GPSDO and signal quality (of satellites)

[tkamiya]

I am considering moving my GPS antenna into attic.  My roof is made of just regular stuff.  Asphalt shingles, plywood, and so forth.  I know this will work, as I even had this antenna in my equipment room. 

What I wonder is, is GPS signal quality matter to quality of 1pps output, and ultimately the 10Mhz signal?  I always have lock on at least 6 satellites of various signal strength.

[kj7e]

Texaspyro would be able to better answer this but Ill try.  Just from what I have found out myself, the number of satellites with a stable lock is more important that having an excellent signal over an okay signal.   As long as the signal is adequate for a stable lock, a stronger single wont help improve reference stability much.  If you have a quality gain antenna with low feed line loss you should be just fine.

[texaspyro]

Having strong signals is better than weak ones.   Whenever the receiver changes the satellites it is tracking, you can get a change in the GPDSO output signal.  The size of the change depends upon the GPSDO and its filtering algorithms.  That said, you should be fine  with the antenna in the attic.  An attic antenna can have a better sky view than an outdoor antenna.

Lady Heather has displays for your signal levels. the antenna signal level patterns, and the tracked satellite count.  After collecting a few hours of data you can see where obstructions are blocking the signal and optimize your antenna position.

[David Hess]

In my experience the attic is better than the roof unless the GPS module is firmly attached or under a radome.  Squirrels chewed through the cable on mine and carried the GPS off.

[nctnico]

I've made this GPS coverage image using Lady Heather and a BG7TBL GSPDO + the standard antenna a while ago. My antenna is under a wooden roof.
The data is collected over a period of 48 hours.

[metrologist]

In my experience the attic is better than the roof unless the GPS module is firmly attached or under a radome.  Squirrels chewed through the cable on mine and carried the GPS off.

I'll be pulling my antennas off the roof this afternoon. Thank you...

[jpb]

The attic is easier to fit in but where I live the roof is tiled and it rains almost continuously, water is generally very good at absorbing microwave energy (though generally at X-band) so I suspect that there will be a big signal loss through it but I've not yet tried.

These posts have inspired me to perhaps go up the attic with my daughter's laptop and my GPS mouse to see what the signal strength is like compared to outside.

[chickenHeadKnob]

I live on the south east end of Vancouver Island which has a temperate rainforest climate with so called Mediterranean summers. So called not because they are warm, but because they precipitation drops right down with August often rain free. November-December has rain every day pretty much. My lab is in an upstairs bedroom and I expected to see significant rain fade between summer and winter but that wasn't the case. With two lea-6t receivers I would consistently have 6 sats or more reading 25+ db on ublox control center.

[TheSteve]

I live on the south east end of Vancouver Island which has a temperate rainforest climate with so called Mediterranean summers. So called not because they are warm, but because they precipitation drops right down with August often rain free. November-December has rain every day pretty much. My lab is in an upstairs bedroom and I expected to see significant rain fade between summer and winter but that wasn't the case. With two lea-6t receivers I would consistently have 6 sats or more reading 25+ db on ublox control center.

Have you logged 24/7 to verify there is absolutely no loss of signal?

[chickenHeadKnob]

Have you logged 24/7 to verify there is absolutely no loss of signal?

No not at all. Only two short test runs of about 12 hours one during a medium rain in November and the roof was wet, the other in the middle of a hot dry spell in August. I am not a fully formed time nut.. yet. I was expecting a noticeable drop but it couldn't have been more than a few db from what I remember in August if that. Since the signal of interest is inherently digital I didn't think it matters. As long as the sats were above some angle of elevation (I think around 15-20 degrees) the tracking seemed solid. It sure correctly  locates my lab down to a few meters. My house is standard north american wood frame, hardi board exterior and asphalt shingles. One possible important point: my lab has a 6 by 4 foot south facing window and two 1.5 by 5 foot east windows. Additionally I noticed an improvement when I moved the stock antenna up from the floor and placed it on top of a metal cabinet, 5 to 6 feet off the floor. That is close to how I will finally run them. Before I got into this I was worried I would need an external antenna. You should be able to tell I am a noob with these things.

[tkamiya]

Thanks everybody for input!  I am going to be removing the antenna from outside.

[jpb]

I went up my loft today to store some empty boxes and took the opportunity to take a laptop and my GPS puck to compare the signal quality. It wasn't raining but it had been so the roof tiles were wet.
Here is a comparison between the loft signal, the signal on the fence near my study which is on the north side of the house and the signal on top of my car on the south side (optimal).
The signals are on different days (sorry -not very scientific but they are what I already had).

