Variable pressure (?) power steering pump?

[69427]

Helped a buddy the other week replace the p/s pump on his wife's car (fwd Cavalier or something). The p/s pump had an electrical connector on it. I'm curious what it does. Does it allow the pressure to be modulated by the ECM? I'm curious about a custom system that would open a bypass in the pump to reduce the pump drag at WOT or near WOT, and also possibly help reduce the fluid temp at high RPMs. Can anyone enlighten me on these pumps?

Thanks.

[BBShark]

That is an EVO valve. It regulates the bypass pressure based on vehicle speed and handwheel rate ( full boost for evasive manuvering). I have looked into making a controller for an EVO valve and it would be pretty easy for the vehicle speed but a little more difficult to package the steering wheel rate sensor.

My feeling is that it could be done with a 2 axis accelerometer.

[mrvette]

Originally Posted by BBShark

That is an EVO valve. It regulates the bypass pressure based on vehicle speed and handwheel rate ( full boost for evasive manuvering). I have looked into making a controller for an EVO valve and it would be pretty easy for the vehicle speed but a little more difficult to package the steering wheel rate sensor.

My feeling is that it could be done with a 2 axis accelerometer.

What does this bring to the table???

[denpo]

Originally Posted by BBShark

That is an EVO valve. It regulates the bypass pressure based on vehicle speed and handwheel rate ( full boost for evasive manuvering). I have looked into making a controller for an EVO valve and it would be pretty easy for the vehicle speed but a little more difficult to package the steering wheel rate sensor.

My feeling is that it could be done with a 2 axis accelerometer.

As we spoke here :
[Only registered and activated users can see links. Click Here To Register...]
The steering sensor problem remains.
Accelerometer might be a problem soo many vibration to screw the acceleration integration. A magnet on one wheel with an hal sensor would be more reliable.
....
That make me think about a post on hackaday where a guy was using a laser mouse to sense the speed of a robot by shining at the carpet below it.
You could sense the steering column, provided you can keep it out of the dust with a casing and give the sensor a nice uniformly surface to shine at.
It's dirt cheap, works out of the box and interfaces nicely by usb or serial.

[BBShark]

Originally Posted by denpo

As we spoke here :
[Only registered and activated users can see links. Click Here To Register...]
The steering sensor problem remains.
Accelerometer might be a problem soo many vibration to screw the acceleration integration. A magnet on one wheel with an hal sensor would be more reliable.
....
That make me think about a post on hackaday where a guy was using a laser mouse to sense the speed of a robot by shining at the carpet below it.
You could sense the steering column, provided you can keep it out of the dust with a casing and give the sensor a nice uniformly surface to shine at.
It's dirt cheap, works out of the box and interfaces nicely by usb or serial.

The accelerometer actually works well. Transients (vibration) can easily be filtered out. Also hydraulics systems are inherently dampened.

I did this with a Wiimote and plotted the vectors. You can actually draw pictures in the air with it (that is what we were trying to do).

[vette427sbc]

Could the VSS from a newer trans (ie: T56) be used for speed sense, and then an ABS or similar sensor & reluctor wheel (not sure if thats the right terminology or not) be mounted on the steering shaft or rag joint area?

[BBShark]

The VSS could be used for speed sensing and the encoder for the steering wheel could be adapted. Ford makes a standalone module that could provide an output for the rate solenoid (on the pump). However, I think the accelerometer based control would be easier to do. Hell, it's already done. Take a look at products like the G-analyst

[69427]

I'm curious about at what speed "threshold" that the pump assist isn't needed. (I assume it depends on the car weight, tire footprint, and the personality of the steering rack/box) My interest in this sort of project is just for track days, and it would be handy if the assist wasn't needed most of the time, both to free up a couple horsepower and keep the p/s fluid temperature down a bit. IIRC I have ABS wheel sensors in the front of my car (the front suspension is a '96), so it would be easy to build a F/V converter feeding an adjustable threshold comparator for when to disable or enable the assist.
A MAP sensor input might also be usable as this would indicate (depending on the selected comparator threshold) if the car was accelerating hard (indicating that the car was most likely in a straightaway) and the assist could be disabled. The MAP reading could also be used to override the speed input so that there is always assist under part throttle conditions (such as cornering).

There's several questions I don't have answers for yet.

1) What is the worst case horsepower drain to run the pump at full assist?
2) What is the horsepower drain to run the pump when the bypass is open (ie: is this project worth the time and effort)?
3) What kind of signal and amount of current does it take to drive the solenoid?
4) Can the solenoid be turned on 100%, or does it require some PWM control?
5) What GM cars came with this type pump (I need to retrofit a v-belt pulley to it for my application)?

[daveL82]

There's a company out of Denmark called something like EZ steering. Not sure how much but they take stock steering columns and fit their electric motor/controller to add pwr steering to vintage cars. Nice thing is the motor goes on the column under the dash so you can remove the pump and all hoses associated. The also have a speed sensor that goes in the speedo cable so the system is variable assist. Not sure how much or if they have a C2/C3 option.

[69427]

Scoured a junkyard yesterday looking for a pump that ideally has three traits:

1) Aluminum body
2) EVO valve
3) Shaft size allowing swap to vee-belt pulley

Pretty much struck out finding one meeting all three traits, so I might just focus on an EVO pump that I can somehow get a vee-belt pulley to fit. IIRC, the old pumps I have use a shaft size of about .660" while the serpentine pulley pumps look to be closer to .720. One upside, the two wire harness to the solenoids in the cars I looked at used very small wires, making me think the solenoid current was reasonably modest. That makes life easier when sourcing a switching transistor for the PWM circuitry.
When the weather cools down a bit more I'll take a trip to one of the other junkyards to see what they have out in the yard.