- OvmsDev - Open Vehicles

Transport Evolved
by Mark Webb-Johnson 10 Mar '12

10 Mar '12

Join Nikki Gordon-Bloomfield from Green Car Reports, Michael Thwaite from TeslaMotorsClub, and guest Mark Webb-Johnson from the Open Vehicle Monitoring System project as they discuss the week’s news in the world of electric cars, including the latest from the Geneva Motor Show, The Open Vehicle Monitoring System and much more.. http://www.transportevolved.com/transport-evolved-episode-92-the-anti-brick

1 0

Module stops working after couple of days
by Wido den Hollander 08 Mar '12

08 Mar '12

Hi, About a week ago I was able to install the module and get my own server up and running. Everything seemed to work fine from the beginning, but after about 36 hours the module died. First I thought my prepaid SIM-card had ran out of credits, but that was not the case. A manual reboot of the module was needed and it started working again. Yesterday I opened the app and saw "Last update 8 hours ago" Again, a reboot of the module was enough. I checked the LED's and the first time GREEN was solid ON and RED was flashing slowly, that is normal. The second time the GREEN LED was off and only the RED one was blinking. I don't think it's a problem of GSM reception, the module even worked without the antenna while lying on the floor of the car. GSM reception is however very good here in the neighborhood, no high buildings and it's a very dense network. It's also weird that the module doesn't come back online. If it was a short disruption it could be a connection problem, but being offline for 8 hours is long. Has anyone seen this behavior? Thanks, Wido

