Power steering - more feel required!

[V Mad]

imported post

John, the pump is just a power supply for the system which is designed to respond in proportion to steering torque inputs, not to the power supply. It is a classic linear control system.

Think of it like this: the output of a transistor radiois designed to respond in proportion to the position of thevolume control, not to the power supply (battery).

[Ian Durrant]

imported post

Don't transistor radios tend to get quieter as the batteries run down?

I don't think the amplifier analogy is helpful anyway, linearity in an amplifier is way different to linearity in a flow and pressure based hydraulic system.

[GTJones]

imported post


Power Steering
[align=center]






Pump
The hydraulic power for the steering is provided by a rotary-vane pump (see diagram below). This pump is driven by the car's engine via a belt and pulley. It contains a set of retractable vanes that spin inside an oval chamber.[/align]
As the vanes spin, they pull hydraulic fluid from the return line at low pressure and force it into the outlet at high pressure. The amount of flow provided by the pump depends on the car's engine speed. The pump must be designed to provide adequate flow when the engine is idling. As a result, the pump moves much more fluid than necessary when the engine is running at faster speeds.

The pump contains a pressure-relief valve to make sure that the pressure does not get too high, especially at high engine speeds when so much fluid is being pumped.


Rotary Valve
A power-steering system should assist the driver only when he is exerting force on the steering wheel (such as when starting a turn). When the driver is not exerting force (such as when driving in a straight line), the system shouldn't provide any assist. The device that senses the force on the steering wheel is called the rotary valve.

The key to the rotary valve is a torsion bar. The torsion bar is a thin rod of metal that twists when torque is applied to it. The top of the bar is connected to the steering wheel, and the bottom of the bar is connected to the pinion or worm gear (which turns the wheels), so the amount of torque in the torsion bar is equal to the amount of torque the driver is using to turn the wheels. The more torque the driver uses to turn the wheels, the more the bar twists.

The input from the steering shaft forms the inner part of a spool-valve assembly. It also connects to the top end of the torsion bar. The bottom of the torsion bar connects to the outer part of the spool valve. The torsion bar also turns the output of the steering gear, connecting to either the pinion gear or the worm gear depending on which type of steering the car has.








Animation showing what happens inside the rotary valve when you first start to turn the steering wheel
As the bar twists, it rotates the inside of the spool valve relative to the outside. Since the inner part of the spool valve is also connected to the steering shaft (and therefore to the steering wheel), the amount of rotation between the inner and outer parts of the spool valve depends on how much torque the driver applies to the steering wheel.

When the steering wheel is not being turned, both hydraulic lines provide the same amount of pressure to the steering gear. But if the spool valve is turned one way or the other, ports open up to provide high-pressure fluid to the appropriate line.

It turns out that this type of power-steering system is pretty inefficient. Let's take a look at some advances we'll see in coming years that will help improve efficiency.

More fuel on the fire

gary

[jpyke]

imported post

Ian Durrant wrote:

Don't transistor radios tend to get quieter as the batteries run down?

I don't think the amplifier analogy is helpful anyway, linearity in an amplifier is way different to linearity in a flow and pressure based hydraulic system.

Ian, yes they do, but they start clipping. their gain doesnt change due to the reduced supply voltage. The gain is fixed by design.

Sorry if this is too simplistic, not sure where to pitch this so may be way out!

I think the amplifier is actually a good analogy. Sticking with this for a moment, considera simply amplifier with a gain of 2 and +/-10V supply rails. Supplying up to +/-5V input will result in linear output. Example for +/-4V input:



But now, if you apply +/-6V input the amplifier will still attempt to amplify with a gain of 2 but clips when it hits the supply rails:



You'll now get a distorted output with more high frequencies.

Returning to the power steering, the pump would be the supply rails, the drivers effort on the wheel would be the input voltage and the stiffness in the torsion bar would be the gain. The pump has to be capable of supplying full assistance to turn the wheels whent the car is stationary and the engine is idling, so by inference must have way more pressure than is required when travelling along because (a) less effort is needed when the vehicle is moving and (b) the engine is running faster.

The torsion bar is spring steel so the amount of torsion is proportional to the input deflection in a linear manner (as you're not getting anywhere near the elastic limit). The problem of reducing pump pressure is that you get the same effect as illustrated above - you'll make the steering nonlinear which would feel very strange and could even be dangerous. You'd certainly end up with little or no assistance when the vehicle is stationary. Speed variable pump pressure would still create a non-linear system so that wouldn't create predictable results either.

[V Mad]

imported post

jpyke wrote:

A solid pin sounds like a good idea - may I ask what you used? I'm guessing it would have to be stainless or some form of spring steel to avoid premature wear?

I just used what was in my scrap box. Stainless is not the hardest steel so not ideal. Mabye silver steel would be better? Problem is getting one with a precision fit; you would probably need to get one specially made to do it properly.

[V Mad]

imported post

Ian Durrant wrote:

Don't transistor radios tend to get quieter as the batteries run down?

I don't think the amplifier analogy is helpful anyway, linearity in an amplifier is way different to linearity in a flow and pressure based hydraulic system.