As might be expected, the loft signal shows more even spread of satellites but the maximum level is around 40 instead of 50 so I guess the roof accounts for around 10dB of loss.

For me, the easiest location to attach is the study fence but it is not good for lots of satellites and I guess may lose signal. The loft is the next easiest location - lots of satellites but longer cable and extra loss. The south side of the house is good for an ideal comparison but no good for practicality.

For timing though, perhaps the fence near the study is best as it gets good signals from around 3 satellites and they are at high elevation (since those at low elevation are blocked by our house and our neighbour's house.

[jpb]

Have you logged 24/7 to verify there is absolutely no loss of signal?

Additionally I noticed an improvement when I moved the stock antenna up from the floor and placed it on top of a metal cabinet, 5 to 6 feet off the floor.

This may well be because the metal cabinet is providing a ground-plane rather than the added closeness to the sky. Depending on the antenna, something like a 10cm ground plane is recommended - I bought a couple of laser cut aluminium disks off ebay in a fit of extravagance but I guess anything metallic will help. (It is dependent on the type of antenna and what it was designed for).

[jgalak]

For those with antennas in the attic, how do you handle a long run down to the rest of you gear (assuming it's long)?  Long antenna cable?  Some sort of amp?  Put GPSDO in the attic and run a long coax for the 10MHz signal?

For me the total run would be on the order of 30-50' as I'd need to drop a cable from an attic above the 2nd floor down to the basement, and then back up to the lab on the 1st floor (unless I ran the cable outside, which I really don't want to do...)  Not sure what that would do to attenuation...

[chickenHeadKnob]

Have you logged 24/7 to verify there is absolutely no loss of signal?

Additionally I noticed an improvement when I moved the stock antenna up from the floor and placed it on top of a metal cabinet, 5 to 6 feet off the floor.

This may well be because the metal cabinet is providing a ground-plane rather than the added closeness to the sky. Depending on the antenna, something like a 10cm ground plane is recommended - I bought a couple of laser cut aluminium disks off ebay in a fit of extravagance but I guess anything metallic will help. (It is dependent on the type of antenna and what it was designed for).

I thought of that. Actually I did a brief experiment: first I had the GPS sitting ontop off a steel mid tower computer case 50 cm off the floor then on a chest high wooden shelf, then on the wooden floor then on the highest point a steel cabinet sitting on that wood shelf. I had ublox control center running on the same midtower PC whilel I was using another PC nearby and occasionally glancing over. At each station I let the unit run for an hour or more, long enough to let the reception stabilize. The reception was fine  at any of those points but showed a small improvement the higher I placed it. I think there may have been a small ground plane effect but it wasn't anything dramatic. Essentially I was just happy I didn't need to make any external plumbing. I have to get back to this project but I am swamped with work.

[David Hess]

For those with antennas in the attic, how do you handle a long run down to the rest of you gear (assuming it's long)?  Long antenna cable?  Some sort of amp?  Put GPSDO in the attic and run a long coax for the 10MHz signal?

For me the total run would be on the order of 30-50' as I'd need to drop a cable from an attic above the 2nd floor down to the basement, and then back up to the lab on the 1st floor (unless I ran the cable outside, which I really don't want to do...)  Not sure what that would do to attenuation...

The remote antenna usually has a built in RF amplifier.  50 feet of RG-6 (75 ohm cable TV coax) will yield a loss of about 6dB at 1.5GHz so it is nothing to worry about.

[jgalak]

75 Ohm, not 50?  Hmmm, I have this spool of RG-6 lying around for no good reason...

[David Hess]

75 Ohm, not 50?  Hmmm, I have this spool of RG-6 lying around for no good reason...

Some GPS antennas and receivers are intended for 75 ohm cable because RG-6 is low loss and cheaply available.  The SWR mismatch is more than made up for by the lower loss.  In practice either can be used.

I usually have more problems with contaminated cable if it is old or an end is exposed to the weather.

[jpb]

Since the signal of interest is inherently digital I didn't think it matters.

It does matter. The content is digital but for timing and positioning it is the phase of the signal that gives the information - e.g. the digital content may give the information as to which satellite it is but to make use of this the receiver needs to have accurate phase information.

If you only want to know the time to within, say, a microsecond then it matters a lot less. Also integrated over time the phase error drops but that is not inherently to do with the digital nature of the signal.

[SoundTech-LG]

I have two antennas at different heights above the roof line, and one indoors pushed all the way up inside a skylight, at, or slightly above the roof line. All three seem to be of equal signal strength, all three antennas are of different manufacture... I have a lot of nasty squirrels around here, but I don't think they are chewing the cable...  YET. Maybe should be thinking of moving more in the skylight, or attic.