3 14

Re: [Ovmsdev] 0% SOC Reported in v1.5 Roadster
by Mark Webb-Johnson 07 Mar '12

07 Mar '12

Jack, Tom, Yep, Jack is right. It is feature #14. 0x0e. I seem to have get hexed. P.S. Tom's problem has been narrowed down to a faulty PIC board. The soldering seems ok, so probably the CAN bus driver chip. The module doesn't work in a 2.x roadster either. We're sending him a new board (I'm going to QC one tonight).. On 7 Mar 2012, at 3:34 PM, Jack West wrote: > Tom, > > Carbits is actually feature 14 and would need set to be set to 1 for a 2008 Roadster. > > Feature 15 set to 1 is to enable writing to the CAN bus. Set to 0 is 'listen' only. > You only need to do this once. > > So > SMS > FEATURE 15 1 (allows you to do all the fun stuff) > FEATURE 14 1 (because you have a 2008) > > That should do it (once you get the unit working). If you are using the new iPhone app you can do it from the Cars/I/Control tab. > > I don't think Feature 12 is anything yet. > > Jack > > > > On Mar 6, 2012, at 3:45 PM, Mark Webb-Johnson <mark(a)webb-johnson.net> wrote: > >> Tom, >> >> Scott451's original (ground breaking) post is at: >> >> http://www.teslamotorsclub.com/showthread.php/4388-iPhone-app/page2?p=49456… >> >> He talks there about the 2008 roadster, and confirms 1MHz bus speed. >> >> Your confirmation of the wales gps location and 0 SOC means the module is definitely not getting any can bus messages. >> >> It is either a faulty board, or some other incompatibility. >> >> Can you ask Cathy (she is your hardware support, right?) to check the continuity on connections on the MCP2551P and pins #23 and #24 on the PIC CPU? >> >> Do you know anyone close by with a 2.x roadster that you could plug the module in, to rule out 1.5 vs 2.x incompatibility? >> >> Regards, Mark. >> >> On 7 Mar 2012, at 1:35 AM, Tom Saxton wrote: >> >>> I did the "FEATURE 12 1" SMS command, but got no response. Are those >>> commands documented somewhere, or is it a "read the code" deal? (It is cool >>> to be able to read the code.) >>> >>> I get the same "not charging" status with the charge port door open. >>> >>> When I do a GPS command, I get back a link to 52.042460,-3.944089 which I >>> assume is the default position in Wales. >>> >>> So, it seems the OVMS module isn't getting any of these CAN messages. Is >>> there a way to tell if the board is getting any messages at all? >>> >>> With Cathy's help, I checked continuity and all four pins going to the MCU >>> board. They all check out, as do the four twin wires to the other connector. >>> >>> When plugged in, after the start up sequence, I get a solid green LED on >>> each board and a blinking blue LED on the modem board. >>> >>> Finally, I'm sure my diag port is functional as it has been working with my >>> Tattler (although the Tattler is going through some brain surgery right now >>> so I can't confirm that it's working today). >>> >>> BTW, I notice the generated GPS link takes you to a static map that can't be >>> scaled or panned. This makes it impossible to tell where the pin is if you >>> don't recognize any features in the static map. >>> >>> I like using a simpler link that gets you a fully functional map: >>> >>> http://maps.google.com/maps?q=52.042460,-3.944089 >>> >>> Also, I'm also not a big fan of the Tesla logo for the pin, but I have an >>> unusual history with the company that may be coloring my attitude on that. >>> >>> Tom >>> >>> on 3/5/12 5:38 PM, Mark Webb-Johnson wrote: >>> >>>> Tom, >>>> >>>> Good to have you aboard. >>>> >>>> We haven't been able to test the module on 1.5 cars (which is why we're only >>>> shipping to end-users with 2.x cars). The problem we've