-
Want to remove ads? Register an account and login to see fewer ads, and become a Supporting Member to remove almost all ads.
Caution - Annual Service - Car power-down - GPS issue
- Thread starter jfischer
- Start date
-
- Tags
- clock gps broken gps rollover
#0879 had the SERVICE REMINDER trip on 14-August-20, and then VDS thought it became 01-Jan-01. Thanks to all who posted the many helpful comments on this thread. I was able to make my own update cable and execute the update from all the info provided herein [I already had a USB-Serial dongle, so cable-making was pretty straightforward]. I went in through the VMS connector (for anyone who is keeping score). VDS now thinks that it's 19-August-20. Thanks, again, to all.
Good for you! Did you hack into the USB-Serial dongle to access the USB 5V supply like I did to make a dedicated cable, or did you just supply 5V from a separate source and connect into the VMS connector with separate pins?
pd_australia
Member
One of my friends quickly made this cable based upon the instructions in the forum. Haven't tried it yet, but will soon.
Keen to hear if anyone has had success/failures with this cabling.
Keen to hear if anyone has had success/failures with this cabling.
It's interesting to see the different variations that people come up with. That looks like a little Sparkfun USB-serial breakout board, but it doesn't match any of the ones I see currently shown on their website. I'm curious whether the output is at TTL level or RS232.
BTW, mounting the big AMPSEAL connector in some kind of small box is a good idea because it takes some force to mate and separate the connector, so loose wires may become broken.
pd_australia
Member
Roughly the same as this one
Arduino Compatible USB to Serial Adaptor Module | Jaycar Electronics
All on 5V with FT232RL chipset.
Thanks for the advice on the small box!
OK, that's TTL level, and by configuring the FT232RL chip to invert the RX and TX signals to be inverted then the polarity will be correct. That's the way I built the first version of my cable which successfully updated the first car where I tested it. But the second person to try my cable ran into trouble and observed that the output of the Garmin GPS was bipolar, so I made my second version of the cable using a dongle outputting RS232 levels.
Please report back whether this one works for you.
pd_australia
Member
So I tried this today, but no success ...
Used Win7 and Win 8 in a Virtual Machine, bridging the USB directly to the machines.
Had to install drivers for the FTDI chipset (D2XX Direct Drivers
Used FT_Prog to try to inverse R/TX signals. FTDI Utilities
I also used CoolTerm 32 bit to monitor what's on the Serial port...
1. Configured COM3 for 4800 Bps
2. Used CoolTerm and could see data being send every now and then (but not in the format expected or readable, even after changing bauds) -> to me, this confirmed the GPS is powered up
3. Disconnect CoolTerm and used SNSRXCFG (running as Admin) to connect to COM3 on 4800bps, but nothing.
Tried the same without the RX/TX inverse ...
Going to try and connect from a physical Windows 10 now ... but if anyone has any other ideas ...
Used Win7 and Win 8 in a Virtual Machine, bridging the USB directly to the machines.
Had to install drivers for the FTDI chipset (D2XX Direct Drivers
Used FT_Prog to try to inverse R/TX signals. FTDI Utilities
I also used CoolTerm 32 bit to monitor what's on the Serial port...
1. Configured COM3 for 4800 Bps
2. Used CoolTerm and could see data being send every now and then (but not in the format expected or readable, even after changing bauds) -> to me, this confirmed the GPS is powered up
3. Disconnect CoolTerm and used SNSRXCFG (running as Admin) to connect to COM3 on 4800bps, but nothing.
Tried the same without the RX/TX inverse ...
Going to try and connect from a physical Windows 10 now ... but if anyone has any other ideas ...
You should see a stream of valid characters at 4800 baud when power is applied to the GPS unit. It is not necessary to send anything to the GPS in order to start the data stream. I was going to suggest a bad ground connection as a possible cause for garbage characters, but in this instance you must have ground connected or the GPS would not power up.
Do you have an oscilloscope that you could use to see what the signals coming from the GPS look like?
Also, if you try Windows 10, remember that you need administrator privileges to control the serial port.
Do you have an oscilloscope that you could use to see what the signals coming from the GPS look like?
Also, if you try Windows 10, remember that you need administrator privileges to control the serial port.
pd_australia
Member
I am an idiot. Now it worked. I reprogrammed the the FT232R to inverse RX/TX .. but probably didn't disconnect and reconnect.
GPS data coming in at 4800bps, and Garmin updater worked flawlessly.
So to summarise the process I followed:
1) Inhibit APS
2) Disconnected black cable at VMS
3) Used a USB-to-Serial board and connected to AMPSeal
4) Used Win7 in VMWare Fusion on Mac
5) Used FT_Prog to inverse RX/TX (don't forget to save and disconnect the USB)
6) Used CoolTerm to test for incoming GPS data @4800 bps on COMX
7) Used SNSRXCFG (running as Admin) to update
Thanks to the great advice and material on this forum!
. I learned so much today.
Cable available for anyone on this side of the world (Australia, APJ)...
cheers,
Pieter
GPS data coming in at 4800bps, and Garmin updater worked flawlessly.
So to summarise the process I followed:
1) Inhibit APS
2) Disconnected black cable at VMS
3) Used a USB-to-Serial board and connected to AMPSeal
4) Used Win7 in VMWare Fusion on Mac
5) Used FT_Prog to inverse RX/TX (don't forget to save and disconnect the USB)
6) Used CoolTerm to test for incoming GPS data @4800 bps on COMX
7) Used SNSRXCFG (running as Admin) to update
Thanks to the great advice and material on this forum!
