2
\$\begingroup\$

The cheapest class of GPS receivers is what we all have inside our smartphones etc.

Then we have devices like this U-blox.

In the end there are professional GPS like this SPAN combined with INS (IMU).

The latter one is basically 1 box including support for GPS+Antenna+INS.

It's understood why the INS component is expensive (MEMS gyro etc.), how the INS correcting the GPS and same with the Antenna.

This raising the question what is the difference solely in the GPS component inside the professional expensive device versus the cheaper one? Because if it's essentially the same, one can take $50 GPS receiver and by adding Antenna get same result like buying the expensive professional GPS...

Thanks,

\$\endgroup\$
8
  • 3
    \$\begingroup\$ IMO you are comparing apples to oranges. Even if you weren't, "cheap vs expensive" is an unending debate in any field. \$\endgroup\$
    – Wesley Lee
    Mar 21, 2017 at 13:18
  • \$\begingroup\$ @WesleyLee Why apples to oranges? Inside the professional device you do have pure GPS component. \$\endgroup\$
    – michael
    Mar 21, 2017 at 13:29
  • \$\begingroup\$ 1) I disagree that smartphones have the cheapest receivers, because smartphones are used in suboptimal conditions (indoors, bad antenna) you might need a better, more expensive receiver to get decent performance. 2) the price of product does not always relate to the price of the components used. 3) Professional does not always mean "better components", it can also be same component but better antenna, more optimized software, better support. 4) what do you mean by "pure GPS component" ? Is a GPS receiver in a cheap phone not "pure" ? And if so, why ? (show facts, no assumptions). \$\endgroup\$ Mar 21, 2017 at 13:32
  • \$\begingroup\$ @FakeMoustache Inside the professional box you have {pure GPS, INS, Antenna}, in "pure GPS component" I simply refer to all the electronics related to the functionally of the GPS, and not the INS and the Antenna. \$\endgroup\$
    – michael
    Mar 21, 2017 at 13:40
  • \$\begingroup\$ @FakeMoustache Smartphone can cost $100 and the GPS is not the most important & expensive component there. ublox is $50. The professional staff can be even $30,000. Yes, price is a function of marketing & demand. However difference between $50 to $30,000 is more that that. \$\endgroup\$
    – michael
    Mar 21, 2017 at 13:43

3 Answers 3

6
\$\begingroup\$

The differences between a cheap GPS and a high end commercial GPS:

Narrow band RF front end Vs wide band front end.
A cheap GPS will only track L1 C/A which is a fairly narrow signal, a good GPS will also track the P(Y) code which is wider bandwidth.

1 bit Vs 2 bits sampling.
A low end GPS will use a 1 bit front end, a better one will normally be 2 bits.

16 MHz Vs 25/50 MHz sample rate.
Lower end systems tend to use a lower sampling rate, that's all you need for C/A code.

Single frequency / signal Vs multi-band, multi signal.
A basic GPS only tracks L1 C/A, in addition to the already mentioned L1 P(Y) code a high end one will also track signals on the L2 (L2C and P(Y)) and L5 (L5C) bands.

Dual antenna support
Some high end GPS systems will support two antennas and run an RTK like correction between the two allowing you to get good accuracy heading information even when stationary. Low end GPS you can get direction of travel accurately when moving but that doesn't tell you which way you are facing and only works well above a certain speed.

RF shielding.
GPS signals are really weak. If you mount a GPS antenna close to the receiver or other electronics then you risk noise getting into your GPS system. In the past I've spent weeks stopping the 37th harmonic of a clock signal from leaking into a GPS antenna in order to gain an extra 0.5 dB of SNR. For a cheap system no one would care about that, it only matters when you are trying to get the last little bit of performance.

Running completely stand alone and only tracking L1 a good GPS will (with a lot of assumptions, hand waving and approximations) give an RMS error that is around half that of a cheap consumer unit.

And this is before we get into the extra CPU power and investment in terms of software engineer time required to allow things like RTK corrections to improve accuracy down to 1 cm with just GPS alone, no IMU needed.

And then you have all of the extra stuff needed for GLONASS, Beidou, Galileo and all of the other GNSS things that are outside GPS that a good receiver will do.

Those NEO-7P modules from ublox are very good position accuracy for the price but are very similar in terms of hardware to the $15 base model NEO modules. The extra cost isn't for parts, its for the development effort behind the software they are running. They are still a long way off from a high end of GPS, list price for a high end GPS with all the features enabled can be over $10,000. Those $10,000 systems aren't just a Ublox with a GPS antenna put together in the same box.

\$\endgroup\$
1
  • \$\begingroup\$ Fully agree. Add semi-codeless P(Y), narrow correlators (needs high sample rate), delta-delta-DLL, coherent discriminators (all of these pushing baseband complexity), vector tracking and RINEX-output (controller processing load). \$\endgroup\$
    – Andreas
    Mar 21, 2017 at 19:19
5
\$\begingroup\$

See How many GPS channels make sense?

That long answer details how adding more channels to a reciever (at additional cost) can:

  • Speeds up satellite acquisition
  • Reduces power consumption
  • Reduces likelihood of losing a 3D fix even in urban canyons
  • Provide better sensitivity, allowing fixes in dense forests, and even in some tunnels
  • Provides better positioning accuracy

Edit: channels are one factor. A big factor seems to be the fiber-optic gyros in those other units; those are something I can't even find pricing for on the internet. But really I suspect a lot of the costs for the professional kit are non-recurring expense (NRE) of various sorts. More complex software, especially for sensor fusion. Proprietary position-enhancement techniques. Certification and compliance (I see "export compliance" on one of those feature lists, maybe there's a legal cost associated with that?)

High NRE divided across small number of units sold = very high unit cost for seemingly modest performance improvements. Not unusual when dealing with lab or test equipment.

\$\endgroup\$
3
  • \$\begingroup\$ This is it? Just the channels thing? \$\endgroup\$
    – michael
    Mar 21, 2017 at 13:41
  • 1
    \$\begingroup\$ I've added some other possible factors. \$\endgroup\$
    – pjc50
    Mar 21, 2017 at 13:57
  • 1
    \$\begingroup\$ Be careful on channel count, it depends on what you count as a channel. e.g. Say you have a high end dual band system (L1 and L2) then when tracking a satellite you will typically be running 6 correlators in two sets of 3, early, punctual, and late, for each frequency. Within the group of 3 they are tightly locked together and only a limited variation is allowed between the two sets. So is that 1 channel, 2 channels or 6 channels? The answer is that it depends who's marketing department wrote the datasheet. \$\endgroup\$
    – Andrew
    Mar 21, 2017 at 15:56
2
\$\begingroup\$

These have an accuracy down to a cm. Does your phone do that?

\Between these two models there is a huge difference.

enter image description here



Compare to this one with 555 channel, all-constellation, multi-frequency positioning solution, Built-in Wi-Fi support, and 16 GB of internal storage

enter image description here

\$\endgroup\$

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