invalid dates on gps tracking systems

I also saw few devices with the same problem. I don't remember about the year, but the time was 00:00.
If you are talking about 'A' and 'V', probably, you use $GPRMC sentence.
Once I worked with a GPS receiver that puts 'A' into an empty $GPRMS message - there were no coordinates at all. It was something like $GPRMS,,A,,,,,,,N*46. I think, your application may have a bug for such case (my application had it).
Actually, this message comes each second.
I don't know for sure, I think, that this time problem we see when we do not have GPS fix.

Hi

The gps units in question can get a correct date/time. However the times when their date and time is incorrect it is normally 2005.

Jaime, would your answer still apply given that the units can get the correct date/time?

I believe they get their date/time from the gps system itself and not from their own hardware. I know you can't speak for the specific hardware i use, but if this is unlikely,please let me know.

thanks

This is really a strange situation, I have read about GPS that need to be updated to provide a more accurate algorithm every 5 years, but this is more applicable for aviation or naval devices.
GPS uses the time to calculate but not necessarily the date, if your GPS has a backup battery which is already lasted, then this can be causing conflicts with date. Are you losing also other information like waypoints?

Hi

The gps units are installed in vehicles. They are programmed to poll every minute from when they detect the ignition is switched on. Sometimes when an ignition is turned on and the unit starts polling data it can take 3-5 minutes to produce valid gps data. The units know if their data is valid because they include a status code of A(valid) and v(valid). I have been looking at what is sent when the data is invalid. Sometimes the date and time is correct but the gps data is wrong in one of 2 ways
a) not in the uk which is where i am based
b) lat/lon is 0.0000.

Sometimes the date/time is incorrect. Whenever the date/time is incorrect, lat and lon are always 0.0000. The incorrect date is almost always january 2005. I thought this must mean something.

thanks

well, if lat/lon is 0.00 then date won't be correct. GPS algorithm relies on a time-based algorithm, if it doesn't show correct position, it is because it doesn't have accurate time info from satellites, during warm up (3-5 minutes is ok for an old or cheap GPS). So basically GPS date time info is blanked at that time.

You are talking about Qualcomm GPS?
So you have anpty $GPRMC messages with 'A' (ID:25387010). You need to skip them.

Hi

No I'm not talking about Qualcomm.

I don't think I am making myself very clear.  I understand that the dates and gps information are not valid because the unit has not received accurate data from the satellites. That wasn't my initial question. My question was, why, when the date is invalid, is it almost always january 2005. It seemed both odd and interesting to me that the date was jan 2005. That's all i am asking

>>My question was, why, when the date is invalid, is it almost always january 2005. It seemed both odd and interesting to me that the date was jan 2005
This is a particular design behaviour. Your GPS should be designed to contain a date of minimum 1/1/2005, so when date is blank it shows the minimum.
Other GPSs can show 1/1/1970 (as Unix's minium date) or maybe 1/1/2000

Hi

I don't think my hardware uses the GPRMC sentence but it is very similar.

Here is an example string

$ID,354879020144852,IOP,I,0x10,GPSEX,A,D,140909,T,140413,La,52.97056,N,Lo,2.18173,W,H,182.30,V,8.2,DD,316311,*

I wrote the application that uses the software so there is no driver. It must be a default value within the receiver. You'd think their technical support would know things like this wouldn't you