had is that all the >>>> developers involved have 2.x cars. >>>> >>>> That said, the module _should_ work in a 1.5 car. Scott451 did the original >>>> bus decodes on a 1.5 car, and while there are some differences, he did >>>> document the messages (but we haven't been able to verify this as we have with >>>> the 2.x cars). >>>> >>>> First, some trivia: >>>> >>>> Our current master document on CAN bus messages is at: >>>> https://github.com/markwj/Open-Vehicle-Monitoring-System/blob/master/vehicl… >>>> ar%20Module/canbus_notes.txt >>>> and that includes all of Scott451's published work, plus our own additions. >>>> >>>> The SOC message is on the CAN bus - message ID#100, B1=0x80. There is >>>> supposedly no difference between the 1.5 and 2.x models for this message. >>>> >>>> To get the charging messages correctly, you will need to set the CARBITS >>>> feature to a 2008 model roadster. An SMS command: >>>> FEATURE 15 1 >>>> should enable bit#1 of the carbits, and permanently record this as a 2008 >>>> roadster. >>>> >>>> The "Not charging" just means the charge port door is closed (or bit#2 of B2 >>>> on id#100 B1=0x96 is zero). The SOC and Ideal range being zero is similar zero >>>> (default) values. >>>> >>>> So, either none of these messages are getting from the CAN bus to the module, >>>> or the messages are completely different on a 1.5 car. I suspect the former. >>>> >>>> Can you try: >>>> a) A "GPS" SMS command - do you get a location in Wales (default), or the true >>>> location? >>>> b) Open the charge port door, and repeat the "STAT" command - do you get a >>>> charging reply? >>>> and let us know the result? >>>> >>>> The problem could also just be a faulty module (we had some factory soldering >>>> issues with early modules), so please verify the soldering looks good, and >>>> trace continuity test, particularly on the CAN bus side? Or, a loose >>>> connection on the CAN bus cables (two pins CAN-L and CAN-H), so can you check >>>> that with a continuity tester? If you have the opportunity, you could try it >>>> in a 2.x roadster. >>>> >>>> There is a wiring schematic at: >>>> >>>> https://github.com/markwj/Open-Vehicle-Monitoring-System/raw/master/vehicle… >>>> r%20Module/OVMS_V5a.png >>>> >>>> My guess for this problem is a loose/broken cable on the CAN bus side - >>>> somewhere between the DIAG connector and the telco-style jacks connecting to >>>> the PIC motherboard. >>>> >>>> It would be really helpful if you could help us test this and iron out any >>>> wrinkles for the 1.5 cars. We've had some enquiries from 1.5 end-users, but >>>> have told them to hold-off for the moment until we can verify things work in a >>>> 1.5 car. >>>> >>>> Regards, Mark. >>>> >>>> On 6 Mar 2012, at 8:39 AM, Tom Saxton wrote: >>>> >>>>> I just got the OVMS board plugged into our v1.5 Roadster. I haven't jumped >>>>> through the hoops to get a data plan on my SIM card yet, so I'm just using >>>>> SMS. >>>>> >>>>> I registered my phone and changed my password. Those commands worked as >>>>> expected. I then did a STAT and got back: >>>>> >>>>> Not charging >>>>> Ideal Range: 0 Km >>>>> SOC 0% >>>>> >>>>> It's true that it's not charging, but the car is actually at 96% and 185 >>>>> ideal miles. >>>>> >>>>> What should I do next? >>>>> >>>>> Tom >>>>> >>>>> >>>>> _______________________________________________ >>>>> OvmsDev mailing list >>>>> OvmsDev(a)lists.teslaclub.hk >>>>> http://lists.teslaclub.hk/mailman/listinfo/ovmsdev >>>> >>> >>> >> >> _______________________________________________ >> OvmsDev mailing list >> OvmsDev(a)lists.teslaclub.hk >> http://lists.teslaclub.hk/mailman/listinfo/ovmsdev

