problem: my Venus GPS not getting fix

[Admin]

I'm using this Venus GPS:
http://www.sparkfun.com/commerce/product_info.php?products_id=9133

All I'm getting from it is:

$GPGSA,A,1,,,,,,,,,,,,,0.0,0.0,0.0*30
$GPRMC,120554.000,V,0000.0000,N,00000.0000,E,000.0,000.0,280606,,,N*7B
$GPVTG,000.0,T,,M,000.0,N,000.0,K,N*02
$GPGGA,120555.000,0000.0000,N,00000.0000,E,0,00,0.0,0.0,M,0.0,M,,0000*6B
$GPGSA,A,1,,,,,,,,,,,,,0.0,0.0,0.0*30
$GPRMC,120555.000,V,0000.0000,N,00000.0000,E,000.0,000.0,280606,,,N*7A
$GPVTG,000.0,T,,M,000.0,N,000.0,K,N*02
$GPGGA,120556.000,0000.0000,N,00000.0000,E,0,00,0.0,0.0,M,0.0,M,,0000*68
$GPGSA,A,1,,,,,,,,,,,,,0.0,0.0,0.0*30
$GPRMC,120556.000,V,0000.0000,N,00000.0000,E,000.0,000.0,280606,,,N*79
$GPVTG,000.0,T,,M,000.0,N,000.0,K,N*02
$GPGGA,120557.000,0000.0000,N,00000.0000,E,0,00,0.0,0.0,M,0.0,M,,0000*69
$GPGSA,A,1,,,,,,,,,,,,,0.0,0.0,0.0*30
$GPRMC,120557.000,V,0000.0000,N,00000.0000,E,000.0,000.0,280606,,,N*78
$GPVTG,000.0,T,,M,000.0,N,000.0,K,N*02
$GPGGA,120558.000,0000.0000,N,00000.0000,E,0,00,0.0,0.0,M,0.0,M,,0000*66
$GPGSA,A,1,,,,,,,,,,,,,0.0,0.0,0.0*30
$GPRMC,120558.000,V,0000.0000,N,00000.0000,E,000.0,000.0,280606,,,N*77
$GPVTG,000.0,T,,M,000.0,N,000.0,K,N*02
$GPGGA,120559.000,0000.0000,N,00000.0000,E,0,00,0.0,0.0,M,0.0,M,,0000*67
$GPGSA,A,1,,,,,,,,,,,,,0.0,0.0,0.0*30

I'm on a 2nd floor, with my duck antenna and GPS right next to a very large window. My coworker says he gets a signal in his car, so I know they aren't scrambling the signal (I'm on a Navy base right now).

I have the device powered using VBAT, and tried both 3.3V and 5V.

I also tried using the SkyTraq software for it, and the data is coming out fine (I think), its just not getting a fix.

Anyone familiar with this? I'm going to try and take out a laptop later today and try again outside, but doubt it'll work.

[Soeren]

Hi,

It took me 20..30 minutes to get a fix on my Garmin GPS when I was in Belgium last week. This was outdoors, free access to the sky and everything - I've never experienced something like this before.
Wondered if the satellites were being "maintained" or something.
Probably don't help you a bit though.

[Admin]

Yea, if I was in Europe/Asia, I'd assume just no signal. But I'm in Washington, DC, and like 3 miles away from an international airport.

I just got my new work laptop to try GPS outside, but unfortunately it looks like its about to rain. Doh!

Does the output I got at least look like its the right format? A lot of commas and what looks like hex at the end . . .

[Admin]

Problem solved, sorta.

Apparently there were no sats at the time I think. After 18 minutes I only had 2 sats, then suddenly got 6 more in the next few minutes. So the data was all correct.

Then I unpowered everything for ~5 min, then turned it all back on again, and got 8 sats in under 2 minutes. I'm using it near my big window inside.

The altitude and direction seem very fickle . . . but whatever . . .

the results:
By matching it up using Google maps, its accurate within 60 feet. By recording position history, it fluctuated by about 40 feet as it locked on. The realtime 2D RMS over 30 minutes fluctuated up to 15 feet and as low as 1 foot. The datasheet claims to be accurate under 8 feet . . .

Oh, and I'm using a regular duck antenna.