As jpyke explained, when the batteries run down, the electronic circuit starts to malfunction. The same happens in the steering system if you reduce the pump pressure. If you want to try it, dont tell your insurance company

[John Wright]

imported post

If taking out the pump at speed doesn't provide any steering feel, then this implies that the steering geometry (caster, camber, toe in) are not well designed on the Stag.

The Stag was expected to sell well in the USA, and probably would have done if it had been reliable. This may have resulted in the steering being matched to be closer tothe US cars that have absolutely no feel at all.

John

[V Mad]

imported post

I think to some extent this is true John. Even when new the Stag was critcised for lack of feel by some motoring journalists. But when you get one that has all steering system components in good condition, they can be quite nice. Ask Surreyone, he is delighted with his since I replaced his suspension ball joints, and inner and outer track rod ends. Don't forget these cars are up to 40 years old now, and some arenotin tip-top condition.

[Gistan]

imported post

John Wright wrote:

If taking out the pump at speed doesn't provide any steering feel, then this implies that the steering geometry (caster, camber, toe in) are not well designed on the Stag.

The Stag was expected to sell well in the USA, and probably would have done if it had been reliable. This may have resulted in the steering being matched to be closer tothe US cars that have absolutely no feel at all.

John

John

I think you have identified an important point re the steering geometry ie camber castor and toe-in, I took my Stag to Capesthorne Car Show on Sunday and got into a discussion with a guy who had restored a Gordon Keeble and also owned a Stag.

He was very interested in my slight modifications, and asked how I found the steering, due to the reduced weight of the RV8 compared to the TV8.

I commented that the steering was lighter with less feel than than my previous Stag (TV8), and was considering converting to a manual set up from the Triumph 2000.

He stated that slight changes in castor angles can have dramatic effect on steering weight and feel, he had experimented with alterations to correct steering issues with his Gordon Keeble.

From what I understand of the Stag steering geometery the castor angle seems to bedictated by the drag strut and havung stripped and rebuilt front suspensions a few times have simply tighted up thenut that holds the drag strut to the body to an unmeasured compression of the bushes.

I wonder how many of us have re-assembled the suspension and steering ckecked the toe-in and assumed camber and castor are ok, or am Itheonly bodger :shock:

Gary

[Ian Durrant]

imported post

Ian, yes they do, but they start clipping. their gain doesnt change due to the reduced supply voltage. The gain is fixed by design. Sorry if this is too simplistic, not sure where to pitch this so may be way out!

Don't worry about the level of your pitch, I did electronics and am also a radio amateur aswell as a computer hardware engineer, the behavious of transistors is well known to me.

The difference is I don't believe the PAS steering becomes non-linear in nearly the same way. The nearest I can see to the analogy of the clipping effect of the transistor type circuit making the system non-linear is the viscosity of the fluid. If we were to say halve the pump pressure then yes, for small applied torque figures, the spool valve would only open a tiny amount and flow would be insufficient to make the assistance effective, unless you only wanted to turn the wheel very slowly. But we are not talking about massive variations in output pressure, what we are after is improved feel. so maybe 10-20% would improve feel and not become non-linear enough to cause any adverse side effects. Certainly American Hotrodders seem to have been lowering output pressure of the pump by these kind of amounts for many years.

In any event PAS already is (apparently) non linear by design. The output pressure limiting valve also has a smallish drilling to limit flow (from what I have read) and this means that you don't get the same effect if you turn the wheel slowly as if you turn it rapidly, i.e when turning it rapidly you get less assistance because of the flow control which means its harder to yank the wheel and get a rapid large steering input so rapidly at speed the car is easy to get out of control.

I am not really arguing here about the way the system works, just that small reductions in output pressure can possibly help feel by slightly (10 - 20%) lowering assistance, which is the part being disputed. Some say this won't work, some of us can't see why it won't given that American hotroders do it, monarch sell a pump that claims to do it.

Frankly I don't see the point in arguing it won't work unless you have either tried it, seen it tried or have a convincing argument as to why people that are already doing it are wrong and don't know it.

[jpyke]

imported post

Ian, I agree about the arguing

However, I was all for buying the Monarch kit a little while ago and posted asking if anyone had tried it and immediately got the response that it would not work so this set me thinking. I didn't see any responses from anyone who'd tried it who'd said that it did what they wanted and if you look at Johnboy's post (the 2nd post in this topic) you'll see that he has tried this and it didn't do what he wanted.

What I'm keen to avoid is little or no assitance when the vehicle is stationary or slow moving. If this were the outcome then I may as well go for a Triumph 2000 manual steering rack. I've driven the stag without the belt connected to the pump and have to say that it was bloody awful although I will accept that PAS racks don't behave the same as manual ones when there is no pump pressure as you're having to drive the hydraulics as well as the steering.

So, as you say, the proof of the pudding is in the eating so what I plan to do is to experiment with some spring steel, the lathe and mill and the spare spool valve that I have and try making a stiffer torsion bar (this is if I get no positive response from Adwest). If nothing else it will be interesting to see what happens and if there is any improvement. I expect if there is I'll have to go around the buoy a few times to get the feel just right but I'll be sure to post the results so everyone can find out.


It's a crazy idea but it may just work...


James