|
|
|
|
|
|
BMW Garage | BMW Meets | Register | Today's Posts | Search |
|
BMW 3-Series (E90 E92) Forum
>
More Radio signal details
|
|
02-09-2014, 05:04 PM | #1 |
Lieutenant Colonel
113
Rep 1,951
Posts |
More Radio signal details
I'm back to working on my Nexus7 install. Software-wise, the N7 will go into deep sleep when power is removed, and will wake back up as soon as power is reapplied. Basically, that means I need to switch power to its USB port, while it's still plugged in.
The question is, what signal should I use to do the switching? I'd like to have the N7 come alive as soon as the headunit does, and go to sleep as soon as the headunit does. The problem is, the headunit's wake state is controlled by the CAN bus. Power is supplied from Terminal 30g, which turns on when the car is unlocked and turns off 30m later. That's not ideal - having a glowing nexus for 30m after I leave the car. Unfortunately, the amp remote line isn't exactly the answer either. In WDS this signal is labeled RAD_ON. Unfortunately, the radio turns on as soon as the car is woken up (ANY event - door open, lock, unlock) and turns off between 40s and 1m after the car is left (the last to occur of key out, lock, unlock, etc). So if you're tapping off the remote line, whatever you're switching will remain on for a minute after you leave. So on the turn-on side, the amp remote line is useful - I wouldn't mind if the N7 came on as soon as I unlocked the car. But the turn off isn't ideal - I'd like the N7 to turn off as soon as the key comes out of the car. The last promising line I looked at was the illumination signal on pin 14. I thought, man, this would be perfect if the N7 powered on as soon as the radio lit up, and powered off when it went dark. That's basically exactly what I want. Unfortunately, that's not what this pin does. It's known as Terminal 58G, and it's actually used to signal BRIGHTNESS of the interior dash lights. It's set by the "brightness" setting in the BC. Sadly, it's also only on when the headlights are (which is the only time dash brightness settings come into play, with the headlights off they're always max). But let's look into this more, how does it work? It turns out, 58G is a pulse width modulated (PWM) signal determining desired lighting brightness from 0 to 100%, with arbitrary apparent brightness at the 0 and 100 marks. The signal itself is an 80hz (12.5ms period) square wave, roughly 12v peak to peak (zero volts to ground low, actually between 10.5 and 12 volts to ground high, about what you'd expect). The minimum brightness setting corresponds to a high time of 1.04ms, or ~8.25% duty cycle. The 1 notch brightness setting corresponds to a high time of 2.64ms, or 21% duty cycle. The 2 notch brightness setting is a high time of 4.32ms, or 35% duty cycle. And so on: 3=5.92ms(47%), 4=7.6ms(60%), 5=9.36ms(75%), 6=10.9ms(87%) then 7 notches is actually max brightness at 100%, even though there are two notches left to go in the setting. Not QUITE linear, but close enough. Sadly, after all that, the signal still isn't super useful to me, since it's only on when the headlights are on, and also doesn't exactly follow radio-on timing like one might hope. When my power switching PCB comes back in a couple days, I'll probably just tap off the remote line and call it good enough, it's probably not going to kill me if the nexus glows for another minute after I lock the car. Though I'm still looking for something a bit better (and I'd rather not open up the headunit to tap off the backlight power itself).
__________________
|
02-09-2014, 05:27 PM | #2 | |
Major General
237
Rep 5,118
Posts |
Quote:
1. Retrofit factory USB. 2. Use an aftermarket single din head unit with USB. You can hide it in the dash or in the ashtray. Using the connects2 harness you will get steering wheel controls and the aftermarket head unit would turn on and off just like stock. |
|
Appreciate
0
|
02-09-2014, 05:35 PM | #3 | |
Lieutenant Colonel
113
Rep 1,951
Posts |
Quote:
Incidentally, factory USB on its own doesn't play (very) nicely with android anyway. You can browse the filesystem for songs you have downloaded, but the interface is wonky at best and doesn't handle any other audio, i.e. navigation, pandora, spotify, rdio, etc. There's also no chance in hell that the MULF will charge a Nexus 7, or any tablet for that matter. Hell, any phone. Even with my power injector as I had it before, there's just no way to get the device to draw >500mA when connected to the MULF, which isn't enough for a tablet with its screen on. My plan for audio integration is already taken care of - I've got a Tune2air wma1000 I'm currently using with my iPhone that will also work with the N7 for audio and wheel controls. Easy-peasy.