[c02cutter]

Just because you have a window available does not mean you will get signal. In the United States the satellites for positioning are predominately to the south. Most all the satellites are south for all communication in the United States. GPS's need a particular number of satellite fixes for positioning. Depending on software of the unit, it may tell you how accurate it's positioning is at a particular time. A  window can reflect the signal depending on your position to the window from GPS to satellite. GPS's are only as accurate as the the information received and the programming involved in the unit. Another point is that on occasion the satellites are updated and information needs to be downloaded to your GPS for it's programming to position correctly. Since GPS's were opened up to close to military accuracy, acquiring of satellites is a bit more complex.

[kd5kfl]

if a GPS has been off for a while, or has moved 500 or more miles since it was powered on, it has to initialize. figure out where it is, from scratch. takes time.

[Webbot]

Yea, if I was in Europe/Asia, I'd assume just no signal. But I'm in Washington, DC, and like 3 miles away from an international airport.

Us 'poor' Europeans managed to drag ourselves out of the slime a while ago. Yes - we have 'airports', 'satellites', etc - we even have 'cities' like your 'Washington DC' - but we tend to give them real names rather than acronyms ;-).  We even have windows in our rooms and on our disposable laptops! We also invented quite a lot of stuff: computers, radar, car tyres, jet aircraft etc etc. Sorry - it was just too good let it go!

I live one mile from any other house in the middle of the countryside  - I cant get a cellphone signal but my GPS gets a fix in about 40 seconds.

Does the output I got at least look like its the right format? A lot of commas and what looks like hex at the end . . .

Have sent you you links to formats. But that hex stuff at the end is a checksum. Ignore it at your peril (WebbotCLib uses it) - otherwise you may think you are in Europe one moment and, a few seconds later,  in Thailand.

Cold boot - normally takes a minute or so. Others refer to 20 minutes which is just crazy! Something to check on data sheets though.
Warm boot - the GPS retains some info 'on chip' and so if re-started before it looses the charge then it gets its fix a lot quicker.

[sdk32285]

Hi
The NMEA strings that you have look correct.
The 20minutes to acquire a good lock I have only seen when moving significant distances (CMU to S. America) or when using commercial corrections (omnistar etc...).

With no corrections the GPS speed and altitude are very coarse.

Where are you getting direction from?

Admin welcome back to the states!

[Webbot]

All that extra stuff: altitude, direction (compass), speed etc are pretty useless. Guess they are built for cars doing 60mph not a robot going at a snails pace.Think direction and speed may well be calculated from previous readings over time. So 'slow robot' = 'bad readings'. It will all be based on longitude/latitude anyway - and those reading may be a few meters out. May not sound much but if the GPS thinks you've jumped by a meter in 1 second then it will think you are going much faster than you probably are.

[sdk32285]

Hey
Also you need to be careful if you are writing the driver.
For a reported position of LLMM.mmmmm the LL is degrees and the MM.mmmm is the minutes   

[Admin]

if a GPS has been off for a while, or has moved 500 or more miles since it was powered on, it has to initialize. figure out where it is, from scratch. takes time.

I'm trying it out again this morning. It appears the Venus has a *very* slow cold start of 20+ minutes. Arg! Now I understand why there is a VBAT connection - for a battery to avoid cold starts . . .

(the datasheet says 29 seconds, but that just isn't happening)


sdk32285, welcome back to SoR!

[Webbot]

Time-to-first-fix (TTFF) is crucial with these devices.

If its any help I find the modules available from GlobaSat to be very good. http://www.easydevices.co.uk/scp/GPS_Receivers/GPS_Modules.html with a cold start TTFF of about 45 seconds.

Think that Parallax used one of these in their GPS receiver. http://www.parallax.com/tabid/768/ProductID/396/Default.aspx

I use the EM-406 module - its just a 1 inch square box (of negligible height) with a cable.

Look at the Parallex link above and its just that thing with the pink border.

The cable output connection is very small and fiddly, and the matching connector from Digitkey is SMD. I just cut off the existing plug and added a Molex connector instead - warning - the lead is only one to two inches long and so doesn't allow you many goes at getting it wrong and the cable is hardwire into the device at the other end. But it saves you the cost of paying someone else to add a mounting board. I just use velcro on the bottom of the device to stick it to my robot.

I get a TTFF of about 45 seconds and I live 'in-the-wild' - but I can still get a signal if the receiver is locked in my desk drawer. Note: the device has a built in aerial so no worries with that.

I recommend this bare module - but replacing the connector with a Molex is not for newbs or people with a faint heart !!