[kj7e]

The small square active antennas may work but I had much better results with a dedicated outdoor active antenna.  I found a PCTEL GPS-TMG-HR-26N antenna brand new on eBay for $30;
http://www.antenna.com/apg_search_new.html?type=details&id_num=11589

50' run of LMR-240 (about 5db of loss), the antenna is mounted behind the house and between the first and second story, only visible from the rear of the house.



Only 8 hours of survey, but the signal strength is great, the house shades the singles to the north east.

[BravoV]

About out door antenna for GPSDO, can I use ship's GPS antenna ?
Planning to install it like kj7e did above.

Garmin GA-28 photo, randomly grabbed from Google, click to enlarge.

[metrologist]

that may work fine. Is it tuned for the same GPS band?

[kj7e]

About out door antenna for GPSDO, can I use ship's GPS antenna ?
Planning to install it like kj7e did above.

Garmin GA-28 photo, randomly grabbed from Google, click to enlarge.

If you don't already the GA28, I would recommend looking at some newer offerings, the GA28 is quite old and from what I can tell about 15db of gain.  Search eBay for "GPS Timing antennas" which are tuned for GPS L1 1575.42 MHz and often times have a bit more gain in the 24-26db range.  The extra gain is nice to overcome the feed line loss for a longer run outdoors.  Check the antenna supply voltage on your GPSDO and on the outdoor antennas, most outdoor active antennas will work down to 3.3 volts but 5 volts is recommended.  The BG7TBL GPSDO comes set to 3.3v, need to move a resistor on the PCB for 5v.

[BravoV]

If you don't already the GA28, I would recommend looking at some newer offerings, the GA28 is quite old and from what I can tell about 15db of gain.  Search eBay for "GPS Timing antennas" which are tuned for GPS L1 1575.42 MHz and often times have a bit more gain in the 24-26db range.  The extra gain is nice to overcome the feed line loss for a longer run outdoors.  Check the antenna supply voltage on your GPSDO and on the outdoor antennas, most outdoor active antennas will work down to 3.3 volts but 5 volts is recommended.  The BG7TBL GPSDO comes set to 3.3v, need to move a resistor on the PCB for 5v.

Thanks for the recommendation. My GPSDO board is Trueposition.

Maybe I should re-write the question, will this kind of outdoor antenna even only 15dB gain + cable loss approx. 3-4 meters, is still better than those small match box sized in door GPS antenna ?

I'm looking at local listing, and coincidentally there is a sale for this GA-28 quite cheap, less than 10 bucks. Other choices like your recommendation is not an option as it will be expensive to ship to my country.

[kj7e]

At 1.5GHz, wood and roofing materials add considerable path loss.  A few db of feed line loss is a good trade off for a clear view of the sky.

[BravoV]

At 1.5GHz, wood and roofing materials add considerable path loss.  A few db of feed line loss is a good trade off for a clear view of the sky.

Thats the plan, thanks for confirming and also for inspiring me with your installation photo, my planned spot for installation actually similar to yours, but with bigger view of sky.

[jhenderson0107]

Below is the signal strength plot aggregated over 24 hrs by Lady Heather of the same PCTel GPS-TMG-HR-26N antenna biased at 5V mounted just above the northwest eave of my home office, driving 7m of cable into a 2017 GB7TBL GPSDO.  Typically, nine satellites report SNR >= 40dB. 

[jpb]

It is interesting a lot of the plots shown on this thread where the better reception is all mixed in with less good levels.
The plot for my antenna is as shown and it is good for high angles where it has a good view of the sky and rather worse at lower angles where houses and trees get in the way (it is mounted on a fence about 6ft (~1.8m) off the ground.

It is a Tallysman timing antenna GPS only bought several years ago. I use a 10cm aluminium disk as a ground plane. There is 15m of cable followed by 1m of different cable and several connectors. I did  try using a line amp that I got off ebay but it showed no improvement, if anything it made it worse so I took it out again. The signal goes through an HP splitter (4 way).

[0culus]

I live on the south east end of Vancouver Island which has a temperate rainforest climate with so called Mediterranean summers. So called not because they are warm, but because they precipitation drops right down with August often rain free. November-December has rain every day pretty much. My lab is in an upstairs bedroom and I expected to see significant rain fade between summer and winter but that wasn't the case. With two lea-6t receivers I would consistently have 6 sats or more reading 25+ db on ublox control center.

I wouldn't think significant rain fade should be a problem at the frequencies GPS operates at. Could be wrong though.