Things to do with a spare ABS valve block & pump

[FridgeFreezer]

Having replaced the brain of my Freebie's ABS unit, I find myself with a spare ABS unit, complete, but with a damaged brain. Pump and solenoids work, but the bit that does the maths is dead. I'm fairly confident I can bypass that and just fire the solenoids direct - which means 4-wheel control of something or other is possible.

First/most obvious idea is fiddle brakes, but does anyone have any other ideas for landy related projects?

[LandyManLuke]

I'd be interested to learn how the block works, both in terms of getting to the solenoids, and how the block is routed/ported.

Fiddle brakes implemented through an ABS valve block would be compact, and benefit from the master cylinder's servo assist. The other advantage is your right foot is fully trained in operating brakes, and it would leave your hands free to steer/operate winches/gesticulate at others etc.

[elbekko]

I've been toying with the idea of making my own traction control system with one of those. As long as you can read the sensors, and know what they're telling you, and know what kind of input the valve block is expecting, it should be a doddle

Once you've got that part figured out, you could have selectable maps (much like terrain response) for how quickly the system responds etc.

Oh, and you can't forget the fancy graphics on a computer screen telling you wheel rpms

[LandyManLuke]

FWIW, I've done that, though at about 10:1 scale. it did work though.

[FridgeFreezer]

I've been toying with the idea of making my own traction control system with one of those. As long as you can read the sensors, and know what they're telling you, and know what kind of input the valve block is expecting, it should be a doddle

Once you've got that part figured out, you could have selectable maps (much like terrain response) for how quickly the system responds etc.

Oh, and you can't forget the fancy graphics on a computer screen telling you wheel rpms

Elbekko - the ABS unit (when working) already does all of that, all you need is a very small micro with CANbus to tell it what to do.

[elbekko]

Oh. That's cheating

[FridgeFreezer]

Oh. That's cheating

Well, it would still require a bit of piracy of the CANbus. Also, you would need a working ABS unit. Although it's possible to make your own ECU from scratch (and yes I did briefly consider it) IMHO it would take a hell of a lot of development time to get anywhere near what the regular unit can do. It would also mean adding in wheel-speed sensors or similar, already present on some vehicles with ABS but never fitted to Series 3's

[simonr]

Instead of ABS or pure traction control - how about steer control?

This is an idea I've been rolling around for a while now.

You need a steer position sensor to work out how much steer input there is. You then calculate the relative speeds of all four wheels - and force then to be within a few % of that.

In a situation with minimal traction it should give the vehicle the best chance of executing the line of a turn as demanded. I think it should work particularly well when traction is very low or variable - the times when things tend to go wrong!

A bit like traction control - but for steering. Something you can't do with traditional difflocks.

Si

[elbekko]

Interesting idea Si. If I understand correctly, you mean something like a forced skid steer, where you calculate the difference in wheel speeds during cornering and make sure it doesn't deviate from that. This way you have a 'locked' diff, but can still corner normally.

That *should* be a fairly simple extension to a stock traction control system (if they don't already have something like that, I have a feeling terrain response for example does), as long as you have a steering position sensor as you say.

[simonr]

Yup that's exactly it. Like with all three diffs locked in a slippy surface, the truck goes in a straight line - this just sets the direction it will not deviate from with the steering wheel.

I was told by a grown-up the reason they don't do it is brake pad wear. To be useful it has to be fairly aggressive and they are worried that before you know it - you wouldn't be able to stop at all!

On separate calipers however - who cares!

Si

[Manxey]

Instead of ABS or pure traction control - how about steer control?

This is an idea I've been rolling around for a while now.

You need a steer position sensor to work out how much steer input there is. You then calculate the relative speeds of all four wheels - and force then to be within a few % of that.

In a situation with minimal traction it should give the vehicle the best chance of executing the line of a turn as demanded. I think it should work particularly well when traction is very low or variable - the times when things tend to go wrong!

A bit like traction control - but for steering. Something you can't do with traditional difflocks.

Si

JCB use a version of this on some of there smaller hydrodrive telehandlers... Seems to work well until a sensor breaks!

[calle-fas]

That is a really brilliant idea Si!

[simonr]

That is a really brilliant idea Si!

Cheers! One day I'll get around to trying it in anger!

Si

[Gazzar]

Cheers! One day I'll get around to trying it in anger!

Si

Here is another idea - ABS brakes for a towed trailer.

