Anyone's Infill G4 GPS not resuming from hibernation? Pls help!

**Carz** · 03-25-2009, 05:37 AM

Hi Curiosity, thanks for taking your time out for me.

Originally Posted by Curiosity

When selecting devices in the list, just select one and not a hub. Try it for a few days, and if it works better leave it, if it's worse deselect it.

At the moment, I'm not getting consistent test results with the com1 gps resuming correctly when isolating different variables i.e. it fails more often than it succeed.

If when I'm using XPort to split the ports for other usage, I've not selected any usb devices in XPort as I don't think I'm having any of them misbehaving when resuming.

The next time the com1 gps resumed correctly out of hibernation, I'm not going to change anything (while taking note of other programs that were running before hibernation and if they resume correctly, including the usb devices I have) and let it hibernate the same way for a few days and runs to see if there's any consistency.

If the answer to my previous qns if other usb devices could prevent the com1 from resuming correctly is yes they do then I suppose I'll have even a harder time to troubleshoot as none of the usb devices seems to misbehave. If it comes down to this, then I'll take note of your advise above.

I still don't understand why you are setting everything to 9600 when the GPS is sending at 4800. It's just causing more problems.

Following your previous explanation, XPort does detects the com1 as 9600 when the com1 is working correctly. Does that means the baud rate settings is correct at 9600?

I can't recall manually setting com1 baud rate to 9600 as it should already be the default. Perhaps in CF I did as the default could have been 4800.

Do I still have to set them to 4800 both on com1 and CF despite both G4 and CF manuals state to use 9600?

Will a opened GPS appz be it CF Navigator or Garmin caused any issues if left opened during hibernation and susequently resume? Or it depends whether the baud rate are set correctly in these appz? If it does affect, I could disable CF Navigator or set it to 4800 as advised and close Garmin (cos Garmin does not have baud rate settings) before hibernating but this makes using a car pc so painful. There were times during troubleshooting the com1 gps resumed correctly despite these appz were opened and appz obviously carried on working where it left off.

Wonder if the following screen shot gives you any clues. As you can see when the com1 gps port does not resume correctly, you'll see garbage/weird characters in the status bar. Hitting scan does nothing as it will say no gps found.

Changing ACPI in the BOIS without reinstalling Windows can cause sleep and hibernate to not work properly. You should at least update your kernel in Device Manager, under "Computer". You should be using USB 2.0 and ACPI enabled.

My vendor previously use the default bios cfg so I assume it was shipped with the OS preinstalled this way. I did read newer OS (including WinXP) will still turn on ACPI even if it's off in bios. I did try with it on but overlook checking the device manager ACPI. I appreciate very much that you have brought up stuffs I need a second look. Wonder where the drivers to update ACPI comes from?

As far as USB, can you list all the devices plugged directly into the PC and how many hubs? What's plugged into the hub(s)? Are they powered?[/QUOTE]

The only device I have plugged in directly to 1 of the 2 rear usb ports was the keyboard which I've since plugged out and now plug in to the front port only when needed as the rear port now is dedicated to my Touch HD Mobile PC as it had issue going through a hub. Even then as part of my troubleshooting, I do not plug in the Touch HD. The other rear port is connected to a quality Berklin usb hub (unpowered at the moment) located at the back of my boot. Connected to this hub is the Huawei E270 USB modem and Asus Wireless G Adaptor (combined 500mA+300mA if I'm not wrong and though appear drawing more than usual 500mA per port available but everything on here seem to work fine) which so far does not appear to have any resuming issues. This hub was previously powered by either drawing power from one of the internal unused usb port of the previous Hi-info N5 car pc or combined in such way from the battery as well (not sure of the details as it was done by my dealer) but he say G4 is much more stable and no longer need such a mod.

Again, thank you so much for assisting which I know does take alot of your time.

**Curiosity** · 03-25-2009, 07:56 AM

I don't think any USB devices would be causing the problem. It's hard to imagine that hub working well though. That's about 400mA over the limit, but then who knows the true current draw. Why fix it if it's working, right?

