signal strength on a GPS

[Jim Whitby]

I'm asking this here in the hope someone knows something about GPS
receivers.

I'm a decent tech, but not an engineer...

I have a Tomtom 2535 gps. Nice unit blah blah.

Some ( software? ) engineer decided to change the gps info display. It
now contains a pic of earth with dots for satellites instead of a bar
graph. Ok. No biggie.

Except now there is a signal strength percent! Percent of what???

Full quieting? 20 db quieting? 10 db sinad? or ????

I ask because at less than about "80%" I get incomplete data ( speed,
direction, etc ). This is with 7 or more "locked" satellites.

This occurs quite often and I'm trying to figure out if its bad or
"normal" operation. If its weak signal ( 100% is 10 db sinad ), I can
understand it. If 80% is 80% of full quieting, then it must be broke.

Does anyone have a clue what 100% is relative to?

Thanks

Jim

 

[miso]

sci.geo.satellite-nav

 

[Spehro Pefhany]

A software engineer probably doesn't know what a dB is.

dataBase?


Best regards,
Spehro Pefhany

 

[Jim Whitby]

On Sat, 15 Sep 2012 18:04:12 -0700, Jeff Liebermann wrote:

You might get a more specific answer in sci.geo.satellite-nav

Thanks will chk there.

on TomTom
support site: <http://us.support.tomtom.com/app/home/locale/en_us/>
or on multiple TomTom user forums.

Not much help there, from what I've seen. Can't get to anyone at Tomtom
that might know what the answer is ( I suspect (sh)he is locked up in a
Faraday cage).

Agin thanks for lead to sci.geo.satellite-nav

Jim

 

[MrTallyman]

A software engineer probably doesn't know what a dB is.

You're an idiot.

 

[SoothSayer]

Well, the way a GPS receiver works, there is no way to actually
measure the signal. It is below the noise of the receiver!
What they do is autocorrelate the signal with a deciphering key
and if the data passes a checksum, it is good. So, they must be measuring
the ratio of good to bad decoded data blocks.

Jon

They measure timing flag packet arrival times from three satellites.

 

[[email protected]]

Probably signal to noise ratio, in fact bit error rate.
In GPS (the US system) all sats transmit at the same frequency,
their data xored with some pseudo random code that is different for each sat.

If you use a 4-quadrant (Gilbert cell) multiplier (mixer) instead of a
XOR gate, the despread signal will definitely have a positive SNR in
the 1000/50 Hz bandwidth) and the absolute signal power should be
easily measurable.

 

[qrk]

I'm asking this here in the hope someone knows something about GPS
receivers.

I'm a decent tech, but not an engineer...

I have a Tomtom 2535 gps. Nice unit blah blah.

Some ( software? ) engineer decided to change the gps info display. It
now contains a pic of earth with dots for satellites instead of a bar
graph. Ok. No biggie.

Except now there is a signal strength percent! Percent of what???

Full quieting? 20 db quieting? 10 db sinad? or ????

I ask because at less than about "80%" I get incomplete data ( speed,
direction, etc ). This is with 7 or more "locked" satellites.

This occurs quite often and I'm trying to figure out if its bad or
"normal" operation. If its weak signal ( 100% is 10 db sinad ), I can
understand it. If 80% is 80% of full quieting, then it must be broke.

Does anyone have a clue what 100% is relative to?

Thanks

Jim

If it's Trimble, the measurement is relative to AMU (Arbitrary Mystery
Units).
http://gauss.gge.unb.ca/papers.pdf/SNR.memo.pdf

 

[[email protected]]

No, you misunderstand the GPS system.
The satellites all transmit at the same frequency, their digital signals are
xored before modulation with a pseudo random code, a so called 'Gold code'.
This code is different for each satellite.
Basically this is done with a shift register.
In the receiver AFTER demodulation the digital signal is run through a similar
shift register, and one by one tested for the correct Gold code (each sat has its own code).
In case of noise, you do not get an exact match against the know codes, and the number of wrong
bits is then a measure for the signal to noise ratio.
100% good signal, no bit errors.
Weak signal lots of bit errors.
http://en.wikipedia.org/wiki/GPS_signals

The GPS signal is an ordinary direct sequence spread spectrum (DSSS)
signal. It can be received with a mixer/multiplier/xor gate by
multiplying the received signal with the same chip clock (in this case
1.023 MHz satellite specific PRN sequence) _synchronized_ with the
transmitter modulator chip clock. After the demodulator the about 1
MHz wide spread spectrum signal is despread to something about 1 kHz.
At this point you could also make power measurements. After this, the
50 bit/s data signal is extracted.

