[insert name of project] worklog

**tripzero** · 09-16-2010, 11:20 PM

Amidst all the clutter, beyond all the obstacles, aside from all the static, are the goals set. Put your head down, do the best job possible, let the flak pass, and work towards those goals.

-Donald Rumsfeld

Mp3car was kind enough to sponsor an experiment in what I perceive as the future of car computing. I hope to turn the results of this worklog into something that everyone in the forums can reproduce. However, before I dive right into what this project is all about, lets discuss the why.

Causality-there is no escape from it, we are forever slaves to it. Our only hope, our only peace, is to understand it. To understand the "why." "Why" is what seperates us from them..."Why" is the only real source of power, without it you are powerless.
MEROVINGIAN, The Matrix Reloaded (2003

There are issues in todays car computers. Forum members are adept at solving many of these issues but many of them are not addressed. Here are a few that provoke this experiment:

1) Car computers are too big. A measurement of 6" x 10" (or bigger) usually means you shove the carpc in the trunk. Mac mini's were small enough to shove in the dash or in the glove compartment, but...

2) Car computers consume too much power. Law of conservation of energy states that energy really can't be created or destroyed. It just moves around in different forms. In a carpc, the electricities you put in your computer usually end up as heat. Tight spaces without much circulation such as the dash or the glove compartment make heat problems worse. This causes hardware to act funny and software to crash.

In addition to heat, there is another problem with high power consuming computers. Electricity in your car is usually transformed from kinetic energy when your engine spins your alternator. This requires gas. Gas costs money. In addition to the extra cost, there is also the risk of killing your alternator because of the extra load, or draining your battery to the point where it no longer holds a charge. The dead battery has been such an issue from the beginning that smarter power supplies were created to address the issue. Today, a standard carpc can only remain running for a few hours before it's killed by the smart power supply as to avoid draining the battery below what would be required to start the car. This short lifespan when the engine isn't burning gas and turning it into electricity brings us to the third problem:

3) Short offline lifespan limits the potential of the carpc. If you've been privy to any of my rantings, I'm a big sucker for things you can do offline (ie, when the engine is off). With the right software, I can have the car doing security checks, making sure the doors are locked, lights are off and alert me if any of that changes. If I have a webcam, I can take snapshots any time someone moves around my vehicle. If security isn't your thing, what about automation? By monitoring the presence of my cell phone, I can have my car unlock when I walk up to it. I can have the car started before I even reach my keys in my pocket.

4) CarPCs are too slow. There are loads of threads on this forum about operating system optimization, frontend optimization and how to get the most out of your software. For some, it takes a full minute to cold boot to their frontend of choice. Lots of software seems to be built with core 2 duos or i3-7's in mind. Or even worse, software makers expect you to resume from hibernate which itself takes some time.

The goal of this project are to solve the above issues. In subsequent posts, I will try to explain how these get solved and hopefully define the project in a clear way.

**tripzero** · 09-16-2010, 11:21 PM

As mentioned in Part I, the goal of this project is to solve the issues that limit or frustrate car computing. This is just one attempt. There are likely many solutions to the problems. This is the approach that I think will solve all of them. First, lets lay down some goals. The main goal of this project is to build a car computer that others can reproducibly make that will:

1) Fit a basic set of hardware components required for a carpc into the dash.
2) Consume very little electricity and produce very little heat
3) Stay on for long periods of time without draining battery
4) Instant-on capabilities and smart power management.

**tripzero** · 09-16-2010, 11:22 PM

Here's what we have to work with:

- Bybyte double din case
- igep V2 RevC 720MHz ARM, 512MB RAM, Wifi, Bluetooth
- DCDC-USB PSU @ 5V with standard smart battery monitoring/cutoff safeguards.
- Boomzbox HD Radio
- Lilliput 629 with DVI input
- Powered USB Hub.

The bybyte case is the obvious choice for fitting a carpc in the dash. It works with the lilliput eby701 and the 629 7" touchscreens.