[Soeren]

Hi,

Yea, if I was in Europe/Asia, I'd assume just no signal. But I'm in Washington, DC, and like 3 miles away from an international airport.

And why do you think the Global Positioning System would work differently in Europe or Asia than in Washington?

Usually I get a fix very quickly with my Garmin, a bit faster than my Navman (which isn't slow either).
My point was simply that I had this strange incident where it looked like all the sattellites were switched off for a period.

I am looking forward to the European system however, it should be more up to date on both precision and coverage (probably due to a higher power transmission). And for a number of years, the navigators will be able to receive from both systems, so that will be the right time to shop

[Admin]

Webbot, I already bought 2 Venus GPS modules . . . so kinda stuck for now. But I guess I'll need to try others and compare . . .

And why do you think the Global Positioning System would work differently in Europe or Asia than in Washington?

My technical friends in Thailand all tell me the GPS signal there is exceptionally weak/bad. I'm not really sure why. Figured maybe some areas are disadvantaged in terms of sat orbits . . .

So uhhh, when will the Euro GPS ever be finished?

[kd5kfl]

http://www.kowoma.de/en/gps/orbits.htm

tell them Thais not to breathe the fumes when they burn the weeds. no reason Thailand should be different from India, Belize or Japan in re gps

[Admin]

http://www.kowoma.de/en/gps/orbits.htm

tell them Thais not to breathe the fumes when they burn the weeds. no reason Thailand should be different from India, Belize or Japan in re gps

Weeds?

I'm not saying they don't get GPS, as I've certainly used an automobile GPS over there without a signal problem. They only told me that the signal is very weak. And I've been told this from several unrelated/connected groups of people. I was planning to test out my Venus GPS over there during my next trip . . .

Anyway, I've been testing it even more, and this Venus GPS really sucks . . . the datasheet says:
Hot start: 1 Seconds
Cold Start: 29 Seconds

but I'm getting more like:
Hot start: 1-2 minutes
Cold Start: 20-30 minutes
This morning I've had it running for 1.5 hours and it still hasn't got a sat lock.

My only theory is perhaps my duck antenna is a bad choice, but the datasheet and Sparkfun both don't recommend any antenna lengths. Wikipedia says 'Civilian use signal, broadcast on the L1 frequency (1575.42 MHz)'. So using this calculator:
http://www.crompton.com/wa3dsp/hamradio/antcalc.html
It suggests 3.6" for half-wave, and 1.8" for quarter-wave. I'm using a 4.5" duck antenna. I've also tried helical antennas meant for 2400 MHz, and a few others around 4"+. Should I find a 3.6" antenna and try again?

What length antennas are you guys using?



Just as a side note, I've noticed that the red LED on the Venus blinks when there is a sat lock.

[Webbot]

I dont use any antenna at all !!

[Admin]

I dont use any antenna at all !!

not even those built in chip or PCB trace antennas?!

[Webbot]

Just one of these modules http://www.globalsat.co.uk/product_pages/product_em406.htm

Has built in antenna etc. So its just plugs into a UART header and off you go.

Check out those acquisition times !

[Admin]

Just one of these modules http://www.globalsat.co.uk/product_pages/product_em406.htm

Has built in antenna etc. So its just plugs into a UART header and off you go.

Check out those acquisition times !

Did you measure/verify those acquisition times? The Venus *claims* much better times than that . . . I'm just curious.

Also, the accuracy on your GPS is kinda poor (for my application):

10 meters, 2D RMS 5 meters, 2D RMS, WAAS enabled

I measured the Venus at RMS of 3 meters. At least when it actually has a sat lock . . .

[Admin]

Further investigating the problem, I found something interesting . . .

The Venus datasheet says:
vcc 2.8-3.6V
vbat 1.5V-6V

The sparkfun Venus schematic says vbat and vcc are NOT connected. It says there is a jumper, of which I cannot find. They are also two separate pins. I only use the vbat pin. I've occasionally connected my Venus vbat to 5V regulated, and noticed it was getting strangely too warm. Wasn't sure if that was normal, so I investigated further.

Then I looked at the sparkfun Venus board file - and guess what! vcc and vbat ARE directly connected!

Why the heck they included a vbat pin if its directly connected to the vcc AND not have an schematic, I have no idea . . .

The Venus is probably damaged, and maybe why its not working properly . . .

