November 2017 - OvmsDev - Open Vehicles

Higher resolution events
by Tom Parker 10 Jan '18

10 Jan '18

Hi, To implement the Leaf's remote climate control, I need to send a frame, wait 50ms and then repeat another frame every 100ms for 25 repetitions. On the v2 hardware I implemented this with a 50ms delay and the ticker10th function. I wasn't able to find any delay functions in v3 so either I didn't look hard enough, nothing needs one, or busy-waiting isn't the right thing to do. Ticker10th isn't implemented. I could refactor the existing ticker task to fire 10 times a second instead of once per second, or I could register a new task for the climate control and cancel it after the the 25 repetitions. Or maybe something else altogether is appropriate? Thoughts?

4 7

SD CARD
by Mark Webb-Johnson 10 Jan '18

10 Jan '18

Any electronics guys out there can help out with SD CARD? It is frustrating as hell. The last part of the hardware puzzle. Once we solve this, we can hit the button to start making these. China and I have been struggling with this for the last month, with little progress. We are just following the standard Espressif schematics for this. Here is Espressif’s master document on it: https://github.com/espressif/esp-idf/tree/master/examples/storage/sd_card <https://github.com/espressif/esp-idf/tree/master/examples/storage/sd_card> Our circuit has been built for 4 wire SD mode, using the standard pins and pullups (as documented by Espressif). Result is that it is unreliable. Fails to mount 90% of SD CARDS, and those that do mount fail and corrupt data when writing lots of data out. The core problem (we think) is that the GPIO2 and GPIO12 pins used for SD CARD are also required to be at specific values during boot (to select different boot modes). Espressif suggest to work around this using the transistor logic they have for controlling boot mode pins using the USB-to-serial converter and DTR + RTS async control pins. We want 4-wire, but let’s start with a simple 1-wire diagram on a breadboard (as per Espressif instructions): ESP32 pin SD card pin SPI pin Notes GPIO14 (MTMS) CLK SCK 10k pullup in SD mode GPIO15 (MTDO) CMD MOSI 10k pullup, both in SD and SPI modes GPIO2 D0 MISO 10k pullup in SD mode, pull low to go into download mode (see note below!) GPIO4 D1 N/C not used in 1-line SD mode; 10k pullup in 4-line SD mode GPIO12 (MTDI) D2 N/C not used in 1-line SD mode; 10k pullup in 4-line SD mode (see note below!) GPIO13 (MTCK) D3 CS not used in 1-line SD mode, but card's D3 pin must have a 10k pullup N/C CD optional, not used in the example N/C WP optional, not used in the example Also: Connect GPIO0 and GPIO2 using a jumper (so transistors drive GPIO2 and GPIO0 low when entering download mode (10K pullup otherwise) 3.3V connect to SDCARD GND connect to SDCARD We have a variety of ESP32 development modules to try this with. Espressif said that the original DEVKIT C had a problem with the boot pin control lines and transistor logic (but never said what exactly that problem was). With original (old) ESP32 DEVKIT C, I can’t get it to work reliably at all. Sometimes ok. Sometimes not. With OVMS test code (that writes a lot of data), it never works. With our OVMS boards, all versions, I can’t get it to work reliably at all. Sometimes ok. Sometimes not. With OVMS test code (that writes a lot of data), it never works. I don’t think it ever worked. Just before our test was so simple that it didn’t show the problem. So, I just purchased one of these: https://www.banggood.com/ESP32-Development-Board-WiFiBluetooth-Ultra-Low-Po… <https://www.banggood.com/ESP32-Development-Board-WiFiBluetooth-Ultra-Low-Po…> Geekcreit® ESP32 Development Board WiFi+Bluetooth Ultra Low Power Consumption Dual Cores ESP-32 ESP-32S Board (Very similar to DEVKIT C, but more modern) It works 100%. Although after flashing it cannot boot (waiting for download) - need to disconnect GPIO0 - GPIO2 link, reset, then reconnect GPIO0-GPIO2 link. Then perfect. I think our problem is related to that transistor auto-flash reset circuitry. GPIO2 is the line for data transfer, so if problem will cause data corruption and unreliable (like we are seeing). Espressif SD CARD notes here: https://github.com/espressif/esp-idf/tree/master/examples/storage/sd_card <https://github.com/espressif/esp-idf/tree/master/examples/storage/sd_card> Schematic for Geekcreit board is here: https://www.dropbox.com/s/jefwxxtufgwg0ex/esp32_Schematic%20Prints.pdf?dl=0 <https://www.dropbox.com/s/jefwxxtufgwg0ex/esp32_Schematic%20Prints.pdf?dl=0> Schematic for latest version of Espressif DEVKIT C board is here: http://espressif.com/sites/default/files/documentation/esp32_hardware_desig… <http://espressif.com/sites/default/files/documentation/esp32_hardware_desig…> The original DEVKIT C (v1) had 1K resistors on that circuit. This is the version I have. The Geekcreit has 10K resistors on that circuit. OVMS has 10K resistors on that circuit. The latest DEVKIT C (v3) has 12K resistors on that circuit. DEVKIT C v1 schematic is here: http://dl.espressif.com/dl/schematics/ESP32-DevKitJ-v1_sch.pdf <http://dl.espressif.com/dl/schematics/ESP32-DevKitJ-v1_sch.pdf> DEVKIT C v2 schematic is here: http://dl.espressif.com/dl/schematics/ESP-WROVER-KIT_SCH-2.pdf <http://dl.espressif.com/dl/schematics/ESP-WROVER-KIT_SCH-2.pdf> The latest Espressif ROVER KIT development board schematic is here: https://dl.espressif.com/dl/schematics/ESP-WROVER-KIT_SCH-3.pdf <https://dl.espressif.com/dl/schematics/ESP-WROVER-KIT_SCH-3.pdf> Note that they are now using 100K resistors, and a separate transistor on IO0 (to allow it to be independently controlled after boot). I am reasonably certain our SD CARD problem concerns the GPIO0-GPIO2 link. Can you look at these, and see if you can find the difference? Apart from resistor sizes (which Espressif keep messing around with), I can’t see the practical difference. Regards, Mark.