The igep V2 is a controversial choice. It's ARM based which means you are limited by what software will run on it. ARM means windows CE or Linux as the OS and any open source frontend that will run on it. That limits your software considerably, but we will talk about that more in another section. This board is ideal for this project because 1) it's tiny, 2) it consumes only a few watts and 3) it comes with the bare essentials onboard (wifi, bluetooth and built in RAM). For equivalent hardware, you usually end up spending a bit more.

The power supply is the dcdc-usb. It produces very little heat and can take a very wide voltage input (which is good in the car). It also has a usb interface to read input and output voltages.

Any standard headunit will be able to play the radio. We want our system to at least meet those capabilities and offer more because heh, it's a computer not a headunit. Because the igep is so small, we have extra room for more devices. Mp3car provided me with a boomzbox and I'm going to jam it in the box and make some boom with it.

Finally, we need something to interface with the human part in the car. Because the igep has hdmi out, we need a lcd touchscreen that is capable of such an input. It also needs to fit in our case. The Lilliput 629 is the perfect choice and meets those requirements.

Here's a pic of the hardware we need to be able to fit in the bybyte case:
Name: 20100916_001.jpg
Views: 1767
Size: 56.1 KB

Name: 20100916_001.jpg
Views: 1767
Size: 56.1 KB

Just as a proof of concept, I solved the problem of making it all fit here:
Name: 20100916_002.jpg
Views: 1709
Size: 63.1 KB

We also need to fit the cables:
Name: 20100916_003.jpg
Views: 1657
Size: 46.0 KB

And for bonus points, fit also the following hardware:
- Fusion brain
- 2x Hall sensors
- 1x temperature sensor
Name: 20100916_004.jpg
Views: 1651
Size: 40.5 KB

Name: 20100916_004.jpg
Views: 1651
Size: 40.5 KB

The fusion brain addition will allow us to better monitor temperature, our current usage and the current usage of accessories. I may also throw in a few relays to turn off the powered hub and lcd screen if I need to and have the room.

**tripzero** · 09-16-2010, 11:22 PM

This is probably the most challenging part of the whole project. But it's also the part I will enjoy the most. Here's what the software must be able to do:

1) The OS must boot quickly and get out of my way.
2) must be able to play music, video, internet, install apps
3) must be able to manage power

#3 is the most important to solve our objectives. We need to be able to turn on and off accessory devices, the display to save power when the engine is off. To achieve instant-on functionality, we will simply turn on/off the display similarly to how your cell phone works.

To achieve these objectives, the following software bits will be included:

- MeeGo for fast booting and highly touchscreen optimized UI
- Bluemonkey which will manage all our power managing needs.

We will probably add applications later to support the boomzbox radio and other carpc-ish fuctions, but this is good for now.

**Tidder** · 09-17-2010, 12:01 AM

Sub.

**chunkyks** · 09-17-2010, 12:36 AM

Subscribing.

**madadd** · 09-17-2010, 01:30 AM

Hi
New to this forum (how i missed this over the years is beyond me).
I have read several of your threads tripzero and am very interested in this project. I do think that software is the key.

First subscribed thread

**Ryven** · 09-17-2010, 03:59 AM

definitely want to watch this and see where it goes. An all-in-one, efficient carpc with double-din form factor? That smells like profit potential.

**Sonicxtacy02** · 09-17-2010, 08:20 AM

wow you guys managed to get this project off the ground quick. Paying close attention.

**Bugbyte** · 09-17-2010, 08:39 AM

Subscribed. A great project and one that I'm sure will generate a lot of interest.

I like your point about the power consumption. Standard in car PC's consume so much power and kill a car battery so quickly that we've had to create hardware to prevent that! I hadn't really thought about it that way, but I guess if you posted a poll asking members if their car PC had ever killed their battery even with an intelligent power supply, you'd get a very high percentage of yes votes. ...in fact, maybe I'll post one....