As far as the GPS, if it's working at 9600, keep it there. Sorry, I had assumed it was 4800 as that's the defaut for most GPS receivers. I'm not sure if it's worth messing with at this point. If it was working fine and now it's not, then there may not be anything you can do. I think you should focus on the BU-353 and see if you can get that working. Using XPort will help with hibernation.

The drivers for ACPI are on the Windows XP CD. Just insert the CD and right click on the one in Device Manager under Computer. If you have ACPI enabled, select "ACPI Uniprocesseor PC" or Multiprocessor, whatever you have.

**Carz** · 03-25-2009, 09:43 PM

Originally Posted by Curiosity

If it was working fine and now it's not, then there may not be anything you can do. I think you should focus on the BU-353 and see if you can get that working. Using XPort will help with hibernation.

Sorry if my long post lost you. I'll try my best to summarize it here.

I bought my car new 21mths ago, G4 11mths ago and started using CF and hibernation 5mths ago.

Like many users, I notice overtime, the battery does not support G4 on standby long enough till the next day I start my car (maybe more than 12hrs) i.e. power gets cut off to the G4 due to battery drain.

However, other than that I do not notice anything's wrong with the battery i.e. cranking time to start engine remains short and normal. Even if it's time for me to change battery (I will soon on the next major servicing), keeping the pc on standby for extended time will eventually still wears new battery faster than it should.

It is a major hassle the start of each day waiting for the G4 to do a fresh boot after the power get cuts off. If I remember correctly, even after a fresh boot when power is cut off, the com1 gps does not initialized properly even though com1 is visible in device manager with no error. It seems it was because XP did not go through a proper shutdown when power was cut off as restarting XP does not help. Only shutting down XP and then powering up again will the com1 gps initialized properly... double the hassle, double whammy as essentially it means booting up twice!

Even if I remember incorrectly and that the com1 gps does indeed initialized on the 1st boot after the power is cut off, it is still not good enough for a car pc that is meant to take over the oem ICE and what's not.

To get around the long bootup hassle at the start of each day (and if you include saving the battery), I started to use hibernation. However, as I mentioned, the com1 gps does not resume from hibernation (again even though com1 is visible in device manager with no error).

The above behaviours were discovered and noted ever since the battery could not support a longer suspend time and ever since I started using hibernation so I cannot say the com1 gps was working fine and now it's not.

Also, I've reasons to believe the com1 gps is not faulty as the following are consistent 10 out of 10 times:

I don't use hibernate during the day. So if I get back to the car before the power cuts off, the com1 gps resume properly from standby.

If G4 was shut down manually which also means properly, the com1 gps will initialised properly on the next boot.

During troubleshooting, if I were to hibernate and resume immediately after the hibernation process is completed, the com1 gps port resume properly. However, I do not know how "long" hibernation must be before resume becomes a problem but based on my usual schedule the next day I start my car, it will be a problem.

Whenever the com1 gps does not resume from hibernation the next day, if I were to put XP back to suspend and then resume it from G4 power button as soon it finish suspend, the com1 gps resume properly. It's still not ideal but at the moment the quickest workaround that allows me to start using the car pc asap instead of a fresh boot.

I can't emphasize enough the above 4 pts are consistent.

Therefore, the only real issue is the com1 gps failing to resume properly from "long" hibernation.

As you said earlier in your 1st reply,

There's problem that could drive you crazy. Being able to sleep/wake and get it working again would mean that it can reset without shutdown or unplugging, so some kind of software solution should work. Unfortunately XPort doesn't reset real COM ports. I'll see if I can add those to the list though.

I think if XPort can reset real com port, it should resolve the problem. If it's not asking too much, I would really love to see you add this feature at your own leisure time.

The drivers for ACPI are on the Windows XP CD. Just insert the CD and right click on the one in Device Manager under Computer. If you have ACPI enabled, select "ACPI Uniprocesseor PC" or Multiprocessor, whatever you have.

I checked that ACPI Uniprocessor drivers were already installed so obviously XP was installed with ACPI enabled. Nevertheless, I've re-enabled ACPI 2.0 and tried to let XP find newer drivers but was prompted already the latest drivers were installed. So I instead chose from the list Advanced Configuration and Power Interface (ACPI) PC rebooted. Don't think it'll make any difference but shall monitor in any case in the mean time.