__________________
Last edited by alexwhittemore; 02-09-2014 at 05:42 PM.. |
|
Appreciate
0
|
02-11-2014, 03:42 AM | #4 |
Lieutenant Colonel
113
Rep 1,951
Posts |
SUCCESS! (Kind of)
After reading a good 50% of WDS, and a bunch of other support/training material, I found that the signal I'd really like is called Terminal R (which roughly and probably falsely stands for Radio). Terminal R is active any time the key is in the slot, and inactive any time it's not. You can test Terminal R by your brake pedal - if R is active, pressing the pedal will activate your brake lights. If it's not, your brake lights won't come on. More on that later, turns out it's important.
This is more or less what I'd like - the Nexus 7 can really come on at any time from unlocking the car on, I don't much mind, but MUST turn off when the key is removed and the car is left. I don't want to walk away with a still-glowing tablet. This is the biggest disadvantage of the amp remote signal - it's on for a full minute after the car is left alone. Thus, I lock and walk away, and the tablet would turn off a minute later. Not ideal. Additionally, it comes on ANY time the car is woken up. For example, if I lock my car to double-check that I remembered to, even if it's already locked, the amp remote will come on for a minute and change, thus lighting the tablet. On the other hand, Terminal R has a disadvantage too - if a passenger were in the car without the key and wanted to listen to the radio, they can click the radio on, but Terminal R will still be off. Thus the radio will be on, but the tablet will be unpowered. I deem this acceptable, since in such a situation the passenger could simply turn the tablet on manually. It will not be externally powered, but any connected accessories will be (separate supply, on Terminal 30g), and will be usable. Not QUITE ideal, and perhaps unacceptable for a hard install where the buttons are inaccessible, but fine for me and for now. This is much more desirable than the tablet remaining on unattended. At worst, an override switch could be installed. Okay, so what about this Terminal R? It turns out, Terminal R isn't a real terminal. The status of Terminal R is signaled only by CAN bus messages. There's no wire to tap to see whether it's active. or is there? There are a couple of suspicious signals I found in WDS: R_BLS, R<KUPPL, R_MRS and R_SBE. R_MRS is a Note they're all explicitly prefixed R_, I wonder what that could mean. Spoiler alert, it seems to mean they're derived from R and follow it's timing (at least, as far as I can test and tell). R_MRS feeds the Multiple Restraint System, presumably as an arming signal (I suppose for reliability, though I'm not sure why any of these signals are physically present). Anyway, that one's out. Call me superstitious, but there's no chance I'm touching anything going into or out of the module that controls how/whether I survive a crash. More importantly, there's no way to test that I haven't messed something up. I'd hope it'd be redundant enough to throw a code, but I wouldn't bet my life on it! So what about the others? R_SBE feeds the seat occupancy detector. Again, related to airbag function, but mostly I'm sure it's not conveniently located. Not that I even checked, because R<KUPPL looks great! This feeds the "clutch switch module" - a module used to determine whether your clutch pedal is pressed. I have an AT car, so the idea of a totally unused connector hanging around waiting to be plugged into sounded great. Unfortunately, after pulling off the pedal trim and hunting around, I couldn't find anything. I suspect the connector is populated and still exists on the harness, but at best it's buried in a big huge cross section of wire, and I don't fit into the footwell well enough to have two hands free to go hunting around in there. Anyway, right next to where I'd hoped to find the plug for my nonexistent clutch pedal switch, the last signal, R_BLS (brake-light switch) was staring me in the face. After hitting it with the multimeter, it was exactly what I wanted (on when the key is in, off when it's out). Sadly, there's not enough room to focus with my head still in there. Anyway, it's the white plug hanging off to the right - the black switch clipped into the shiny aluminum casting next to it is the pedal sensor. Pin 1 on this plug (purple wire) is the R_BLS signal I am interested in. This pic is a little better, since I didn't spot it and my head wasn't in the way. I used a T tap to splice into it. Basically, it wraps around both wires, and with pliers you compress in a blade that breaks the insulation of both sides and shorts them together. Doesn't require cutting the wire, which would be nearly impossible in the cramped space. From here I routed a short lead to a spade connector, zip-tied up by the FRM for easy access. Plugged into that is a longer lead that runs right along the edge of the pedal trim into the center of the dash where the radio is installed. I'll route it from here up through the cut out for vents into the dash trim, where my power PCB will live.
__________________
|
Appreciate
0
|
09-09-2024, 12:54 PM | #5 |
Major
244
Rep 1,294
Posts |
Love this!!!! Thank you thank you! I had been searching for the same form of ignition wire for weeks. Now I found it. If you're in central PA, I owe you a beer. Thanks!
|
Appreciate
0
|
Bookmarks |
|
|