power steering hacking

[V8Cobrakid]

keep in mind that VSC is Vehicle Skid Control and also funtions to steer the car in case of a slide ( give extra torque in either direction)

[patsparks]

Originally Posted by ken1784

Hope your application will have a very good fail-safe feature, or people will be killed.

Ken@Japan

Wouldn't the failsafe be that there is a person in the drivers seat keeping a vigilant lookout? BMW have a system under developement where the car can drive itself.

[anthony.pray]

Yah!!, looking over TIS I have found the torque sensor it is the 4 wires in the green plastic.
thay are as fallows
TRQ1 - Torque sensor 1
TRQV - Power
TRQ2 - Torque sensor 2
TRQG - Ground

TRQV - outputs 7.5 to 8.5 V

TRQ1 and TRQ2 need to have a 0.3 V or less difference.

Hear its the best part.
If
TRQ1 and TRQ2 are between 2.5 to 4.7 V this tell the Power Steering ECU
to turn the car right.
If
TRQ1 and TRQ2 are between 0.3 to 2.5 V this tell the Power Steering ECU
to turn the car Left.
It looks like this should be easer then I thought.

Car is all back together for tonight.
Test for tomorrow.
1) Grab voltages for controlled steering.
2) Write simple app to simulate this back to the PS ECU.
3) Will post video if I can get it working.

Thanks to V8Cobrakid for the grate link!!!

[V8Cobrakid]

yw. that does sound rather simple. that means that it's veriable where 2.5 is center?

[anthony.pray]

Originally Posted by ken1784

Hope your application will have a very good fail-safe feature, or people will be killed.

Ken@Japan

The plan is to use relays that are normal set to factory operation wiring.
so in a case of failure the power is removed form the relays allow for normal operation to be restored.

There will be much more to come.

Remember this is just a "prototype" to show that the Prius could be a great platform for people, even DARPA.
The great thing about the Prius is excluding the steering weal. form what I have been told the whole car is "drive by wire" so by finely getting the steering to be "by wire" you could very ease connecting in with a PC and drive the car... Without any new servo and actuators. The car already has them there. The reason for not making the steering by wire is, one word power, or lack of it. now are the breaks by wire or just supported by wire?

[ken1784]

Originally Posted by patsparks

Wouldn't the failsafe be that there is a person in the drivers seat keeping a vigilant lookout? BMW have a system under developement where the car can drive itself.

toyota.jp ƒNƒ‰ƒEƒ“ ƒnƒCƒuƒŠƒbƒh > ˆÀ‘S�«”\
Toyota is already selling Crown in Japan with the Lane Keeping Assist.
We know Toyota can supply enough guarantee and have capability to pay millions of dollers for reparation about accidents.

Ken@Japan

[anthony.pray]

Originally Posted by V8Cobrakid

yw. that does sound rather simple. that means that it's veriable where 2.5 is center?

yah!, well the center is 2.3 to 2.7 V

[kohnen]

Originally Posted by anthony.pray

The plan is to use relays that are normal set to factory operation wiring.
so in a case of failure the power is removed form the relays allow for normal operation to be restored.

If you plan on using relays, then you must be aware of this:

As relays vibrate, the contacts that you think are closed aren't always - they will bounce around and chatter a bit due to the vibration of the car. This could cause some unexpected results.

This means when you have the relays on and think that you are using your alternative design, you will get intermittent opens.

When the relays are off and you think you are using the factory operation, you will get intermittent opens too.

You will likely confuse the poor computer (likely causing immediate power steering shutdown due to a mis-match between the torque sensors).

Contrary to what you might think, relays aren't a good way to make good solid electrical contact for computer signals.

That said, good luck and be careful!

[ken1784]

Originally Posted by anthony.pray

The plan is to use relays that are normal set to factory operation wiring.
so in a case of failure the power is removed form the relays allow for normal operation to be restored.

Failure is not only a power fail, but also in many places.
I think you'll use CCD camera to detect highway lanes.
A CCD has possibility to detect an artifacts or smear. You'll have to detect your lanes without any error.
Smear - Wikipedia
How are you going to detect correct lanes and to eliminate the smear?

Remember this is just a "prototype" to show that the Prius could be a great platform for people, even DARPA.
The great thing about the Prius is excluding the steering weal. form what I have been told the whole car is "drive by wire" so by finely getting the steering to be "by wire" you could very ease connecting in with a PC and drive the car... Without any new servo and actuators. The car already has them there.

JP and EU Prius already have IPA, or Intelligent Parking Assist, and we can show Prius performance to people without any additional servo and actuators.
And, it is in a walking speed and no one killed.

We don't want anybody killed including yourself.

Ken@Japan

[anthony.pray]

Originally Posted by ken1784

Smear - Wikipedia
How are you going to detect correct lanes and to eliminate the smear?
Ken@Japan

I was waiting for this post, let me make this clear right now.
my plan is to get my car to drive a farley strait line with out my help. thats it..

I am playing around with 2 different libraries.

BlobDetection by v3ga
Myron by Josh Nimoy
The code is still very much in work but it gets closer every now and then.(I know....). Programing in Java, well it can be fun.I am hoping to set up a test reading of the Torque sensors, when the car is at highway conditions . Using diode porbs there is no chance of feedback in to the ECU.