Aid to getting it perfectly level

Caravan Levelling Aid

This is a really simple project to understand, but the results are quite astounding.

This design is my copyright, but you are most welcome to reproduce it for your own use. If you think you could make it a worth while and saleable item, I would be more than happy to discus it and there would be absolutely no up front fees payable to me.  I have had it in regular use for the past four years, so it is a tried and tested design.

I am well aware that some might find it unnecessary, that its high accuracy is even less necessary, but its main point is that of speed and convenience of setting up level, without need for an assistant. The display unit is intended to hang in your front window, just at those times you require to use it and at other times can be stowed out of the way. You just drive (or use the mover) onto site, then watching the display via your interior mirror, pull the van onto your ramp until it shows side to side level. After unhitching, you wind the jockey until it shows end to end as being level. No fuss, no confusion, no SWMBO misinterpreting a bubble level, swift, simple to understand and absolutely accurate every time.  It is visible in full bright sun as well as in total darkness.  

As said the display goes in a window, the sensor unit can be installed anywhere you like, where a 12v supply can be found, it only draws a few mA, which can be tapped off any circuit in the van.  It does of course need to be where it will not be disturbed, or knocked once calibrated A good location will likely be under a front bed, or in a top front cupboard, but the sensor must be firmly fixed. The sensor is then connected to the display via a multi-core flexible cable, the length of which is not important what ever length suits the application.

In use it is rather counter intuitive. As you would expect - One of the pair of opposing LED’s will light up, to indicate the high side, but then as the van approaches very close to level, both opposing LED’s will flash alternately. Perfect level is when the two LED’s flash on for an equal period of time – you are not aiming for both opposing LED’s to be out, as that is almost impossible to achieve in practise.

Calibrating

Calibrating the sensor is a matter of setting your van to its best compromise of level, then calibrating the sensor to that level. We all know that the floor will not agree with the worktop and the worktop not agree with the sink plus hob and these not match the needs of the drains - so you set the van to what ever is the best compromise, then calibrate this unit to that best compromise of level. Each time you arrive on a site after that, you can then be sure of reproducing that exact level.       

Construction

No doubt you will have to do some interpretation depending upon what you have available and your resources.

This is my second version, with an amendment at the end, to improve the base.

For this one I used an old cassette tape case, as the case for the display. It is permanently powered, ready to use, so a switch is included in the side of the display unit. Build method is my usual quick and dirty for a one off design – just solder the components and wires in place. Case contains 4x superbright white LED’s for the four directions, 1 superbright red LED as both a power on indicator and a reference point for which other LED’s are lit when working in absolute darkness setting up, a switch for turning it on and off and a diode to protect against the power being connected the wrong way around.
__________________

Wiring Diagram and components.










Intention is to  hang the display in the front window or just place it so it can be seen only whilst setting up. The damp patches are where I superglued the LED’s in place.





My Mk II sensor uses a length of 20mm plastic electrical conduit, with a coupling with a threaded lug end and a threaded locking ring. The latter secures it to a sheet of double sided PCB, that PCB is then fixed to a second sheet of PCB, with 4x4mm brass screws and three 4mm brass nuts per screw (12 nuts in total), so the spacing between the boards can be micro adjusted at each side for calibration purposes. Total height of the 20mm conduit from face of PCB to the top is 6.5”. Once the screws were tightened, I ground their heads down to reduce the amount they protruded.



Linking the display to the sensor unit, I used 4 pair (8 wires) flexible type network cable, but only 6 wires are needed. The network cable is fixed to the lower PCB only, with a couple of tiewraps, holding it clear of the underside of the end of the conduit and its locking ring. A 5amp two way connector is fixed to the lower left corner for connection of the 12v +ve and -ve, with a third tiewrap




The upper PCB has its corner cut off, to enable access to the 12v input terminals. Here are both units ready for final connection.




