Old Philips Navigator Mk9 DGPS date problem

[Hydrozoan]

I use a laptop and a GPS ‘puck’ for navigation, but have an old, fixed Philips Navigator Mk9 DGPS as backup, and to supply position to the DSC radio. The Philips takes a little longer to pick up satellites, but is otherwise generally reliable (I’ve changed its button ‘back-up’ battery several times). However, I have just noticed that it holds and supplies to the VHF the right time (in GMT) and day of the week, but the wrong date – today it was providing ‘Sep 10’ – and there appears to be no way of altering the date, which I presume it gets from the GPS system anyway.

Has anybody encountered the problem and knows of a solution, or must I get a new fixed GPS set? There does not seem to be much choice if one does not want a plotter - and in any case I’d prefer not to have to fill the hole left by the Philips if there is another solution.

 

[earlybird]

You may well already know, but on my old Mk. 8 a "power-up reset" was achieved by switching on whilst holding the "E" button. This was recommended for position or time errors. Wouldn't hold my breath though!

 

[lpdsn]

I use a laptop and a GPS ‘puck’ for navigation, but have an old, fixed Philips Navigator Mk9 DGPS as backup, and to supply position to the DSC radio. The Philips takes a little longer to pick up satellites, but is otherwise generally reliable (I’ve changed its button ‘back-up’ battery several times). However, I have just noticed that it holds and supplies to the VHF the right time (in GMT) and day of the week, but the wrong date – today it was providing ‘Sep 10’ – and there appears to be no way of altering the date, which I presume it gets from the GPS system anyway.

Has anybody encountered the problem and knows of a solution, or must I get a new fixed GPS set? There does not seem to be much choice if one does not want a plotter - and in any case I’d prefer not to have to fill the hole left by the Philips if there is another solution.

Do you know if it is consistently wrong across sentences? ZDA and RMC both contain the time and date but the dates are in different formats. GGA & GLL just have time.

Of course I'm taking a guess at what sentences it outputs but it would be handy if you could capture them and see what's in there.

It will get the date from the GPS signal and so there may be a risk of it being a few seconds out if it doesn't correctly account for UTC leap seconds, but six months is a bit much.

 

[Hydrozoan]

You may well already know, but on my old Mk. 8 a "power-up reset" was achieved by switching on whilst holding the "E" button. This was recommended for position or time errors. Wouldn't hold my breath though!

Yes, but thanks for the reminder: the Mk9 manual refers to that reset and I tried it – but I perhaps did not quite synchronize my switch-on and “E” pressing, so I will try again.

Do you know if it is consistently wrong across sentences? ZDA and RMC both contain the time and date but the dates are in different formats. GGA & GLL just have time.

Of course I'm taking a guess at what sentences it outputs but it would be handy if you could capture them and see what's in there.

It will get the date from the GPS signal and so there may be a risk of it being a few seconds out if it doesn't correctly account for UTC leap seconds, but six months is a bit much.

All I yet know is that the DSC radio display and the Philips which feeds the DSC consistently show the right position (the same as the PC GPS puck, for all practical purposes) and time, but the same incorrect date.

I have a standalone GPS test program on the PC which shows the sentences (as well as the same feature in the PC navigation software itself). I will try outputting the Philips signal to it*, if another attempt at a "power-up reset" (see above) does not work.


My thanks to you both: the Philips must be about 24 years old so I cannot complain!


*PS I say I’ll do that, but realize I don’t know how – presumably I’d need to take the Philips output to the DSC and connect it instead to the PC via a USB plug? I’ve never wired one of the latter ...

 

[lpdsn]

My thanks to you both: the Philips must be about 24 years old so I cannot complain!

*PS I say I’ll do that, but realize I don’t know how – presumably I’d need to take the Philips output to the DSC and connect it instead to the PC via a USB plug? I’ve never wired one of the latter ...

I've only ever sailed on one boat with a Philips GPS, and that was about 20 years ago, but I thought it was pretty good. Philips were very good at producing good kit then failing to exploit it commercially.

NMEA0183 is either RS232 or RS422 depending upon the vintage. It won't work wiring it directly into USB but if you have something that interfaces NMEA0183 to USB you should be able to pick it up on the PC easily. OpenCPN which is free has a useful NMEA debug window.

 

[Hydrozoan]

I've only ever sailed on one boat with a Philips GPS, and that was about 20 years ago, but I thought it was pretty good. Philips were very good at producing good kit then failing to exploit it commercially.