1 0

GPS SMS Response
by Tom Saxton 07 Mar '12

07 Mar '12

I notice that the SMS response to the GPS yields a gigantic link to a static map that can't be zoomed or panned. This makes it very difficult to figure out where the car is if you don't happen to recognize any of the features shown in the map. I prefer a simpler link that takes you to a fully functional map: http://maps.google.com/maps?q=52.042460,-3.944089 Also, I'm also not a big fan of the Tesla logo for the pin, but I have an unusual history with the company that may be coloring my attitude on that. Tom

4 3

Re: [Ovmsdev] 0% SOC Reported in v1.5 Roadster
by Mark Webb-Johnson 07 Mar '12

07 Mar '12

Tom, Scott451's original (ground breaking) post is at: http://www.teslamotorsclub.com/showthread.php/4388-iPhone-app/page2?p=49456… He talks there about the 2008 roadster, and confirms 1MHz bus speed. Your confirmation of the wales gps location and 0 SOC means the module is definitely not getting any can bus messages. It is either a faulty board, or some other incompatibility. Can you ask Cathy (she is your hardware support, right?) to check the continuity on connections on the MCP2551P and pins #23 and #24 on the PIC CPU? Do you know anyone close by with a 2.x roadster that you could plug the module in, to rule out 1.5 vs 2.x incompatibility? Regards, Mark. On 7 Mar 2012, at 1:35 AM, Tom Saxton wrote: > I did the "FEATURE 12 1" SMS command, but got no response. Are those > commands documented somewhere, or is it a "read the code" deal? (It is cool > to be able to read the code.) > > I get the same "not charging" status with the charge port door open. > > When I do a GPS command, I get back a link to 52.042460,-3.944089 which I > assume is the default position in Wales. > > So, it seems the OVMS module isn't getting any of these CAN messages. Is > there a way to tell if the board is getting any messages at all? > > With Cathy's help, I checked continuity and all four pins going to the MCU > board. They all check out, as do the four twin wires to the other connector. > > When plugged in, after the start up sequence, I get a solid green LED on > each board and a blinking blue LED on the modem board. > > Finally, I'm sure my diag port is functional as it has been working with my > Tattler (although the Tattler is going through some brain surgery right now > so I can't confirm that it's working today). > > BTW, I notice the generated GPS link takes you to a static map that can't be > scaled or panned. This makes it impossible to tell where the pin is if you > don't recognize any features in the static map. > > I like using a simpler link that gets you a fully functional map: > > http://maps.google.com/maps?q=52.042460,-3.944089 > > Also, I'm also not a big fan of the Tesla logo for the pin, but I have an > unusual history with the company that may be coloring my attitude on that. > > Tom > > on 3/5/12 5:38 PM, Mark Webb-Johnson wrote: > >> Tom, >> >> Good to have you aboard. >> >> We haven't been able to test the module on 1.5 cars (which is why we're only >> shipping to end-users with 2.x cars). The problem we've had is that all the >> developers involved have 2.x cars. >> >> That said, the module _should_ work in a 1.5 car. Scott451 did the original >> bus decodes on a 1.5 car, and while there are some differences, he did >> document the messages (but we haven't been able to verify this as we have with >> the 2.x cars). >> >> First, some trivia: >> >> Our current master document on CAN bus messages is at: >> https://github.com/markwj/Open-Vehicle-Monitoring-System/blob/master/vehicl… >> ar%20Module/canbus_notes.txt >> and that includes all of Scott451's published work, plus our own additions. >> >> The SOC message is on the CAN bus - message ID#100, B1=0x80. There is >> supposedly no difference between the 1.5 and 2.x models for this message. >> >> To get the charging messages correctly, you will need to set the CARBITS >> feature to a 2008 model roadster. An SMS command: >> FEATURE 15 1 >> should enable bit#1 of the carbits, and permanently record this as a 2008 >> roadster. >> >> The "Not charging" just means the charge port door is closed (or bit#2 of B2 >> on id#100 B1=0x96 is zero). The SOC and Ideal range being zero is similar zero >> (default) values. >> >> So, either none of these messages are getting from the CAN bus to the module, >> or the messages are completely different on a 1.5 car. I suspect the former. >> >> Can you try: >> a) A "GPS" SMS command - do you get a location in Wales (default), or the true >> location? >> b) Open the charge port door, and repeat the "STAT" command - do you get a >> charging reply? >> and let us know the result? >> >> The problem could also just be a faulty module (we had some factory soldering >> issues with early modules), so please verify the soldering looks good, and >> trace continuity test, particularly on the CAN bus side? Or, a loose >> connection on the CAN bus cables (two pins CAN-L and CAN-H), so can you check >> that with a continuity tester? If you have the opportunity, you could try it >> in a 2.x roadster. >> >> There is a wiring schematic at: >> >> https://github.com/markwj/Open-Vehicle-Monitoring-System/raw/master/vehicle… >> r%20Module/OVMS_V5a.png >> >> My guess for this problem is a loose/broken cable on the CAN bus side - >> somewhere between the DIAG connector and the telco-style jacks connecting to >> the PIC motherboard. >> >> It would be really helpful if you could help us test this and iron out any >> wrinkles for the 1.5 cars. We've had some enquiries from 1.5 end-users, but >> have told them to hold-off for the moment until we can verify things work in a >> 1.5 car. >> >> Regards, Mark. >> >> On 6 Mar 2012, at 8:39 AM, Tom Saxton wrote: >> >>> I just got the OVMS board plugged into our v1.5 Roadster. I haven't jumped >>> through the hoops to get a data plan on my SIM card yet, so I'm just using >>> SMS. >>> >>> I registered my phone and changed my password. Those commands worked as >>> expected. I then did a STAT and got back: >>> >>> Not charging >>> Ideal Range: 0 Km >>> SOC 0% >>> >>> It's true that it's not charging, but the car is actually at 96% and 185 >>> ideal miles. >>> >>> What should I do next? >>> >>> Tom >>> >>> >>> _______________________________________________ >>> OvmsDev mailing list >>> OvmsDev(a)lists.teslaclub.hk >>> http://lists.teslaclub.hk/mailman/listinfo/ovmsdev >> > >