. I learned so much today.Cable available for anyone on this side of the world (Australia, APJ)...
cheers,
Pieter
Sorry, just saw this [forgot to tick e-mail notifications]. I pulled the 5V supply from the second USB port on my laptop (I didn't want to cut into my standalone USB-RS232 dongle as I have other uses for it). I have many-many-many original, but now almost useless, original iPhone-USB cables lying around, and I sacrificed one of them for its male USB2-A connector, and simply pulled the 5V from the second USB port on my laptop [a very old HP TC1100].
See attached photo---I used your [slcasner] brad trick as the male connectors into the VMS female connector (color coded by your [slcasner] scheme), the RS-232 female connector connects to the USB-RS232 dongle, and the male USB2A connector connects into the second USB laptop port for the 5V to power the Garmin.
Calling all cars in Canada: The cable I made for updating the firmware in the GPS receiver is wending its way to Calgary with a planned second stop in Montreal. If there are cars in other Canadian cities needing the update it would make sense to send the cable around within Canada before it comes back home to me in California.
For the benefit of those who have not read this thread, the purpose of this cable is to connect a laptop to the GPS sensor in the Roadster to update its firmware. This corrects a problem caused by the week-number field in the GPS satellite protocol rolling over to a new epoch. The symptom in the car is an incorrect date displayed on the VDS. The date actually goes backwards by 19.7 years but may appear to go forward by several months if you don't see the year.
Anyone else in Canada who would like to borrow my cable to do this update should PM me to make arrangements. A word of warning to set expectations: To use the cable requires some physical dexterity to work in the passenger footwell and disconnect a cable from the VMS box (the car's main computer).
For the benefit of those who have not read this thread, the purpose of this cable is to connect a laptop to the GPS sensor in the Roadster to update its firmware. This corrects a problem caused by the week-number field in the GPS satellite protocol rolling over to a new epoch. The symptom in the car is an incorrect date displayed on the VDS. The date actually goes backwards by 19.7 years but may appear to go forward by several months if you don't see the year.
Anyone else in Canada who would like to borrow my cable to do this update should PM me to make arrangements. A word of warning to set expectations: To use the cable requires some physical dexterity to work in the passenger footwell and disconnect a cable from the VMS box (the car's main computer).
CruzInVA
#1422 Silver
Well, it looks as if I’ve joined the Time Traveler club as well, though I am not sure exactly when it happened.
I’ve been sadly negligent in keeping up with Forum activity, so I only heard about the problem this week, but sure as heck, Roadster 1422 (a 2.5 Sport), thinks it is 2001!
Is that a Garmin GPS puck or a Flux Capacitor in there?
Took this shot yesterday, Jan 14, 2021:
Interestingly, I had just picked the car up from the SC following an annual. Either the date scram happened after they were done with it, it had already occurred but they didn’t notice the discrepancy in the logs, or they noticed it but chose to ignore it. The SC certainly said nothing to me about it.
I wish I had been on top of things back in Dec 2019 when @JohnGarziglia had @slcasner ’s rig in Reston.
@sethr - do you eventually get yours sorted out? I am a little afraid to even mention it to the SC, in light of recent experience.
Any idea when one of the connector rigs might become available?
Thanks guys (from the halcyon days of 2001!)
Ted
I’ve been sadly negligent in keeping up with Forum activity, so I only heard about the problem this week, but sure as heck, Roadster 1422 (a 2.5 Sport), thinks it is 2001!
Is that a Garmin GPS puck or a Flux Capacitor in there?
Took this shot yesterday, Jan 14, 2021:
Interestingly, I had just picked the car up from the SC following an annual. Either the date scram happened after they were done with it, it had already occurred but they didn’t notice the discrepancy in the logs, or they noticed it but chose to ignore it. The SC certainly said nothing to me about it.
I wish I had been on top of things back in Dec 2019 when @JohnGarziglia had @slcasner ’s rig in Reston.
@sethr - do you eventually get yours sorted out? I am a little afraid to even mention it to the SC, in light of recent experience.
Any idea when one of the connector rigs might become available?
Thanks guys (from the halcyon days of 2001!)
Ted
The rollover event occurred on April 6, 2019 but the GPS sensor would continue to maintain the correct time after that event because it could keep track of the fact that the rollover occurred. The incorrect time arises when power has been removed from the GPS sensor long enough for its backup battery to be depleted so its microprocessor stops and it forgets about the rollover. So the likely cause for your car was that the power was shut down during the PEM service for a long enough period. There have been cars go through annual service without inducing the time symptom, so perhaps those were performed more expeditiously.
My cable just cleared customs on its way into Canada. After the cars in Calgary and Montreal are updated, I have a request pending in South Carolina. The cable could go to you from there if you would like to try this update yourself. If so, please PM me with your email and postal addresses.
I have made cables that cover two possibilities:
One cable is a USB-to-RS232-with-5V power that connects to the PC. One of the others is then connected to the RS232 interface. The cable with the black box is used for connection via the VMS cable. The cable with the white plug can be used for direct connection to the module. The connection via the VMS cable has been verified to be functional on two roadsters so far. The other has not been tested. I didn't need it but couldn't resist the temptation to make it as I had the parts. PM me if interested.
One cable is a USB-to-RS232-with-5V power that connects to the PC. One of the others is then connected to the RS232 interface. The cable with the black box is used for connection via the VMS cable. The cable with the white plug can be used for direct connection to the module. The connection via the VMS cable has been verified to be functional on two roadsters so far. The other has not been tested. I didn't need it but couldn't resist the temptation to make it as I had the parts. PM me if interested.
Anyone have the VDS in their car showing a date in August 2001 now (instead of the current date April, 2021)? If so, the firmware in your GPS sensor needs to be updated. The cable I made for performing said update has finished a tour of duty in Canada and is now heading to Virginia. Read the earlier posts in this thread about the update procedure, and if you would like to borrow my cable before it comes home to me in California, PM me with your postal address.