Looking forward to your reply on the possibility of adding reset of real com port to XPort.

**Curiosity** · 03-26-2009, 02:39 AM

Yeah, I think hibernate is the best option for most computers except laptops or special timers. In the long run it will save money on battery replacement.

Hibernate cuts power while standby doesn't, so I'm thinking that the GPS retains it's memory for a very short time and after that time something is lost. I don't think it will help to restart COM1 though. You can test it by exiting all programs including xport, going to Device Manager, Ports, right-click COM1, disable, then enable, and see if it works. If it does, DevCon may be another method to look into, as it handles all types of drivers. I don't think it will work though. It would be easier to tap into the power for the GPS and put a switch on it.

**Carz** · 03-26-2009, 03:57 PM

Originally Posted by Curiosity

You can test it by exiting all programs including xport, going to Device Manager, Ports, right-click COM1, disable, then enable, and see if it works.

Thanks! I'll do just that.

I'm beginning to suspect the issue may be related to the Huawei E270 USB Modem which I did not isolate due to the fact it's not very accessible at the back of my boot.

It may appear if a dial-up connection was alive when entering hibernation the com1 gps cannot resume (or initialize correctly in the case when power was cut off). I need to duplicate this later to confirm my suspicion by hibernating with and without live connection.

If it is true, then I may have found my answer and try XPort to stop/start the modem. If it doesn't help, I've to find ways to initiate auto disconnection when hibernating.

It may seem unrelated but Huawei uses com3 and com4 which perhaps interfered with the com1 in some ways

**Curiosity** · 03-27-2009, 07:40 AM

I really doubt it has anything to do with com3 and com4, but you may be onto something with the modem. You should try disconnecting it. Since that hub is underpowered, it could still be causing the problem. See, all the devices including the wi-fi, modem, hub, PC, GPS, everything has capacitors that discharge over several minutes when the PC turns off, so waiting say an hour causes a larger surge than waiting several seconds. Too much can drop the voltage too low and mess things up when they are trying to initialize. That could very well be happening. You could test by unplugging the modem or wi-fi, or both. If that does fix it, find out how that hub was powered before.

**Carz** · 03-28-2009, 01:10 AM

Thanks! I'm taking in everything you're suggesting and hope to resolve it soon.

Yesterday when I resumed from hibernation from previous session where I specifically make sure no connection was active, the gps resumed fine!

I ended the day hibernating with an active connection just so I can duplicate the gps problem later today. I've a strong hunch it will happen. If so, I have to find a solution to terminate the connection automatically before hibernation sets in. Then I may be able to avoid meddling with the hub since the wireless adaptor and modem can resume without issue.

Fingers crossed.

**Carz** · 03-30-2009, 05:06 PM

It seems the com1 gps not resuming from hibernation issue only happen when I'm using MWconn. Weird as it can detect the system entering into standby or hibernation as there will be a MWconn pop-up status in the forground showing so. If I'm not using MWconn, it doesn't matter if my internet is connected or not as the com1 gps can resume fine from hibernation.

**Curiosity** · 03-31-2009, 10:17 AM

So don't use MWconn?

Maybe MWconn is keeping it turned on, and so it's using more power. You could try playing with the Power Management settings for the modem or the hub in Device Manager under USB.

**Carz** · 03-31-2009, 06:22 PM

Yeah. I've currently disabled MWconn till I'm sure the com1 gps can resume consistently for a week or so before I can be certain. I can do w/o MWconn though. I'm on an unlimited plan so I don't use it to manage my connection but rather to relay the connection mode and signal strength reading to Logodash plugin for CF on the main page where I can see these readings easily rather than having to toggle to the desktop.

When I'm certain I will look for a solution if it's worth it to use MWconn again cos there are pple who is complaining that they can't hibernate due to active connection and others recommending them to use MWconn which will terminate the connection before hibernating and yet while I'm not facing hibernation issue as a whole, I'm facing problem resuming from hibernation when using it.

Not sure if the hub has the option to "Allow the computer to turn off this device to save power" (my Asus Wireless-G doesn't have but previous Compex have, probably such support are built into certain drivers as default) but I've certainly enable it for the modem. Will look into this area as well.