1 0

0% SOC Reported in v1.5 Roadster
by Tom Saxton 07 Mar '12

07 Mar '12

I just got the OVMS board plugged into our v1.5 Roadster. I haven't jumped through the hoops to get a data plan on my SIM card yet, so I'm just using SMS. I registered my phone and changed my password. Those commands worked as expected. I then did a STAT and got back: Not charging Ideal Range: 0 Km SOC 0% It's true that it's not charging, but the car is actually at 96% and 185 ideal miles. What should I do next? Tom

4 3

A few announcements
by Mark Webb-Johnson 02 Mar '12

02 Mar '12

A few of announcements: 1] We are now open for end-users to use the system. 2] MPT has done a nice write-up in greencarreports.com: http://www.greencarreports.com/news/1073583_open-vehicles-community-release… 3] iOS App v1.2.0 (including vehicle charge control, lock/unlock, valet mode on/off, feature and parameter setting, cellular usage charts, etc) has just been submitted to the Apple iTunes App store for approval. Hopefully it will be through in a couple of days. 4] Tesla Motors Club have kindly offered to host the infrastructure to support the project. This is clearly a perfect fit and we are grateful for their support. They will host the production server, and we will be moving over all existing users from the Hong Kong development server in the coming few days. http://www.teslamotorsclub.com/showthread.php/7705-PR-OpenVehicles-com-sele… Regards, Mark

1 0

OVMS Module for 2.x Tesla Roadsters
by Mark Webb-Johnson 01 Mar '12

01 Mar '12

We're delighted to announce that the production end-user release of the OVMS system for 2.x Tesla Roadsters is now available and will be formally launched on March 2nd 2012. We are giving 24 hour advanced notice to users that pre-registered with us to let us know they were interested in this project, so that you can be sure to obtain the system early. You can order now, with shipping starting mid next-week. Requirements are: A 2.x Roadster (VDS in the centre console, and a DIAG port in the passenger footwell) A GSM SIM card, with either SMS or GPRS Data plan SMS control from any SMS-compatible smartphone App control from iPhone/iPodTouch/iPad running at least v5.0 iOS software App control from modern Android smartphones and tablets Please note that we do not currently support v1.x Tesla Roadsters. The system _should_ be compatible with v1.x cars, but we haven't verified this at all. If you have a v1.x car, would like this system, and are willing to help, please let us know and we'll see if we can arrange a nearby developer to work with you. But, for the moment, we are recommending this only for 2.x cars. A rough feature list is: The OVMS car module goes into Tesla Roadsters, plugs into the DIAG port for power and car interface, and comes with a small adhesive external antenna for GSM reception. The modules come pre-programmed with the latest (v1.2) firmware. The end-user will need to add a GSM SIM card. If you only want SMS control, an SMS-only plan is ok, otherwise for data you will need 1MB-2MB a month for normal usage. SMS control is available to examine car status, GPS location, and to configure parameters and features. Apps are available for Android (in the Android marketplace) and iPhone/iPodTouch/iPad (in the Apple App Store). For Apple, iOS v5 software (or later) is required. Notification of charge interruption and other ancillary messages is available via SMS and/or Android/Apple PUSH notification. An emergency alert system is included to repeatedly alert (via SMS and Android/Apple PUSH notifications) should the car be off and the battery SOC fall below 4%. The Apps provide access to car battery status (battery SOC, estimated range, ideal range, and parking meter). The Apps provide access to car environment status (door status, lock status, ambient temperature, PEM temperature, Motor temperature, Battery temperature) and TPMS pressures and temperatures. Some of this data is only available when the car is turned on, or running (e.g.; TPMS). The Apps provide access to car location (shown on a Google map, using the on-board GPS of the car). An optional stream mode will provide location streaming to connected apps (with corresponding increase in data usage), otherwise status and location data is provided once every ten minutes (normal) or every minute (when one or more Apps are connected to the car). A "Friends" mode is a new feature already supported by the car firmware, and is coming soon as a "Friends" tab in the Apps. The car is set to send location, speed and battery SOC updates to a named friends group, and everyone with access to that group can see all the cars on one map screen. Great for multi-car families, events and meet-ups. Controls are available within the Apps to Start charging, Stop charging, set charge mode (standard, storage, range, performance), Lock the car, Unlock the car, Activate valet mode, Deactivate valet mode, Reset the car module, and provide for setting parameters and features. (* required read-write mode enabled). Note that for non-US cars, the lock/unlock cannot affect the immobiliser, so is of limited functionality. A user guide for the system is available for download and that describes both the functionality and installation process in the car. You can also download (for free) the Open Vehicle Apps from either the Apple iTunes App Store or Android Marketplace. To order the system, you will need a paypal account (or credit card acceptable to paypal without requiring an account). The hardware module you need is this one and costs US$99 plus shipping. You will need to create an account on the Open Vehicle Systems website, add the product to your shopping cart and checkout with Paypal. Please note that this is a non-profit, charitable project, with free open-source software and community support. We think the price (US$99) is incredible, and only just covers the cost of the hardware. If you agree you are getting a great deal then please contribute the difference to your favorite charity ;-) Regards, Mark and Sonny.

