System for levelling a motorhome

[Robert Baer]

A friend of mine has bought a 21-seater coach with the idea of
converting this into a motor home. Although it is a bit over the top
(OTT) in my opinion, he wants a system that can

a) Stabilise the vehicle when parked so that wind does not cause it to
move on the suspension. i.e remove the effect of the springs in the
suspension.

b) Level the vehicle if parked on terrain with an incline.

He has asked me if I can design him a system to do this. He is quite
capable of adding hydraulic rams to do the mechanicle work, as he has
excellent mechanical skills and has built a kit-car and done other
related items. However, he has asked me to design a control system to
do this. He wants to push one button and it all work. I'm an
electronics engineer, with quite a bit of experience of using A/D's,
D/A's and microcontrollers.

He intends adding two sensors that can check if the vehicle is level
in two orthogonal (at right angle) directions.

I'd like some advice/guidance from anyone with knowledge of how such
systems might work - I belive they are available at a high price.
Someone has pointed out to me the turret on militrary tanks keeps in
one direction, despite the fact the tank moves on uneven terrain.

My thoughts to date have been

1) Determine from the angle sensors which of the 4 corners the coach
needs to be raised. Lower a ram in the corner most needing raising
until such time as it in contact with the ground with say a 1 degree
correction in the angle of the coach.

2) Raise the corner that requires the second most degree of correction
until there has been a 1 degree change in the coaches angle.

3) Do likewise for the second and third corners.

At this point each ram should be touching the road surface.

Then repeat the procedure, after which the coach should be level.

Another idea is the dead opposite of this.

a) Raise the corner of the motorhome that is already highest, so
making the motorhome in a worst position than originally. Then do this
for the second worst corner. and to this for each corner. This seems
more robust as an algorithm, but needs longer rams, which cost more.

I have nagging doubts about all these.

Do I need pressure sensors to be certain the rams are indeed touching
the surface on which this is parked ? What happens if the surface is
not very hard, and gives a certain amount? What effect would wind have
on this ?

I don't want to completly lift the motorhome, but likewise I do need
to be able to remove some of the give in the suspension.

Dave kirkby

Perhaps an atom bomb could be used to solve all of the problems in a
definitive manner.

 

[Mac]

A friend of mine has bought a 21-seater coach with the idea of
converting this into a motor home. Although it is a bit over the top
(OTT) in my opinion, he wants a system that can

Sounds way OTT to me. I'm sure I wouldn't do it this way. If I did this at
all, I would put in seperate manual controls for all four corners and do
the whole stabilization thing myself, using small bubble-levels for
sensors.

a) Stabilise the vehicle when parked so that wind does not cause it to
move on the suspension. i.e remove the effect of the springs in the
suspension.

b) Level the vehicle if parked on terrain with an incline.

He has asked me if I can design him a system to do this. He is quite
capable of adding hydraulic rams to do the mechanicle work, as he has
excellent mechanical skills and has built a kit-car and done other
related items. However, he has asked me to design a control system to do
this. He wants to push one button and it all work. I'm an electronics
engineer, with quite a bit of experience of using A/D's, D/A's and
microcontrollers.

He intends adding two sensors that can check if the vehicle is level in
two orthogonal (at right angle) directions.

There are micro-macnined accelerometers out there, nowadays. I seem to
remember Crossbow as a manufacturer, in the US. (Not sure where you are.)
I think Analog Devices has them, too. This should be able to provide level
sensing feedback.

I'd like some advice/guidance from anyone with knowledge of how such
systems might work - I belive they are available at a high price.
Someone has pointed out to me the turret on militrary tanks keeps in one
direction, despite the fact the tank moves on uneven terrain.

Yeah, but they would have to use three-axis gyros (or similar
micro-machined devices) and possibly three-axis accelerometers, too, all
with fast update rates. This is overkill for your application.

My thoughts to date have been

1) Determine from the angle sensors which of the 4 corners the coach
needs to be raised. Lower a ram in the corner most needing raising until
such time as it in contact with the ground with say a 1 degree
correction in the angle of the coach.

You don't want angle sensors, you want a two-axis accelerometer installed
perfectly level, or at an offset to perfectly compensate for its own bias.

2) Raise the corner that requires the second most degree of correction
until there has been a 1 degree change in the coaches angle.

3) Do likewise for the second and third corners.

At this point each ram should be touching the road surface.

Then repeat the procedure, after which the coach should be level.

Another idea is the dead opposite of this.

a) Raise the corner of the motorhome that is already highest, so making
the motorhome in a worst position than originally. Then do this for the
second worst corner. and to this for each corner. This seems more robust
as an algorithm, but needs longer rams, which cost more.

I think this is the better algorithm. It could probably allow you to
dispense with the pressure sensor. Then again, you might get thrown off if
you get a gust of wind while you are lowering one of the rams. This might
trick the system into thinking the ram has touched down when it hasn't.
Also, I should mention that I have no experience with this type of
application.