NMEA0183 is either RS232 or RS422 depending upon the vintage. It won't work wiring it directly into USB but if you have something that interfaces NMEA0183 to USB you should be able to pick it up on the PC easily. OpenCPN which is free has a useful NMEA debug window.

Thanks again for that information. Yes, the Philips has been pretty good, and ISTR that they didn't stay long in the GPS market despite being in it quite early.

 

[mcetap]

This is a know problem with this unit, and might not be solvable without new software on the GPS unit (which is unlikely to be available). Your problem is caused by the GPS Week Rollover problem. Every 1024 weeks, starting on Sunday 6, 1980, the GPS "week number", transmitted by GPS satellites, rolls over from 1023 to 0, confusing GPS older receivers. This first rollover occured in august 1999 and the next one will occur in april 2019. It's comparable to the Year 2000 problem and a real design flaw in the GPS system.

See also http://www.npl.co.uk/reference/faqs/when-and-what-is-the-gps-week-rollover-problem-(faq-time)

 

[Hydrozoan]

This is a know problem with this unit, and might not be solvable without new software on the GPS unit (which is unlikely to be available). Your problem is caused by the GPS Week Rollover problem. Every 1024 weeks, starting on Sunday 6, 1980, the GPS "week number", transmitted by GPS satellites, rolls over from 1023 to 0, confusing GPS older receivers. This first rollover occured in august 1999 and the next one will occur in april 2019. It's comparable to the Year 2000 problem and a real design flaw in the GPS system.

See also http://www.npl.co.uk/reference/faqs/when-and-what-is-the-gps-week-rollover-problem-(faq-time)

Thank you very much – I had wondered if it might be something like that, and analogous to the Y2K issue. I doubt if any software fix is possible, but AFAIK the correct position is given nonetheless.

Which raises the question: does the date passed to the DSC matter, if the position and time are correct?

 

[lpdsn]

This is a know problem with this unit, and might not be solvable without new software on the GPS unit (which is unlikely to be available). Your problem is caused by the GPS Week Rollover problem. Every 1024 weeks, starting on Sunday 6, 1980, the GPS "week number", transmitted by GPS satellites, rolls over from 1023 to 0, confusing GPS older receivers. This first rollover occured in august 1999 and the next one will occur in april 2019. It's comparable to the Year 2000 problem and a real design flaw in the GPS system.

See also http://www.npl.co.uk/reference/faqs/when-and-what-is-the-gps-week-rollover-problem-(faq-time)

Would that not have caused the problem to appear at the first rollover rather than this year? What year is it on your GPS, Hydrozoan?

 

[Hydrozoan]

Would that not have caused the problem to appear at the first rollover rather than this year? What year is it on your GPS, Hydrozoan?

I'm sorry to say that it may have been like this for years, as I probably just noted that the DSC had the right position and looked no further.

I'm no longer on the boat, so can't be sure but one date held in the Philips (but not that routinely displayed on it or passed to the DSC) is IIRC one in 1994 - which I took to be manufacture or first installation as the boat was launched in 1994. The displayed date yesterday was simply 'Sep 10' (which is rather ambiguous, until I get back on board and see if it changes daily).

On the bright side, the Admiral says she has every faith that the DSC system would ignore it and apply its own 'timestamp' in an emergency!

 

[Vegable]

I've got this as well. It was ok on initial switch on around a week last wednesday then when it locked on the date changed and it was some date in 1980 but I can't recall what date it was. The time was ok though and this was the first switch on this year. The only reason I've still got it is that it's going to leave a big gap if I remove it. It is a nice bit of kit though. I'm back at home now and was hoping it'll reset itself when I next switch it on but from what's written above that's not going to happen.
Thanks for the info.
Mike

 

[LittleSister]

Your problem is caused by the GPS Week Rollover problem. Every 1024 weeks, starting on Sunday 6, 1980, the GPS "week number", transmitted by GPS satellites, rolls over from 1023 to 0, confusing GPS older receivers. This first rollover occured in august 1999 and the next one will occur in april 2019.

So if you wait long enough it will eventually get itself back in sync!

 

[mcetap]

Would that not have caused the problem to appear at the first rollover rather than this year? What year is it on your GPS, Hydrozoan?

It would probably have worked with an offset from the date / GPS week number it was first manufactured. Still, this would limit it to 19 years after the unit was made, at most.

 

[Hydrozoan]

It would probably have worked with an offset from the date / GPS week number it was first manufactured. Still, this would limit it to 19 years after the unit was made, at most.

It still seems to ‘work’, in terms of very close position agreement (for all practical purposes) with the PC GPS puck, and the time and day of the week.