1 0

Strange preparing, connect-pwr-cable state
by Mark Webb-Johnson 27 Feb '12

27 Feb '12

I'm seeing something strange, that I'd like to know if it is repeatable for others. I previously thought this was some sequence problem with the messages we were sending (specifically, sending a start charge command when the charge port is closed or cable not plugged in), but I can now repeat it with just a normal idle car. Steps to reproduce: Charge the car as normal, and wait for "Charge done". Remove the charge plug+cable and close the charge port door. SMS STAT the car and you should get "Standard - Charging done" as the status. All good. Now, with the car still off, open the charge port door, wait a few seconds, then close it. SMS STAT the car and now we seem to get "Standard - Charging, Preparing". A check on the can bus shows charge state 13, sub-state 7 (preparing-to-charge, connect-pwr-cable). The iPhone App will show the battery with copper-coloured tops, as it seems to be expecting the car to be charging. This doesn't seem to happen with a car in 'stopped charge' state. The VDS shows nothing unusual (which implies that Tesla treat it specially, based on the charge door status). Can some other people try this, and let me know what you get? If you can have an App connected at the time of your test (so the server gets the logs), then let me know your vehicle ID, the time you did the test and what the result was, that would be helpful. It is not hard to handle as a special case in the code (both firmware and apps, or hacked in the firmware to detect the issue and put the state back to something more meaningful), but just seems a bit strange, as the state does not accurately reflect the car status. Regards, Mark.

3 4

SPEEDO feature
by Mark Webb-Johnson 25 Feb '12

25 Feb '12

> What is the status of the digital speedo feature? A developer asked about this off-list, and I thought it might be relevant to the wider community, so here it is. The speedo feature is enabled by defining OVMS_SPEEDO_EXPERIMENT at build time (either in ovms.def, or via configuration parameter). It is currently enabled for the experimental configuration, and off for the other two. The feature must also be enabled, after the module is powered up in the car, by setting feature #0 to a value >0. The value 5 is recommended as a good starting point. This feature #0 is volatile, and will be set to 0 if the module is powered down or rebooted. The reason the feature is classified experimental is because while it seems to work, we are really not sure what affect it has on the car. Something unexpected might damage the car. Use it at your own risk. Our warning is "The feature can transmit a lot of messages on the bus and is spoofing the car message. We can't see any affect (and we've even put an analyser on at the same time to see if any errors come up - they don't), but it is very intrusive to the car. One concern we have is if the car is using that message for anything else (but we can't see how it would as it also contains data for the battery level display, estimated miles, etc - it seems very very specific to that dashboard display). We really don't want anyone breaking their cars with this." If you do use, please let us have feedback on the ovmsdev mailing list. We really need to know more about this and any impact on cars (good, bad or zero). Regards, Mark.

1 0