I have nagging doubts about all these.

Do I need pressure sensors to be certain the rams are indeed touching
the surface on which this is parked ? What happens if the surface is not
very hard, and gives a certain amount? What effect would wind have on
this ?

As I mentioned above, you may indeed need pressure sensors. They would
seem to provide the most reliable detection of when a ram touches down.
However, since any sensor can fail, I would probably put some logic in
there to make sure that if the accelerometer(s) detect(s) a large change in
tilt, you turn off the ram and signal a fault, even if the pressure sensor
is reading a low pressure.

I don't want to completly lift the motorhome, but likewise I do need to
be able to remove some of the give in the suspension.

It might be nice to be able to completely lift the motorhome. I wouldn't
want to jack one up manually to change a tire.

Dave kirkby

Good luck!

--Mac

 

[Paul Hovnanian P.E.]

A friend of mine has bought a 21-seater coach with the idea of
converting this into a motor home. Although it is a bit over the top
(OTT) in my opinion, he wants a system that can

a) Stabilise the vehicle when parked so that wind does not cause it to
move on the suspension. i.e remove the effect of the springs in the
suspension.

Any jacking system will accomplish this. Once the bus is supported by a
set of stiff jacks, movement will be minimized.

b) Level the vehicle if parked on terrain with an incline.

He has asked me if I can design him a system to do this. He is quite
capable of adding hydraulic rams to do the mechanicle work, as he has
excellent mechanical skills and has built a kit-car and done other
related items. However, he has asked me to design a control system to
do this. He wants to push one button and it all work. I'm an
electronics engineer, with quite a bit of experience of using A/D's,
D/A's and microcontrollers.

You'll need pressure (or load) sensors on each leg and two nested
control loop levels. The inner loop will regulate the pressure in each
ram to maintain an equal load distribution. Two outer control loops (one
for the roll axis and one for the pitch axis) will provide error signals
from the tilt sensors to bias the load control loops and achieve the
leveling function.

The algorithm should guarantee that each leg will pick up a minimum
amount of load in the event the bus is initially parked level so that
the equal load criteria isn't satisfied until some minimum amount of
load is taken off the bus springs.

The system will also need some sort of limit switches so it can detect a
situation in which there is insufficient ram travel available to satisfy
the leveling solution.

If the rams have sufficient travel, you might also want to implement a
angular limit safety switch (a pendulum hanging inside a contact ring of
the appropriate size) so that when the system tips the bus close to its
limits of stability, the whole thing is shut down. Needless to say, this
feature shouldn't depend upon the microprocessor to work. The tilt
sensors can probably judge whether the incline is beyond hope of
leveling before entering the leveling algorithm, but the safety system
should shut everything (including the pump) down in the event the rest
of the system goes brain-dead.

 

[Fred Bloggs]

Any jacking system will accomplish this. Once the bus is supported by a
set of stiff jacks, movement will be minimized.




You'll need pressure (or load) sensors on each leg and two nested
control loop levels. The inner loop will regulate the pressure in each
ram to maintain an equal load distribution. Two outer control loops (one
for the roll axis and one for the pitch axis) will provide error signals
from the tilt sensors to bias the load control loops and achieve the
leveling function.

The algorithm should guarantee that each leg will pick up a minimum
amount of load in the event the bus is initially parked level so that
the equal load criteria isn't satisfied until some minimum amount of
load is taken off the bus springs.

The system will also need some sort of limit switches so it can detect a
situation in which there is insufficient ram travel available to satisfy
the leveling solution.

If the rams have sufficient travel, you might also want to implement a
angular limit safety switch (a pendulum hanging inside a contact ring of
the appropriate size) so that when the system tips the bus close to its
limits of stability, the whole thing is shut down. Needless to say, this
feature shouldn't depend upon the microprocessor to work. The tilt
sensors can probably judge whether the incline is beyond hope of
leveling before entering the leveling algorithm, but the safety system
should shut everything (including the pump) down in the event the rest
of the system goes brain-dead.

Good idea- this sounds like a small handful of 741's driving the whole
error bridge thing to zero.

 

[N. Thornton]

Good idea- this sounds like a small handful of 741's driving the whole
error bridge thing to zero.

Not a good idea. This could activate while driving, or become faulty
and partially activate while driving. If the person building it doesnt
understand about error trees and the dangers and reliability issues in
auto environments they shouldnt be building something like this.

In this country it would also make your vehicle uninsurable.

Regards, NT

 

[Alan Balmer]

One of the lads here in Oz has come up with a system, no electronics as I
understand. AFAIK the rams are all fed from a common reservoir and thus
each moves until it hits the deck and I guess a coupla pounds are then added
to take the load.

Sounds good for stabilizing, but it wouldn't level.