4 12

Notifications concept draft
by Michael Balzer 10 Dec '17

10 Dec '17

Some thoughts about how to implement vehicle notifications and data logging. Please check & comment. Concept: - handle all unsolicited transmissions from vehicles by commands - vehicle module signals event if a transmission is due - … with event data = command line to generate output - listeners retrieve the transmission content by executing the command - queueing for async delivery is done by the listeners (i.e. through BufferedShell) → any shell command can be used for transmissions → all transmissions can also be generated on the shell / by script → all transmission output can be sent to any channel → data logging on SD can be done transparently by the system Interactive consoles can output unsolicited transmissions by event type according to the current log level. Events: "send.info" → send text notification "send.alert" → send text alert "send.data" → send data record (CSV, content part of v2 MP) Scheme: MyEvents.SignalEvent("send.<type>", "command [args]"); Examples: MyEvents.SignalEvent("send.info", "stat"); MyEvents.SignalEvent("send.alert", "xrt batt status"); MyEvents.SignalEvent("send.data", "xrt log RT-BAT-C"); → command output: H,RT-BAT-C,4,86400,1,1,4025,3675,4045,5,14,14,19,1 Regards, Michael -- Michael Balzer * Helkenberger Weg 9 * D-58256 Ennepetal Fon 02333 / 833 5735 * Handy 0176 / 206 989 26

5 28

Enclosure for OVMS v3
by Mark Webb-Johnson 04 Dec '17

04 Dec '17

Injection moulding samples are in. Clean and simple. And oh so black. Regards, Mark

3 6

Problems ms_v_charge_mode
by Geir Øyvind Vælidalo 03 Dec '17

03 Dec '17

I’m having problems setting ms_v_charge_mode. I do it this way, but it fails at runtime: StdMetrics.ms_v_charge_mode->SetValue( chademo ? "performance" : "standard"); This is the trace I get, so it seems to fail somewhere else: I (6451) events: Signal(vehicle.charge.mGuru Meditation Error of type LoadProhibited occurred on core 0. Exception was unhandled. Register dump: PC : 0x400883c3 PS : 0x00060633 A0 : 0x8008708a A1 : 0x3ffc6b30 0x400883c3: vListInsertEnd at /Users/goev/esp/esp-idf/components/freertos/./list.c:130 A2 : 0x3ffc6384 A3 : 0x3ffc6c78 A4 : 0x00060620 A5 : 0x00000000 A6 : 0x00000000 A7 : 0x00000000 A8 : 0x00000038 A9 : 0x3ffb33b8 A10 : 0x3ffb131c A11 : 0x00060623 A12 : 0x00060623 A13 : 0x3ffce1ec A14 : 0x3ffce1ec A15 : 0x3ffc6a70 SAR : 0x00000010 EXCCAUSE: 0x0000001c EXCVADDR: 0x00000040 LBEG : 0x400014fd LEND : 0x4000150d LCOUNT : 0xfffffffa Backtrace: 0x400883c3:0x3ffc6b30 0x40087087:0x3ffc6b50 0x400860f1:0x3ffc6b70 0x4008771b:0x3ffc6b90 0x40087818:0x3ffc6bc0 0x400883c3: vListInsertEnd at /Users/goev/esp/esp-idf/components/freertos/./list.c:130 0x40087087: vTaskPlaceOnEventListRestricted at /Users/goev/esp/esp-idf/components/freertos/./tasks.c:4446 0x400860f1: vQueueWaitForMessageRestricted at /Users/goev/esp/esp-idf/components/freertos/./queue.c:2392 0x4008771b: prvProcessTimerOrBlockTask at /Users/goev/esp/esp-idf/components/freertos/./timers.c:839 0x40087818: prvTimerTask at /Users/goev/esp/esp-idf/components/freertos/./timers.c:839 It looks like it has something to do with the Signal event, but I ‘m still not that good using gbd to find out what actually makes this fail. Any one got a clue? Regards, Geir

