Replies to This Discussion

Permalink Reply by Jake Stew on July 16, 2012 at 11:05pm

I never understood the difference, at least as far as the specs.  From the datasheets it looks like a draw at best.

You can get the ublox a lot cheaper than that though...

http://www.sunsky-online.com/view/257610/GPS+Receiver+u-blox+NEO-6M...

IMHO 95% of the reason people rave about ublox is the antenna and/or simply being external gives them better reception.  OTOH maybe it has better power filtering or out-of-band interference rejection.

In any case, my built-in MT unit gives me excellent performance and I'm very happy with it.

• ▶ Reply

Permalink Reply by Justin on July 16, 2012 at 11:24pm

If being external is what makes it better, then why wouldn't people go for the built in one on an external module that's half the price? https://store.diydrones.com/MediaTek_MT3329_GPS_10Hz_Adapter_Basic_...

If you've got the built in one you can answer my other question on how it's mounted to the APM2 main board?

• ▶ Reply

Permalink Reply by Jake Stew on July 16, 2012 at 11:52pm

The GPS, mag, and SD card slot are on a little daughter board mounted on top of the APM2.  There are 6 semi-randomly placed header pins from the APM that go through the daughter board, they look unsoldered and are just for support I think.  There is a 2x6 set of header pins that come from the APM and go through the daughterboard through a passthrough type header connector.

If you look through the pictures in the manual that description should make sense.

• ▶ Reply

3D RoboticsPermalink Reply by Chris Anderson on July 16, 2012 at 11:28pm

Antenna four times the area for better lock and multipath resistance. Battery for near-instant lock (cold-start only needed once; and never again). Next-gen chipset for more accurate position readings.

• ▶ Reply

Permalink Reply by Jake Stew on July 17, 2012 at 12:01am

I get the antenna and battery, but everything claims multipath resistance and anti-jamming technology.  They both throw around techno-gibberish like "indoor GPS" and "supersense technology".

To me the rest really means nothing.  If either had much of an advantage you would think it would show up in the specs, right?

They both seem to be almost exactly the same in receive sensitivity, TTFF, etc..  It's not like one or the other has even a slight edge in everything, it's totally back and forth.  They're both also the latest chipsets from GPS market leaders.

It would be nice to see a real-world comparison to blow through all the marketing bull spit.

• ▶ Reply

Permalink Reply by Justin on July 17, 2012 at 12:19am

I agree, I'd love to see some unbiased comparisons of the two and see how they really compare rather than 'marketing department comparisons'!

Also does the 5hz/10hz make a big difference for a UAV?  Obviously to us more would appear better as it would be twice as frequent and therefore appear 'smoother', but I'm unsure of the difference in a UAV.

• ▶ Reply

Permalink Reply by Dr Mike Black on July 17, 2012 at 1:16am

I dont know if this helps...

As many have been having prob, me included with the MTK units posted elsewhere...

I personally use the ublox, as it deos exactly what it says on the tin, filtering is good, lock is fast..

however, as i had a few ublox units sitting around with broken antenna's i thought that a good test to see, whether the quality of the unit is based purely on the helix (earlier LEA5h units) or the soft/hardware.

It might look a little odd, but i removed the patch from the MTK and soldered it directly onto the ublox.

In essence, done exactly what 3dr has now done with the ublox unit.

I have to say that the patch is very good...very fast lock, strong signals, not quite as good at all angles (roll and pitch in our application, this is fairly obvious as its a patch not a helix), but for those who have broken ublox units, this, at least to me, has proved that the issues experienced, is down to the unit not its ability to recieve data.

I would def buy this unit (assuming its as good, which im sure it is) to replace my ageing, dying GPS units)

Just to be clear about the bias issue, i really dont care what unit we use as long as it works, these conclusions are based upon my personal testing, clearly results are subjective to the individual user.