I'll go rant on the sparkfun site now

[madsci1016]

Admin, I am using the same Venus GPS from Sparkfun.

I have come to the decisions it is just flaky. Sometimes i'll get cold starts of seconds, or sometimes 30 minutes. On average it's around 3-5 minutes.

I am using this antenna http://www.sparkfun.com/commerce/product_info.php?products_id=177

I haven't messed with attaching a Vbat, and if i remember correctly, there is a solder bridge on the breakout board you can de-solder if you want a separate Vbat. It is located right off one of the corners of the chip itself, next to a pin header.

Beside that, i am realizing that it may not be as accurate as i need, some testing has shown on my UGV it can vary as much as 50 feet, so my GPS navigating robot will just need BIG fields to navigate until i invest in a better GPS.

I have used their software to set in flash to only output the sentence i need, which saves my Axon from having to parse or deal with all the sentences.

[Admin]

I haven't messed with attaching a Vbat, and if i remember correctly, there is a solder bridge on the breakout board you can de-solder if you want a separate Vbat. It is located right off one of the corners of the chip itself, next to a pin header.

Nope, it has two jumpers, but neither are for VBAT.


I own two Venuses, and decided to try them both at home, about 15 miles from work. I plugged in the first one, the one that I thought was damaged, and it locked on to 3 sats within 5 minutes, and another 4 sats minutes later. Cold start. It was giving me about 50 meters RMS.

Then I plugged in my second Venus, which I haven't yet used before so not damaged or anything. It sat for 1.5 hours and never managed to get more than 2 sats.

For both, I have it sitting on my windowsill with clear skies.

If the Venus datasheet is correct, I'm thinking that maybe SparkFun screwed up the PCB, perhaps adding the wrong components and thats why its working so poorly.

I'm just going to buy another GPS

[madsci1016]

Wow, so i looked closely on mine and found the Vbat jumper (mine has three jumpers) and then compared to the image on Sparkfun, and it's not there!

Here is where it is located on mine (purple circle).


Time for refund from Sparkfun for selling you and old revision?

Please, let me know if you find another GPS and are happy with it. I might want to replace my Venus sometime soon.

Edit: i just saw on your SparkFun comment about the current issues, sounded like you are powering the GPS off the 3.3V Axon bus? I tried that in the beginning, even though it should work, i was getting sporadic (GPS) resets and poor gps performance when i powered off the Axon. Now it has it's own 3.3V regulator and it hasen't reset or freaked out again. Maybe that could be an issue for you?

[Admin]

Time for refund from Sparkfun for selling you and old revision?

I wish . . . I bought mine like ~2 months ago if I remember right . . .


I'll probably just buy several and try them all out. It might take a month before I report back, however.

[madsci1016]

Cool, let me know.

Also did you catch my edit?

Edit: i just saw on your SparkFun comment about the current issues, sounded like you are powering the GPS off the 3.3V Axon bus? I tried that in the beginning, even though it should work, i was getting sporadic (GPS) resets and poor gps performance when i powered off the Axon. Now it has it's own 3.3V regulator and it hasen't reset or freaked out again. Maybe that could be an issue for you?

[Admin]

I haven't had any sporadic issues or resets with either my Axon or Axon II. My guess is that they removed/changed the capacitors on your version, resulting in those resets.

I of course also tried regulated 5V with the same problems.

[Admin]

I'm thinking of getting this module:
http://www.sparkfun.com/commerce/product_info.php?products_id=8266

Its not SparkFun designed, and has a built in antenna, so I think I can trust those specs.

Anyone tried this module and can verify?


Webbot: good thing you no longer have a required 4800bps baud rate in the WebbotLib GPS library. All the latest GPS devices seem to operate at higher baud rates. This one does 38400 at 5 Hz, for example.

[Rick Brooks]

Admin,

I've used http://www.sparkfun.com/commerce/product_info.php?products_id=8266 and have had much better luck with http://www.sparkfun.com/commerce/product_info.php?products_id=465.

Also, it seems like http://www.sparkfun.com/commerce/product_info.php?products_id=9436 is even better, but I've only been playing with it for a few weeks.  If you go with this one, get the adapter from DIY Drones http://store.diydrones.com/PhotoGallery.asp?ProductCode=BR%2D0008%2D01.

     Rick Brooks

[Webbot]

Has anyone tried http://www.parallax.com/tabid/768/ProductID/396/Default.aspx