help needed in determining materials list

I didn't understand the exact requirement, did you intend the RPI to gracefully power off when the vehicle is not in operation?

If so, there are alternate ways that do not require the LiPo and step-up DC-DC converter - the sleepy pi would then likely meet the need.

Well, that's part of it. I also want to ensure no vehicle battery drain when the system is not in use. In addition, I need some mechanism to ensure the RPi is getting clean power in the hostile power environment present in a car. That was the thought behind the LiPO. The LiPO is meant to ensure a clean stream of power to the RPi no matter what the power condition is in the vehicle.

The requirements are:

1) Ensure that the system can completely power off such that there is not drain on the automotive battery.

2) Ensure a clean and steady power supply to the RPi.

3) Provide a mechanism to allow the RPi to shut itself down gracefully and then "wake up" when the vehicle key is switched to the "run" position. Note the IGN wire in my drawing. That's what starts the whole process.

The sleepy pi may be an option. I am unsure if it the entire system can be completely shut down...meaning does the arduino stay powered on the whole time? If so, that would cause a drain on the automotive battery. Lastly, I see that it provides the power management I'm looking for; however, I really need to be able to fend off the spikes, brownouts, and sustained voltage craziness in the vehicle. I would still need some kind of battery to provide the consistent power I'm looking for...correct?

Sat-Nav and in car entertainment equipment operate from the vehicle battery. They contain processors too, and a LiPo is not needed for cleanliness with these devices.

The sleepy pi website specifies "Extend RPi battery life from hours -> weeks / months" so I'm confident it won't drain the vehicle battery when powered off gracefully (which

is what the sleepy pi does). Basically relays, LiPo's and step-up converters are not needed - saves a lot of effort!

You should take steps to ensure the input won't exceed 17V (the sleepy pi limit) - perhaps using a varistor, TVS or both.

I have been experiencing power-related problems with my RPi project. Log messages harvested from the SD card suggest power-related issues that cause a bunch of horrible things to happen like the network interface to stop functioning. Over and over again I have come to realize that the RPi seems to be sensitive to in-vehicle power fluctuations. This is the largest impediment to my project. If I'm wrong, do you have any ideas what I may be experiencing and why? For example, when the RPi is running off of the vehicle battery and the vehicle is started, the only LED on the Pi that lights up is the second one and it stays red. Nothing else happens. The device will not function until I completely pull the power for several seconds and then re-apply power.

Regarding

" however, I really need to be able to fend off the spikes, brownouts, and sustained voltage craziness in the vehicle. "

You'd have to do that with any circuit intended for a vehicle, it is not eliminated using a LiPo - you have merely moved the problem to the input of the LiPo charger circuitry.

By the way, 17V (the max input rating for the Sleepy Pi board) is quite low for vehicle use, so perhaps you may want to use a linear regulator (e.g. a 10V linear regulator like a LM7810) preceeding it.

It wouldn't consume enough current in shutdown to worry about in a vehicle (about 6mA with typical linear regulators).

On 06/11/14 22:09, shabaz wrote:

Regarding

" however, I really need to be able to fend off the spikes, brownouts,

and sustained voltage craziness in the vehicle. "

You'd have to do that with any circuit intended for a vehicle, it is not

eliminated using a LiPo - you have merely moved the problem to the input

of the LiPo charger circuitry.

Vehicle electrical systems are a horrid environment, but just throwing a

LiPo at it isn't the solution.

 

By the way, 17V (the max input rating for the Sleepy Pi board) is quite

low for vehicle use,

Low but probably just about tolerable, as long as it's the max steady

voltage. There will be momentary spikes above that. Mind you, I once

encountered a (very expensive, German made) explicitly automotive piece

of development kit that would fry a tantalum if subjected to over 16V

for a fraction of a second.

so perhaps you may want to use a linear regulator

(e.g. a 10V linear regulator like a LM7810) preceeding it.

I'd go with a series resistor and MOV rather than a (potentially equally

sensitive) linear regulator

It wouldn't consume enough current in shutdown to worry about in a

vehicle (about 6mA with typical linear regulators).

 

Most OEMs now consider 6mA an unacceptably high quiescent current for

(an individual piece of) automotive electronics.

Hi Shabaz, yeah the proposal is some what over complicated. Also the relays would be too slow switching the power to and from the Pi so really the LiPo would be redundant.

I would suggest using a 5V 1-2A Switching regulator with  an inhibit/enable pin to control it powered directly from the permanent 12V.

you would then need to work on the logic in such away that if the IGN 12V isn't present then the Regulator is inhibited thus will not switch on. This will need to have a bypass probably based on say a 555 monostable which is periodically reset by the RPI's GPIO.

A script running will keep the power on after the key is removed only long enough to shut down cleanly....Voila!

Hi Rob,

That's a great idea, something like a 2.2R 2W would be great if the 17V is a bit too close for comfort (I didn't check to see what device the sleepy pi board uses), and a varistor as you say.

Hi John,

Exactly, I agree, it was unnecessary complications!

The sort of cheap Switching Reg I was thinking of will go to around 35V which I think is good enough for most situations in a car.

I know if the alternator is running and the  battery is removed you can get spikes higher but that really is a fringe case.