I will fly the unit today to get 'real world' data

Pics below

Attachments:

• ublox_with_MTK_antenna (1).jpg, 128 KB
• ublox_with_MTK_antenna (2).jpg, 115 KB
• ublox_with_MTK_antenna (4).jpg, 89 KB

• ▶ Reply

Permalink Reply by Jake Stew on July 17, 2012 at 8:23am

Thanks for the results.  Looking forward to seeing what else you come up with.

You seem to have more than your fair share of trouble with the MT units.  I've had some power cleanliness issues with one of my MT modules.  I've also seen that Fastrax makes a MT module with a saw filter in it, and claim it is useful for applications near transmitters or in noisy environments..

So I'm wondering if that isn't at least part of the issue.  You might be in a location with high interference near 1G or maybe one of your transmitters is a bit noisy.

• ▶ Reply

Permalink Reply by Dr Mike Black on July 18, 2012 at 1:31am

Well, gentlemen,

Was an interesting day yesterday, teh MTK stand alone unit i was using for testing with decide d to 'get with the program'

you have to picture the scene here....

on one of my benches we have a AP1, APM2 & a Remozibi OSD

AP1 has a un-touched ublox LEA5h

APM2 has the ublox with MTK patch attached

OSD has stand alone MTK

All 3 locked up succesfully,

STD ublox in 90 secs from cold

Hybrid Ublox 40-120 secs from cold

MTK 230 secs from cold

______________________________

STD saw 7 sats all day, re-boot after reboot, after reboot, warm start times from 20-60 secs

Hybrid saw 7-9 sats "     "               "                     "        "                           "  30-120 secs

MTK 7-9 sats             "                       "                             "                              50-200 secs, however it would drop out from time to time, (not awful, but still did it 3 times over a 7 hour period)

Now none of these lock times are unacceptable, but what i dont get is...

I have not changed anything.....this is the best i had teh MTK working since owning it.

same PSU, same position, same location!....

I have to agree with Jake here....there is some magical mystery gooo affecting the results, but fror the life of me i cannot nail it down.

Also in a dramatic demonstration of idiocy i managed to blow up the remzibi OSD, grrrrrrrr, nice little hole in the processor, something about 'assumption being the mother of all foulups!'

Thoughts anyone?

• ▶ Reply

Permalink Reply by Dr Mike Black on July 18, 2012 at 12:24pm

HI all,

We in a continueing effort to work out the changes....

I have now taken the patch off and old Locosys and an old 406

Soldered them onto the ublox adn the readings are the highest i have seen (fro my location)

Locosys patch is the yellow one, adn its the best so far.

Tomorrow ill solder one onto the MTK unit and see if the behavious are better.

compared to teh STD ublox the readings are up to 45, (never over 30 normally)

attached are the pics of the readings and the conversions

Something worth noting, is that teh HDOP is .9 on average, (lowest i have seen)

and the 2d acc is 3-5m..again the lowest i have seen...lock sub 30 secs on all occasions..

So maybe were onto something here..

will be very interesting to see the readings from the MTK with a diff patch.

as standard the HDOP is 1.3+ (took ages to get that low)

Anyway i dont know if anyone is reading this or interested, but these are my results so far.

Kind regards,

Mike.

• ▶ Reply

Permalink Reply by Jake Stew on July 18, 2012 at 1:14pm

Nice work.  I'm sure people are reading this thread a ton.

I've dealt with the "mystery goo" a bit and it seemed to me that power supply was the biggest factor, but there's really no way to account for radio interference.

I have a sneaking suspicion that radio interference is also a pretty big factor.  This could really bite me in the butt since I live in a pretty small town area.  Flying into interference could really sink my ship if I let a false sense of confidence set in.

You could try some cell phones near the units.  I know they throw a ton of interference.  Usually with clock radios or other small cheap radios cell phones will make a series of clicks when about to receive calls and at other random times.

Maybe you can use a clock radio to monitor the cell interference then see what effect it has on the GPS.  Hopefully you've encountered this interference before so I don't sound like I've gone mad.

• ▶ Reply