This is the brass bob-weight, turned up in the lathe. It needs to be slim enough to fit in the plastic tube and between the four contacts The RH end was drilled and tapped to accept a 4mm brass screw. The brass screw I drilled through the centre with a 1mm drill, to accept a suspension wire. I made a second similar screw, which screws into a tapped brass insert which was itself turned to fit neatly inside a 20mm plastic bush (the black object on the right). You can just see the brass insert poking out of the plastic bush.

The suspension of the bob-weight is crucial to its accurate operation. It has to be centred, both top and bottom and the suspension wire needs to be conductive, thin and very compliant, so as not to disturb the ‘hang’ of the bob-weight. Network cable is too rigid and will want resume its curl. I used a single core from the telephone wire, as used to link a phone to the socket at the wall. This type of wire has the copper strands mixed in with nylon, which makes it very strong, but very flexible. The downside is that it is very difficult to solder, but not impossible. When soldered it will have the two screw heads facing each other, ready to screw into the bob-weight at one end and the brass bush at the top end. That is stripped at either end, then passed through the 1mm holes in each screw head, then the ends soldered at the far end of the screw from the end (the threaded end). I set it up so the length of the suspension wire, hung the bob-weight with 2/3rds below the contacts, 1/3 above the contacts.




For the four point contacts, making contact with the bob-weight, I used four brass electrical grub screws, removed from an old socket.  Ideally these need to have an end which is slightly pointed or at least slightly domed, so they make a definite contact at a single point on the bob-weight. I drilled a suitable size of the 4 holes, so the grub screws cut their own thread, but which left the grub screws tight to turn in the plastic. You don’t want the screws able to move on their own, once in service, but they do need to be adjustable for setting up.

There is no need to prevent the bob-weight from moving whilst in transit. It cannot move very far between the contacts, the suspension wire can support several kilogram and the bob-weight only weighs a few grams. Besides the slight movement of the bob-weight helps keep the contacts and itself clean. A spray of WD40 does seem to help with this last point.

Here is my latest update on this little project, with some improvements...

I have revised the base – it is now a tripod with three adjusters to set the level up. The two triangular bases are made from some scrap alloy sheet. The base of which has to be fixed into a suitable place in the van, the van then set at what ever level works for that particular van, then the unit is then finally calibrated. The idea is not one of getting the caravan precisely level, but one of being able to quickly reproduce the version of level which makes your caravan work best.



I've used crimp on bullet connectors, which happened to be a perfect fit on the grub screws. Rather than adjusting a top and bottom nut, I have used three springs pushing the two triangular plates apart and tight against the top nut. The vacant holes just above four contacts points, serve no purpose, it was just an attempt to use an alternative contact method, to the one I settled upon. 




Here you can see one of the four grub screw contacts. These are tightened up so as to leave an absolute minimum of gap between their tips and the bob-weight. The bob-weight obviously needs to be able to move, but only slightly.



This most recent version is as yet untested for accuracy, but a couple of weeks ago I carried out some accuracy and repeatability tests on my original version and I have no reason to expect the accuracy of this version being much different from my original. The results were really quite surprising....

For test and calibration purposes, I used a mag-mount LCD engineering level, which reads to 00.00 of a degree - 2 decimal places. I only tested the fore and aft accuracy, the side to side will be the same, but not so easy to test. The process went as follows...

1. Switch unit on and level the van end to end as near as I could get it, as indicated by the unit. In fact the indication was slightly more accurate than I was able to adjust it, using the jockey wheel height adjustment. It is so very sensitive.

2. Set my mag-mount unit on the hitch, switch on and calibrate it to zero.

3. Lower the nose of the van to the floor, raise it to maximum, then adjust back to level as indicated by my HR unit. Finally take a reading from the mag-mount unit.

I repeated step 3 only a total of four times and noted down the errors between HR Unit and my fancy mag-mount unit, these are the results....

Round 1 00.09 degree error
Round 2 00.13 degree error
Round 3 00.01 degree error
Round 4 00.08 degree error


I had originally expected it to work to an accuracy of 1 degree or so, but the above results are quite astoundingly accurate for such an apparently crude device.

Enjoy!