GPS Dropping Out

[Gypsy23]

Claire & Dave
Just to add to the tests, here is my experience over the past 12-18hrs. Since my last post I have loaded the latest Prolific driver and updated Ocpn to 2.21124. Firstly Claire, I am not trying to undermine your observations, just trying to add info to the diagnosic efforts.

As described in a previous post my set up is a Fujitsu S7111 Centrino Duo W7 Pro (32bit). The GPS is a USB attached BU353 and the driver is 3.3.11 152. I ran it for more than 18 hours, mainly as a background program whilst I work. I look every 15mins or so to see if the red boat is still showing. I have not seen it turn grey in all that time.
During this 18 hr period, my PC has dropped to 'sleep' mode on numerous occasions (6hrs when I slept too last night) and Ocpn always revives with the red myship. When the PC is asleep the GPS is still powered and shows a flashing red LED to indicate it has a fix.
So what is different here? Is Ocpn dropping GPS but I am not observing it? My experience is that Ocpn will not automatically return to red if the GPS stops and then returns so I am expecting any failure of GPS to result in a grey myship which will stay grey until I reset Ocpn or the com port.
If I remove the GPS/USB myship drops to grey after a couple of seconds but does NOT automatically resume to red when the GPS is hot-plugged back, even though I see the GPS LED flashing to show it has a fix. In a previous test after a GPS unplug-replug I restarted Ocpn and it recovered to the red myship immediately. This morning I did the same unplug-replug. Same result, no automatic recovery, but this time I went to the toolbox, selected NoGPS, exited toolbox, returned to toolbox and selected the normal Com5 port then exited and the GPS resumed OK without needing to restart Ocpn.
Is there a hardware difference in PC's? It is hard to imagine a hardware problem because your other apps seem to be happy with the data stream.
Is it the Prolific driver. Same response regarding as to hardware above.
Is it the difference in 32/64 bit? This seems to be the main difference but again hard to imagine why Ocpn is affected and not other apps.

Sorry to say, I don't have answers but I hope my test results will help someone else come up with a result.
Ray

[hellosailor]

" When the PC is asleep the GPS is still powered "
You will find that on some computers, the USB power is shut off when the computer goes into any powersaving state. On others, the USB power is suppled all the time.

Usually this is controlled by a BIOS setup selection.

If power to your USB GPS is being shut off during suspends, perhaps it is not resuming properly afterwards, i.e. perhaps the GPS is powering back up with different parameters, different default baud rate, something like that is the issue and setting "power USB always" will solve the problem?

[Gypsy23]

Thanks Hello Sailor, you are right, there are differences in the way PCs handle "sleep". In my case the power IS NOT turned off in Sleep and I DON'T have a problem when it is resumed. It doesn't appear to be part of the problem for Claire either but I mentioned it as one of my tests for the sake of interest to others. Further to the Sleep observation, I can also put the PC into Hibernate and it will recover correctly with the GPS working even though the power to the GPS is turned off during Hibernate.

My earlier testing was with Sleep and Hibernate disabled as when I use Ocpn in real-life I would not want it to Sleep, I would want to be sure it would be available at all times. On the Ubuntu version which I intend to use, I have disabled Sleep but I have set the PC to dim the screen when there is no keyboard activity. That should keep power consumption down but availability up.
What I can't understand is why Claire (and others) has this frustration of GPS dropping when I don't see it on my system so I am just trying to give as much info I can about what I see. Hopefully someone will draw the pieces together.
Ray

[djmarchand]

I am the original poster to this thread. For the last 3-4 days I have been communicating with USGlobalSat about this problem on their support website. If anyone is interested it is at USGlobalSat GPS Forums - Index. Go to the second forum group and then click on the thread "Gps drops out with navigation program"

After much back and forth confirming silly stuff like drivers, etc, USGlobalSat finally gave me a link to a firmware upgrade that presumably would update my BU-353 firmware. After downloading the stuff and perusing it, I have given up. It is incomprehensible to me.

If someone can make use of this, please post on that thread and ask for the updater. Give them your email address to get it and give it a try.

I remain convinced that this is a generic problem for Win 7 systems. FWIW the GPS worked fine with a Vista system.

David

[djmarchand]

Well having given up on the BU-353, I bought a TranSystem GM-2. It is very similar to the BU-353 and costs the same, but uses a new GPS chipset and a Silicon Laboratories usb/uart chip and driver.

After fussing around a bit with drivers and settings, it works. I ran it for 2-1/2 hours with OpnCPN and it stayed connected. Alas it couldn't recover from unplugging and replugging, but closing and reopening OpnCPN got it working again quickly.