3 7

Leaf Updates
by Tom Parker 02 Dec '17

02 Dec '17

Hi, I ported over support for Gen 2 Chargers and polling for the battery Ah capacity and the "Hx" value in https://github.com/openvehicles/Open-Vehicle-Monitoring-System-3/pull/9 3 Questions: In v2 we checked whether transmit mode was enabled, I'm not seeing similar logic in the Telsa Roadster module. I've commented out the checks in the leaf code for the time being. In v2 I abstracted the hardware interactions necesary to write to the canbus into a function void vehicle_nissanleaf_send_can_message(short id, unsigned char length, unsigned char *data). In v3 we call canbus::Write(const CAN_frame_t* p_frame). Do we want to add a similar simpler "send to this buffer with this can id" method to the canbus object? The Leaf doesn't use any of the extended can features so a simple interface for the simple case seems to make sense. For the moment I've implemented it within the leaf codebase, see void OvmsVehicleNissanLeaf::SendCanMessage(uint16_t id, uint8_t length, uint8_t *data) in the pull request. I've also porting over the remote command features, I like the interface presented to the car modules to implement these. I need to wait until tomorrow to test these on a real car so haven't sent a pull request yet. I'm very close to finished on porting the leaf support!

2 4

Real authentication for SSH server
by Stephen Casner 28 Nov '17

28 Nov '17

I have updated the SSH server code in OVMS to remove temporary built-in passwords for jack and jill and public keys for hansel and gretel. Instead there is now real authentication code that fetches passwords and public keys from the config: - We keep the host key for the server in ssh.server:key. If none exists when the server is started then a background task is forked to create one, using the key generator in WolfSSL with entropy from esp_random(), and install it into the config. - The key generation also calculates the SHA256 fingerprint of the host key and stores it in ssh.info:fingerprint so that users can verify the key when connecting from a new client. The fingerprint is displayed when the generation is complete and can be viewed any time with "config list ssh.info". - Passwords may be stored under param password with the instance being the associated username. Alternatively, if there is no password configured for a username then the password stored under instance "module" is used. - RSA public keys may be stored under param ssh.keys with the instance being the associated username. The key format is as generated on a Linux or Mac system by the command "ssh-keygen -b 2048 -t rsa". This could be a key you already have or a new one made for this purpose. Only the one long string of the base64-encoded key should be stored, not including the "ssh-rsa" at the beginning or the user ID at the end. The key is stored with a command like this: config set ssh.keys casner AAAAB3NzaC1yc2EAAAADAQAB...C6p5jcbf4NCnX Public key authentication is really convenient because most client systems allow authenticating the private key once per login on the client and storing the decrypted private key in a key server. Then each session to the OVMS is authenticated automatically without requiring a password to be entered. WARNING: If the config does not contain any passwords (including for "module") nor any public keys, then authentication is considered disabled and any user can log in with no password or a random password. -- Steve

1 0

StdMetrics.ms_v_env_parktime
by Geir Øyvind Vælidalo 27 Nov '17

27 Nov '17

«Quick» question: StdMetrics.ms_v_env_parktime, is it seconds since the car was parked, which means we should update it in Ticker1? Or is it the time when the car was parked? Looking at V2 and the OVMS-V2conversion in V3, it looks to me like the first, but looking at the Twizy-module it looks like the latter. I guess setting the start time should be enough. When sending packets to V2-server, we could easily calculate the seconds: time(NULL)-ms_v_env_parktime. Unless ms_v_env_parktime is 0, then we’re not parked. The same could then be done for ms_v_charge_minutes too. Geir

2 4

Debugging
by Geir Øyvind Vælidalo 23 Nov '17

23 Nov '17

Hi guys! I’ve moved most of the Kia Soul code from V2 to V3. I set the vehicle module and everything is just fine, until I connect it to my car. Then the Kia module crashes. I can’t make out why looking at the little info it prints out. I see some of your are talking about gdb-stub, but I have never used gbd and need some help. How do I debug? What options do I have? Hope someone has a simple guide for a noob like me 😜 Thanks! Geir

3 5

Anybody else having trouble compiling today?
by Greg D. 22 Nov '17

22 Nov '17

Hi folks, I just fetched the latest updates, and I'm getting a bunch of errors in console_ssh.cpp and console_ssh.h. Variables not declared in scope, that sort of thing... I did a 'make app_flash', same as always. Any ideas? Greg

3 4