Turning on Mini PC from car ignition without a micro-controller

Thread Starter

Andrew Lapham

Joined Jun 1, 2018
11
Hey peoples,

I have a small NUC that I'm going to be hiding in my car but the BIOS doesn't have the feature to turn on the device when powered.

has anyone got any ideas on how I can use the car's power on ignition to signal the board's power switch contacts?
I know I can do this easily with an Arduino using a transistor to call high, low signals but I'm looking into doing it with without the help of microcontrollers.

cheers,
 

Yaakov

Joined Jan 27, 2019
1,612
Look for 12V time delay relays. These are common and can be set to turn on for a fixed period after seeing power. The bigger question is how to gracefully shut down to avoid corruption of the filesystem.
 

Thread Starter

Andrew Lapham

Joined Jun 1, 2018
11
yeah, I was just looking at some of those on eBay. surely there would have to be a "jank" way of achieving this with a few components yeah?
shut down would not be an issue as I'm using the device through SSH so I can just send a shutdown command from my phone.
 

Thread Starter

Andrew Lapham

Joined Jun 1, 2018
11
yeah, I was just looking at some of those on eBay. surely there would have to be a "jank" way of achieving this with a few components yeah?
shut down would not be an issue as I'm using the device through SSH so I can just send a shutdown command from my phone.
If all else fails I could just wire a switch into the dash but that's the easy way, haha
 

ebeowulf17

Joined Aug 12, 2014
3,032
I'm just about to start work and don't have time to sketch it, but I've seen very simple circuits that I believe would work (not as simple as a pre made timer module, but cheaper and maybe smaller.)

One capacitor, diode, and resistor to convert off->on transition into a single positive pulse, then a Schmitt trigger inverter, a cap, and a resistor to provide a short, timed high signal before going low again.

If no-one else draws anything similar, I'll draw it up tonight.
 

ebeowulf17

Joined Aug 12, 2014
3,032
Ok, this was a bit rushed and it's already past my bedtime. Sorry. Pretty simple circuit and should do what you want, if I'm understanding your needs correctly...

The basic idea here assumes that you can have both the constant 12V battery power from the car and the switched "accessory" power that tells you when the car is running and the device should kick on. The Schmitt trigger inverters will need to be powered from the constant 12V source, and then the accessory power is the input. The values of C1 and R1 determine how long the positive pulse at the output lasts. Higher values for either one translate into a longer pulse. The circuit output doesn't do anything then accessory power switches off.

So, I think with one cap, one diode, one resistor, and one Schmitt trigger inverter chip you can do what you need...

Although even as I type this I realize if the inverter is powered by 12V, it will output 12V, and you probably need lower signal levels for your microcontroller. We could either do some sort of level conversion to handle that, or switch to running the inverter on the same supply voltage as your microcontroller, then adding more input protections to guarantee that the inverter stays healthy. Hmmm... I've really got to get to bed, but I'll be happy to give this some more thought tomorrow if no-one beats me to it.

schmitt-trigger-pulse-on.png
 

Attachments

ebeowulf17

Joined Aug 12, 2014
3,032
thank you for all the information! after a little googling i also came across this schematic for exactly what i'm after but it seems yours is far more simple to put together.
http://solorb.com/elect/bootcirc/PCautoboot/
I should point out that, although I'm learning quickly and getting better at the conceptual problem-solving stuff, I have very little experience with automotive circuits, which are notoriously noisy.

Having said that, here are some key points about their schematic vs. mine:

1) They show power supply decoupling caps which I didn't include in my schematic. Those caps are important. Nearly every IC that does and kind of digital or logic switching needs them. Mine were omitted simply because it's fairly common when sharing concepts (as opposed to complete, finalized, production-ready schematics) to assume the person receiving the schematic will know to include them. It should be automatic.

2) They are providing the option of either using a relay for the output, or getting a logic low level as the output pulse. I don't know your exact device requirements, but I suspect that those options make more sense than my logic high output. It's probably a good idea to use the same output stages they're suggesting, with a resistor and transistor as the minimum, and maybe using a relay and corresponding flyback diode as well.

3) Beyond the parts mentioned above, much of the remaining complexity is because their circuit creates a sequence of two separate timed pulses, first sending a power on signal, then sending a reset signal. Again, I'm totally unfamiliar with your controller and what it needs - do you need a two stage power up sequence, or will a single pulse signal do the job? If you need two, their circuit looks pretty logical to me. If not, I think mine is probably ok (with the decoupling caps and transistor output stage discussed above.)
 

ebeowulf17

Joined Aug 12, 2014
3,032
I should point out that, although I'm learning quickly and getting better at the conceptual problem-solving stuff, I have very little experience with automotive circuits, which are notoriously noisy.

Having said that, here are some key points about their schematic vs. mine:

1) They show power supply decoupling caps which I didn't include in my schematic. Those caps are important. Nearly every IC that does and kind of digital or logic switching needs them. Mine were omitted simply because it's fairly common when sharing concepts (as opposed to complete, finalized, production-ready schematics) to assume the person receiving the schematic will know to include them. It should be automatic.

2) They are providing the option of either using a relay for the output, or getting a logic low level as the output pulse. I don't know your exact device requirements, but I suspect that those options make more sense than my logic high output. It's probably a good idea to use the same output stages they're suggesting, with a resistor and transistor as the minimum, and maybe using a relay and corresponding flyback diode as well.

3) Beyond the parts mentioned above, much of the remaining complexity is because their circuit creates a sequence of two separate timed pulses, first sending a power on signal, then sending a reset signal. Again, I'm totally unfamiliar with your controller and what it needs - do you need a two stage power up sequence, or will a single pulse signal do the job? If you need two, their circuit looks pretty logical to me. If not, I think mine is probably ok (with the decoupling caps and transistor output stage discussed above.)
Oh yeah, I almost forgot: mine relied on having both constant and switched power available, where theirs appears to be ok just running directly on the switched (accessory) power. So that's one more point in their favor.
 

ebeowulf17

Joined Aug 12, 2014
3,032
Ok, the other circuit you posted may well be the better solution, but I've incorporated some improvements to mine just in case it helps.
schmitt-trigger-pulse-on_03.png
I've added the power supply decoupling caps - although on the schematic they're near the power source in order to keep things un-cluttered. In reality, the decoupling caps need to be as close as possible to the inverter IC power pins.

I also added the open-collector output stage which is probably more suitable for your needs, as well as the optional relay output stage, which should really cover all likely output needs.

Cheers!
 

Attachments

Top