The shift register and bit error detection you are talking about has
to do with signal acquisition _before_ the receiver chip clock
generator has been synchronized with the transmitter chip clock.

 

[G. Morgan]

Google for example:
IS-GPS-800B.pdf
for a lot of technical details.

I didn't understand a damn thing in that document!

 

[G. Morgan]

No, you misunderstand the GPS system.
The satellites all transmit at the same frequency, their digital signals are
xored before modulation with a pseudo random code, a so called 'Gold code'.
This code is different for each satellite.
Basically this is done with a shift register.
In the receiver AFTER demodulation the digital signal is run through a similar
shift register, and one by one tested for the correct Gold code (each sat has its own code).
In case of noise, you do not get an exact match against the know codes, and the number of wrong
bits is then a measure for the signal to noise ratio.
100% good signal, no bit errors.
Weak signal lots of bit errors.
http://en.wikipedia.org/wiki/GPS_signals

Do they x-mit a different freq. for military use?

 

[G. Morgan]

Yes, it presumes some prior knowledge
try this website, the guy build his own GPS stuff, and does a good job at designing too:

A homemade receiver for GPS & GLONASS satellites
http://www.s5tech.net/s53mv/navsats/theory.html

Thanks.

 

[rickman]

OK, explain to us what a dB is.

If you don't know, why should we tell you.

Rick

 

[[email protected]]

Partly,
as you want to display signal to noise for each individual satellite,
you have to identify the satellite in the signal.
For that to happen you need to detet its Gold code directly.
Now you are full circle

So in reality, you need four independent signal strength signals for
the best (geometry) four satellites

 

[[email protected]]

If you don't know, why should we tell you.

You two do make quite a pair.

 

[MrTallyman]

How would a toilet cleaner know that?

Try to tell us, since that is what *you* do for a living.

As for myself...

When I adjust the waveguide output channel of a high power satellite
transceiver, my employer is quite comfortable with my understanding of
logarithmic progression. Also when I adjust the noise source loop that
we pump it with for the test.

You are lost.

As for you...

When you adjust your mop head, be sure to bring it right up next to
your face so your crabs and fleas can hop onto it and migrate to new
horizons. Good job of spreading the vermin, you immature little retarded
characterless honorless bastard. Go back to the kook group, dumbfuck.

 

[MrTallyman]

Try to tell us, since that is what *you* do for a living.

As for myself...

When I adjust the waveguide output channel of a high power satellite
transceiver, my employer is quite comfortable with my understanding of
logarithmic progression. Also when I adjust the noise source loop that
we pump it with for the test.

You are lost.

As for you...

When you adjust your mop head, be sure to bring it right up next to
your face so your crabs and fleas can hop onto it and migrate to new

^
SOME of your crabs and fleas. We wouldn't want you to go
completely without friends.

 

[[email protected]]

?
You need 3 sats for a 2D position, and 4 sats for a 3D position solution.

Three variables require three equations; four, four. Don't forget time.

 

[mpm]

I'm asking this here in the hope someone knows something about GPS
receivers.
[snip]

Does anyone have a clue what 100% is relative to?

Well, it's not going to be SINAD since it's not an analog receiver.
Your choices are most likely Bit Error Rate (BER), though a BER of 20% would seem completely unusable to me. The other option is C/I+N but again, that's used more for analog modulation schemes (or should be, though some companies think SINAD is acceptable for "digital modulation". These people should be tarred and feathered. You could write a book about modulation schemes and their effect on BER, and I'm sure some have. I've never found dBm lookup tables to be particularly good descriptors for BER. Your mileage mayvary.

Bars and SINAD and I+N values aside, if the GPS fades too much, there's probably something wrong with the receiver or the placement of its antenna. Some antennas are powered (i.e., internal active preamps), so you might verify the batteries are fresh, or it's otherwise powered-up with the correct voltage.

Shameless plug: We use a lot of Garmin Montana GPS's (650's and 650t's)forour field work. Expensive, but perfect for what we do. Of course, as with any early adopter there were some early software glitches; like "Continuefour-thousand one-hundred fifty miles, then make a U-turn", but they squashed that bug pretty quick.

 

[josephkk]

Partly,
as you want to display signal to noise for each individual satellite,
you have to identify the satellite in the signal.
For that to happen you need to detet its Gold code directly.
Now you are full circle

Not entirely. The empherides for all the satellites is in flash and can
be updated by data from the satellites which are in turn updated by ground
support stations. Since a time solution is relatively in dependant of the
location solution and the satellites found, it can be use to get the
empherides data and thus know which satellites to look for the location
solution. Of course the receiver remembers the last location it had lock
and keeps time while powered off, the empherides is immediately useful in
most cases.

?-)