satellite. It did not work well just sitting there horizontally.

Still, a good start for some experimentation. For example, what happens if the GPS patch is integrated into a small package with a tuned reflector behind it?

Rg174 has a loss off 1,2 db / meter… I think we must give the normal users of outernet a message that every meter of extra coax can be a working or a not working system… And for so far i know is 3db more of less 50% of a signal… So cut The cable… wil give you extra information.

Oh but it is a misunderstanding that it works like that!
I just inserted 5 m of rg316 cable between the preamp and the RTL stick.
This caused the RSSI to drop down about 6dB to about -125 (from -119)dBm, but it has
no noticable effect on the SNR. Remember the SNR is determined at the preamp input,
as long as there is enough gain in the preamp.

That might be OK with a filtered LNA.

It is not different for filtering or not filtering. What matters is de noise factor of the preamp and the RTL
stick, and the gain of the preamp. The Outernet LNA has so much gain that 6dB loss in the cable should
not matter, and this appears true in practice too.
I got my SMA M/F cable from China today and I tried it with the antenna still on the window sill where
it always was. Only inserted the cable and moved the RTL stick and CHIP a few meters.
With a GPS antenna it will matter how much gain there is in the antenna preamp. However, those antennas
are typically delivered with 4-5 m of RG174 and a GPS receiver is typically designed to operate with
such an antenna (so it is not optimized for ultra-low noise factor either), so the situation should be similar.
Of course a GPS antenna is less effective than the Outernet patch antenna directed at the satellite, but
that matter is not affected by the cable loss.

If filtering is inconsequential, then a RTL dongle with an integrated 19dB preamp and no filter should be able to decode Outernet using only the Outernet patch antenna. It will not do so, in my tests. Nor can it provide a signal quality sufficient to decode NCS STD-C at all, or Aero well. At least in my location. The same unfiltered RTL dongle will not decode Outernet using a helical antenna. It will almost decode some Aero traffic.

Let us be very carefull. Many outernet users have never seen or used a coax… That’s The only reason that i say ‘cut that cable’. It’s not because i can measure my antenne that everyone can do this… The most important thing must be that everyone can use outernet, not only a few freaks :wink:

I don’t think filtering is inconsequential, but I think it is not related to cable length. There certainly is
a case to be made for filtering, as it makes life much easier for the limited dynamic range of the RTL
stick. Apparently it is still possible to receive something on a GPS antenna with the filter removed,
but it would be better to replace it with a suitable filter for 1546 MHz.

Considering cheap GPS active patch antennas, the supplied RG-174 is already out of spec for this frequency, and coupled with the fact it is likely not very “good” RG-174, a long run of it likely will tend to attenuate the amplified signal, and also will be more susceptible to picking up extraneous noise in the environment. With no filtering on the antenna, and also no filtering at the receiver port, all the noise is compounded.

If I were using LMR-240 or better I wouldn’t worry at all about using long lengths of cable betweeen the GPS antenna and the receiver.

My goal is to devise an inexpensive integrated amplifier antenna using a GPS antenna. The next step is to enclose this in a weatherproof plastic pipe which will also serve as the form for a helical passive director of around 5 turns. I wouldn’t mind finding a filter to replace the 1575 one, but it may not be necessary if the satellite signal can be brought up.

Seeing the performance of this cobbled together GPS and backplane, with the larger 40mm patch and possibly a filter, plus the backplane of 120mm or so, I believe a portable low profile antenna solution can be produced at low cost.

There is a direct relation between the gain of the antenna (relative to isotropic radiator), the noise
factor of the preamp, of course the signal strength from the satellite, and the resulting SNR of the signal
at the receiver.
No amount of “raising the signal” is going to change that, no matter how much you hope for it to do so.
As the receiver is imperfect, filtering the signal so it is not going to be overdriven by out-of-band
signals is also going to help.