So, if you are running Win 7, this is the GPS for you.

David

[bdbcat]

David....

Thanks for the report. Sorry you had to go through all this round-about with Win7, but I am glad you have found a workable solution.

Hot plugging of USB devices is a persistent problem with cross-platform applications such as OpenCPN. Each O/S has its own way of notifying the application of USB "events", and each must be dealt with separately.

I hope to build a stable of physical GPS devices on-hand to allow testing all of the permutations available here, or at least until the numbers get totally outrageous. USB hot plugging will certainly be on the list of future improvements for OpenCPN

Thanks again for your persistence on this issue, and your loyalty to OpenCPN.

Dave

[hanschristian38]

I am following this thread with interest as I have been rebooting throughout the season. My earlier post is below. I never had the problem on XP, only on Windows 7 and I had both going side by side.

Thanks again Dave -

[kaptainkriz]

This is working for me on a ACER Aspire One with Win7 Starter:
XPort

It's a port splitter that works pretty well and is donateware.

[bvimatelot]

Good thread! No-one has really mentioned garmin. Here's my problem. Garmin GPS 18 usb (hockey puck) worked on previous version of Open Cpn but does not on latest version. I've changed nothing!

Also, used to work on previous version, my garmin handheld gps 60. This was connected by serial to usb (prolific - and yes, the driver is up to date!) On COM 4. And I mean, like, for days on end! Now with latest version of OC, it drops out after 20 minutes or so.

For the Hockey puck - OC's tools/gps setting is "GARMIN" for the 60, its "COM4".

I've got rid of the "thinks its a mouse" syndrome which occasionally occured, and have used probably all the combinations of switching usb ports, booting up with gps connected, then without gps connected, opening up OC with gps connected then not connected etc.

Its not the end of the world because (even after the frequent bsod) I can always reboot and start again but its such a shame when it all worked so perfectly before!!

Tony - on a Windows 7 64 bit machine. Oh yes - Xport looks great but is only for 321 bit machines1

[kaptainkriz]

Yes, that's a bummer. It also fixes port hot swap and resume after sleep.

Originally Posted by bvimatelot

Good thread! No-one has really mentioned
Oh yes - Xport looks great but is only for 321 bit machines1

[donradcliffe]

Just of word of warning--I was using the BU-353 on an Acer netbook with W7. Couldn't get it to work until I tried Xport. Worked fine with no dropouts, but 2 days ago, the admiral closed the cover of the netbook, then pulled the USB cord out for the GPS. The result was that the computer would not reboot--the operating system was corrupted, and even the hidden backup copy of W7 could not be installed.

Acer will reload W7 if I send it in as the computer is only 2 months old, but I am leaving the country in a few days, and I decided to have the local computer store in Key West load XP into the computer instead. It is only supposed to cost $100 and should get rid of the other W7 issues I have been struggling with...particularly the lack of a working driver for the Alfa wifi antenna.

[bvimatelot]

Ouch!! That sucks!! I'm falling out of love with Win 7 a little bit more every day....I feel a divorce coming soon! But who to re-marry eh?

[fgd3]

Originally Posted by bvimatelot

Ouch!! That sucks!! I'm falling out of love with Win 7 a little bit more every day....I feel a divorce coming soon! But who to re-marry eh?

Looking at how much better open source software is than commercial software--both in terms of features and support--the obvious answer would seem to be an open source operating system.

Fabbian

[bvimatelot]

I'm sure you're right!! But long words like "ubuntu" and short words like "Linux" frighten me a little bit......oh well!

[hellosailor]

Don-
Before you go to that computer store, ask Acer if there are XP hardware drivers for your W7 machine. It may "run" but never run properly. By this, I mean that NT6 (Vista) and Win7) both provide hardware drivers that can do things like powe down the audio system, power down the drive system, saving significant amounts of power by directly controlling hardware in ways that XP simply cannot do, because XP was never written to do them and the XP-vintage hardware didn't have that granularity.

XP could be a honey trap: Looks good, regrets later.

ACER should be able to recommend an authorized tech would can do the rebrick, or send you discs to do it with.

BVI- You might want to check with Garmin support. There are a number of older "hockey puck" GPSes that were intended to save poewr by putting themselves to sleep when there was no host comunications happening. I don't recall the models but the workaround was specifically to apply a small resistor across two of the serial communication lines, so they would see a low-voltage "stay awake" signal.

That applied to serial wired devices, but the larger issue of GPS pucks thinking they can go to sleep, and not having an obvious way to wake them, has been around for at least a decade. Someone at Garmin may have a very simple answer for that, a command the send, or a resistor to add.