The CANary project

[garygid]

You are finding months, days, hours, minutes, and seconds, but missing years?
Finding them where, please?

 

[TickTock]

You are finding months, days, hours, minutes, and seconds, but missing years?
Finding them where, please?

Yes. I posted on this in the Canbus decoding thread and use them to auto-sync the CANary clock. You can get the details in the decoding doc. Carcan 5f9-5fc

 

[TickTock]

Minor, but useful new rev published (174). Now the audible brake monitor is calibrated. brkMonRate in the CONFIG.TXT has units Wh and defaults to 4.0. This means you get one chirp for every 4Wh of energy dissipated in the friction brakes. So every 20 chirps is equivalent to 1 gid of wasted energy. brkMonThr has units Wh/second and defaults to 1.0. This means if you are dissipating less than 1 Wh/second (80 seconds to waste 1 gid) CANary won't chirp at all.

 

[garygid]

Nice work, I see that I read the October posts, and even posted,
but ... what were we discussing?

Yes, young people, use your youth and years of health well, because
growing old may be a blessing, but it can bring its own challenges.

With affection for my many leafy friends, a real blessing indeed.

 

[TickTock]

Nice work, I see that I read the October posts, and even posted,
but ... what were we discussing?

I can't remember - wasn't it something about battery degradation? ... ;-)

 

[TickTock]

I just published a new version. Three edits:
1) Changed the max/min efficiency to only use round trip efficiency (from leaving home to returning home) instead of looking at each trip so you won't get inflated estimates from trips with large net elevation changes. If you drive up a mountain, stop, and then drive down (two separate trips), both your max and min DTE estimate would get updated with values that are not very useful. This is no longer the case.
2) Fixed a bug in the cellpair histogram when the number of buckets exceeded the array size (didn't happen often but when it did it caused a hang).
3) Added a Wh/gid display to the Health Screen (in parentheses next to the gid count) so we can see what number (75-80) we really should be using for computing available energy. Also should show if there are impacts from temperate, SOC, etc.

 

[robot256]

Hey guys, glad to see the project is continuing. I've been busy all winter and haven't been able to play with the schematics of mbed module I was working with back in, what, October? I am still planning to try something with the VSP sounds when I get my CANary assembled this summer. I found the correct part numbers for the connectors on the VSP module in my 2013 and was planning to make a little box with a relay, dummy loads, and audio amplifier to cut the speaker wires to my own audio source and give dummy loads to the stock VSP box. The mbed library for playing WAV files from USB and SD memory also seem pretty functional fro a first trial.

One interesting observation that I'm not sure where else to report: I disconnected my VSP module while I was investigating the pinout etc. and left it disconnected for a month. After that time, I decided a completely silent car wasn't my style, and reconnected it. Everything seemed back to normal, but the driver's window auto-up feature would not work. Auto-down fine, but auto-up did nothing. It was that way all winter, through additional battery disconnects and an encounter with a deer. Then today I got my 12 month service and a recall software update, and the auto-up feature works again. Wonder if it is considered a safety feature and is interlocked with the VSP module or something, or I just caused a glitch in the CAN bus that corrupted a config bit somewhere that got cleared during the update.

Anyway, something to consider if you want to go mucking about with the VSP module.

 

[turbo2ltr]

I've had my VSP disconnected in my 2013 since I got it in Dec and the auto up works fine. Probably just needed to be recalibrated. (I think it's "hold up for x number of seconds")

 

[TickTock]

Yup, just press and hold for 5 seconds or so. I like the VSP box - make sure your audio amp has enough power that you can do justice to the Harley sound effect . Are you going to do a frequency counter on the input feed or tap into the canbus to modulate the output?

 

[robot256]

If I hold the button for 5 seconds, the window will go all the way up, auto-up or not :lol: But now I only have to tap it into the second-level up position and it will continue moving up after I let go, which is what it did when new.

I'm more concerned that the speaker under the hood won't have good bass response. I expect it will work just fine for the Jetson's car sound, but we might have to install a subwoofer for the Harley Davidson or the Star Trek reactor sounds I was testing a while back.

I don't know how I'm going to do the sound modulation, but I figured by linking it to the CANary I would already have the CAN bus information available, no additional integration required. If I did it as a standalone unit, I would tap the CAN bus at the VSP connector.

 

[TickTock]

If I hold the button for 5 seconds, the window will go all the way up, auto-up or not :lol:

Hold for 5 seconds *after it reaches the top*. I guess I should have been more specific. Besides, in the winter 5 seconds may not be enough for it to reach the top

 

[robot256]

If I hold the button for 5 seconds, the window will go all the way up, auto-up or not :lol:

Hold for 5 seconds *after it reaches the top*. I guess I should have been more specific. Besides, in the winter 5 seconds may not be enough for it to reach the top

OOOHHHH.... I had no idea that the feature could enabled/disabled by the user. Guess I should have read my manual more thoroughly :shock:

 

[QueenBee]

If I hold the button for 5 seconds, the window will go all the way up, auto-up or not :lol:

Hold for 5 seconds *after it reaches the top*. I guess I should have been more specific. Besides, in the winter 5 seconds may not be enough for it to reach the top

OOOHHHH.... I had no idea that the feature could enabled/disabled by the user. Guess I should have read my manual more thoroughly :shock:

It's not that it can be disabled it's that it loses it's calibration and needs to be recalibrated, for example when the battery gets disconnected. I think disconnecting the battery is part of the software update procedure so they recalibrate your window for you thus why it worked after taking it to the dealer.

 

[garygid]

Yes, you are correct, disconnecting the 12v battery causes the loss of
calibration for the auto-up window feature.

 

[thew]

SO after all 68 pages I cant tell if One can buy this yet?

 

[TickTock]

This is a DIY project. If you want something pre-assembled then there are quite a few other alternatives that give more than enough data for your typical driver.

 

[TickTock]

I have recently been made aware that the LCD displays currently being shipped are a new revision which uses a completely different display controller. Not only does it require a new library, but the wiring is different. Same part number, etc, but completely different behavior. I guess this is the sort of thing you can expect when you buy parts from 2nd rate companies. :x
I have updated the instructions (http://mbed.org/users/TickTock/code/CANary/) to include instructions for both. Also, anyone using the Rev2.00 version of the display will also have to edit the precompile.h file in the TOUCH_TFTx2_w9341 library and set USE_ILI9341 to 1 prior to compiling.

 

[TickTock]

I just published a new rev. The primary change is the addition of the ability to edit the user watchpoints "in device." Previously, you had to edit the config file to program what msgId and byte you wanted to monitor for each of the eight watchpoint slots. Now, you can just long-press any of the watchpoints and a hexpad will come up on the opposite screen allowing entry of a new 4-character hex value indicating the msgId and byte (mmmb). Performing a "Save Config" will then allow the new value to be restored if power is cycled.

(Don't forget to edit the precompile.h file in the TOUCH_TFTx2_w9341 library and set USE_ILI9341 to 1 prior to compiling if you are using the new displays.)

 

[TickTock]

New rev (rev201) uploaded. Chages:
1) Fixed a bug in the long-touch code that was causing a lag
2) Fixed a bug in the config file save/restore (had the flags for brake monitor and regen monitor flipped causing these settings to load improperly whenever the device was reset)
3) Improved the efficiency computation to stop updating when the car is at a standstill. Before this fix, the DTE gauge would slowly drift to 0 if you were at a stop for a long period of time (~10 minutes). Corrected as soon as you started moving again but I decided it wasn't particularly useful to the driver to be told that they will get 0mpkWh while the car is at a standstill and would rather know what they get once they start moving again based on activity while actually travelling.