I've always had a fear of a trailer jack-knifing when towing, and knowing that the sankey trailers are set up with a master cylinder and all, it seems that all you need is an ABS block, an ECU and a decent 12v source.

G.

[elbekko]

a decent 12v source.

You do know you're talking about a trailer towed behind a Land Rover right?

[Gazzar]

You do know you're talking about a trailer towed behind a Land Rover right?

Fair point, well made. Sorry. Leaving now.

[simonr]

Further to my last idea, if you were to add a flux-gate compass / GPS to the setup, the traction control would know what the car is actually doing in terms of rate of turn. It could then modulate the difference in speed between the wheels to keep the car as close to being on the expected course as possible. So long as it were suitably (critically) damped, it could have a big impact on stability while you are sliding.

Imagine something like a hill climb with limited / variable traction and the front wheels getting a bit light - being able to stay on the steered course has big safety implications. You might have to learn not to react to skids with the steering wheel though - just point in the direction you want to go.

Si

[Gazzar]

Further to my last idea, if you were to add a flux-gate compass / GPS to the setup, the traction control would know what the car is actually doing in terms of rate of turn. It could then modulate the difference in speed between the wheels to keep the car as close to being on the expected course as possible. So long as it were suitably (critically) damped, it could have a big impact on stability while you are sliding.

Imagine something like a hill climb with limited / variable traction and the front wheels getting a bit light - being able to stay on the steered course has big safety implications. You might have to learn not to react to skids with the steering wheel though - just point in the direction you want to go.

Si

Replace the steering wheel with a joystick yoke? Removes instinct associated with the steering wheel.

G.

[Bowie69]

Or, plug it all into an Android powered PDA and use the g-meter, compass and GPS built into that, attach it to a steering wheel which is just there to look pretty, and detatch it from the front axle, which of course would be full hyro steering....

[simonr]

full hyro steering....

Shhhh....you can't say that on here! The SVA Police are listening to E V E R Y T H I N G!

Si

[Deej]

Further to my last idea, if you were to add a flux-gate compass / GPS to the setup, the traction control would know what the car is actually doing in terms of rate of turn. It could then modulate the difference in speed between the wheels to keep the car as close to being on the expected course as possible. So long as it were suitably (critically) damped, it could have a big impact on stability while you are sliding.

Imagine something like a hill climb with limited / variable traction and the front wheels getting a bit light - being able to stay on the steered course has big safety implications. You might have to learn not to react to skids with the steering wheel though - just point in the direction you want to go.

Si

Very cool idea, one problem I can see though is that the refresh rate of standard GPS systems is 1Hz (i.e. once per second). This would cause the feedback to be far too slow, a bit like having an ABS sensor that spat out a single pulse per wheel revolution.

[Bowie69]

It's a tractor, guv, honest!

[simonr]

Very cool idea, one problem I can see though is that the refresh rate of standard GPS systems is 1Hz (i.e. once per second).

I agree - if you use a flux-gate compass for the direction, it can be read continuously. The GPS is then just used for velocity and position which won't change much in a second that interpolation wont take care of.

In fact, just the compass should be enough. If the vehicle is drifting sideways, chances are the driver will adjust the demand course to compensate - so knowing the actual course is not that important.

Si

[FridgeFreezer]

Some very groovy ideas here, not sure I want to add ABS sensors to the portals (or as I'm sure someone would suggest, read the speed from the portal gears themselves) but some sort of forced steering seems most obvious.

I do like the intelligent trailer brakes idea though, could be a real winner for some applications.

[simonr]

Some very groovy ideas here, not sure I want to add ABS sensors to the portals (or as I'm sure someone would suggest, read the speed from the portal gears themselves) but some sort of forced steering seems most obvious.

An inductive pickup drilled through the casing to pick up the gear teeth is the easiest. As the half-shafts are geared down to the wheel, you could probably just read the splines and still have enough resolution.

I wonder if you could actually do it without wheel sensors at all?

If you know the vehicle speed and steer angle, you can work out the rpm of the fastest wheel. You don't apply any braking to that. You know the proportional speeds of the other three wheels so you can apply a proportional amount of braking (using the on/off duty cycle) to the other wheels. It would not work as well - but would probably still work.

Pushing the envelope a bit further - by putting a pickup on either end of each half-shaft, you could measure the amount of twist in the shaft and hence the torque applied to each wheel.

You could use this for some kind of torque balancing or use it to detect when a shaft is about to break and reduce the braking on that axle / kill the engine power briefly as a sort of governor.

Si