To fix this, cut the wire going from the remote-entry unit to the horn, install Molex connectors (or similar) in between, and make it so I can move the connector over to a new unit that provides a much quieter confirmation (via a piezo? Or would blinking the headlight or dome light be better?). (of course, make it so I can easily move the Molex back, to quickly restore the original functionality)
Hopefully this also fixes the issue that it's impossible to tell when the remote-entry unit receives my second tap of the unlock button, and eventually decides to unlock the the passenger doors as well.
Note: I know how to do this with a microcontroller, but that's overkill. Try to figure out how to do it with only a capacitor, transistor, and a few resistors.
Other times, I want to make it appear that I'm hitting the brakes, when I'm really not. (to communicate with the person behind me; or to fuck with someone who is intentionally riding my bumper) [again, I often accomplish this by tapping on the brake pedal so lightly that it doesn't slow the car down at all. However, this is suboptimal in several small ways. It's easier to do "Morse code" with the toggle switch than with your foot. And sometimes it's unsafe to accidentally actually brake (eg. when that semi was riding my bumper that one time)]
So, install a momentary/return-to-center three-position toggle switch. Holding it in the up position will force the brake lights to off, regardless of what I'm doing with the physical brakes. Letting go will return to the center position, where the brake lights act normally/stock/legal. Holding it in the down position will force the lights to on, regardless of what the physical brakes are doing.
(yes, in addition to this being an evil idea, it may be an illegal one as well, even though it only enhances existing methods that already work somewhat well)
(there are COTS solutions available, so at a bare minimum, find out whether one is available that meets my requirements, and compare its cost to the cost of engineering a new one myself)
If I ever get a motorcycle again, this feature will be a MUST, not just a nicety, so remember the details of how I build this for the car, and consider designing it in a compact form factor that would be suitable for burying beneath a motorcycle's dash.
2. When the key is in the off position, it's not possible to roll up the windows. Even if it's not possible to roll DOWN the windows, it should be possible to at least roll them UP. (though this has the exact same safety considerations as above)
When driving in conditions where it's important to conserve your night vision (eg. night racing), but where it's important for the passenger to be able to read (eg. route notes), it's preferable to have a passenger light that 1) spills as little light as possible towards the driver, and 2) is a color that preserves night vision. Red LEDs, being relatively cheap, are the perfect solution. (though doesn't the military use green light for this?)
I calculated how many very bright LEDs I could jam up in this particular hole, and though I'll be running perhaps 200% the amperage that the light switch and supplying wires are rated for, I think I'll be okay. All that's really left to do is figure out how to solder all those LEDs in a tight little space.
Phase Two would be to develop a driver that runs the LEDs at 110% max rating on a low duty cycle (this is actually part of LED specs), to make the perceived brightness be even higher.
Instead of having a dinky flimsy thing sticking out of the cigarette lighter, why not have beefy 5V switchmode powersupply under the dash somewhere, that provides its output via several USB (sans data) plugs? It should be able to supply more power than the cigarette lighter is rated for as well.
(this is also the first waypoint on my path towards a rats-nest of USB perhipherals embedded under the dash, providing lots of sensors for any laptop that wants to plug in)
Phase Two of the microcontroller code would include the ability to precisely count the time in between each pulse, so that when the speed is low, it's still able to accurately calculate the instantaneous velocity. (or, depending on the hall effect sensor's output waveform, perhaps give a precise measure of the waveform's width as the magnet swings by)
Text input would be via... a normal keyboard at first. But I also have dreams of embedding an 8-key chorded keyboard into the grips of my steering wheel.
A circuit that drives a car ignition coil at a tunable frequency (somewhere around 3kHz) squarewave, at 12v and decent amperage. This is the "hello world" of high-voltage electronics — it's often people's first high-voltage project.
I've already worked a little with this single-coil driver. Tweak that until it works well, and understand as much as possible about the circuit. Possibly use a microcontroller or even computer sound card as the signal generator (the needed frequencies of ~3khz are well within the range of human hearing that sound cards focus on).
Enhancements:
(or, is it possible to implement this in my cell phone(Java), or anything else I might carry with me at all times?)