I have just finished rebuilding all the front suspension on my UK spec 1985 HE. This was the last thing to do after I completed my entire rebuild last year, and got the car road going and road registered.

New everything from stub axles upwards. So although I put the new track rod ends in the same position as the old ones on the rack rods, I knew the toe might be off. This is what I did and it seems to have worked:

I was aiming for 0.5mm toe out at each wheel measured as the difference between the hub centre and the wheel rim (ie front of wheel 1mm toe out compared with rear of wheel). This corresponds to an angle of 0.1432° on a triangle with two 8 inch (200mm) sides and one 0.5mm side (see website link below for calcs).

Centred rack and put tape at 12 o clock on steering wheel. Parked the car about 3 yards from the barn door. Took my laser level (10 UK pounds at any hardware shop) and placed it across the wheel centre. I had to tape it to a bit of square ally tubing cut to the right length so it just fitted exactly across the wheel rim edges. Shone the laser on the barn door (from about 3 yards). Marked door. Repeated other side. Measured exact distance from wheel centre to the door. Measured exact distance between marks on door. These were measurments 1

Next, drive car as close to door as possible (about 1 yard from wheel centre to door). Repeat measurements. These were measurements 2

These were the results:
Distance 1, wheel centre to door: 3780mm
Width 1, between marks: 1734 mm

Distance 2 wheel centre to door: 1060mm
Width 2 between marks: 1772mm

Therefore the car had some toe in, as the more distant marks were closer together than the nearer ones. How much toe in?

Here I went to this useful trigonometry site: TRIANGLE CALCULATOR Solves 4 Cases of Triangles - SSS - SAS - ASA - AAA
You can input any combination of triangle side lengths and angles and, if logically possible, get all the other dimensions and angles calculated exactly. So, the length difference between the two sets of marks on the door was 1772 - 1734 = 38mm = 19mm each wheel. A triangle with two sides of 3780 mm and one of 19mm has an angle of 0.28799°. So the wheels were currently 0.28799° toe in.

They need to be 0.1432° toe out, (which is 0.5mm each side at wheel rim) so they needed to be adjusted by a total of 0.42° (effectively minus 0.28° to plus 0.14°).

So I guessed 1 and 1/4 turns each side on the track rod. Redid the measurements and found that I had gone to about 0.9mm toe out each side (ie an angle of about 0.25° toe out was calculated). Wound track rods back 1/3 of a turn and they were bang on 0.5mm toe out each wheel. As far as I can reason it, this is a good method for the home mechanic, as it does not rely on the car being in an exact position, or even the same position for each measurment, or sets of measurments. And it is quick to do and, because the distance of 3 yards or so is quite large, any measurement error of a mm or two, would not greatly affect the results.

Greg

 

Excellent work Greg, congrats!
I have thought that I might try adjusting my wheel alignment after reading some of the different explanations out there including yours...and I do have the laser level. After reading your explanation, I arrived at the following conclusion.
High School Math Grade C = Pay someone else to do the work...besides,
I don't have a barn door.
I am fascinated that you can explain it in such a way that I would never be able to understand it!
My theory and example of getting close to the proper alignment would be to take two 16' 2X4's and lay one on each side of the car, against the wheels. Then take a tape measure and measure the spaces...adjust the wheels accordingly. A flat surface works best, no barn door needed.

Works in spite of...not because of.

Keep up the good work, I have to rebuild my front end also and I have collected the parts. I am still in the procrastination stage.

 

Just wondering why you decided on toe out? Is this a track car?

I am rebuilding my front end now and planning on doing my own alignment, I do not have a flat driveway it slopes away from the garage so I am planning on using the race car method.

4 concrete pavers some vinyl floor tiles (these are to level the car and act as turn plates with grease between the top 2) 4 jack stands 2 bits of electrical conduit 4 clamps and 2 bits of sting.

Level the car in the garage on the vinyl tiles (the concrete pavers are only to get some height so you can get under the car to adjust things). 2 jack stands at the front and rear of the car with the conduit – set the jack stands with the conduit so when the string is tied between the front and rear it passes through the centre of the hub on front and rear wheels. Measure the distance between the jack stands left to right so they are equal and then clamp the conduit to the stands.

You should now have a large rectangle around the car across the front and back have conduit and down the sides string. Centre this side to side by measuring from each hub to the string.

Toe can now be adjusted relative to the rear wheels.

 

A friend has just asked me that same question. The answer is 'that's what it says in the Haynes manual HE special chapter page 328: 0.0 to 1.6mm toe out'. BUT, because of your question, I looked at my copy of the Jaguar published HE supplement to the Jaguar repair operations manual and it says 0.0 to 1.6mm toe IN!

So which is correct? I am guessing the jaguar publication. having said that, the car has always been set to a touch of toe out! What are the theoretical differences in drving characteristics, please, between in and out?

Greg

 

I have read a lot about alignment since I am considering doing my own also. Here is the best description of proper to settings that I have found.
Caster, Camber, Toe

 

It should be toe-IN.

Toe-out will cause the car to be more sensitive to steering input and road irregularity. Twitchy. Good for a race car, not so good for a street car.

 

This is exactly why I asked the question. Toe out will make the car turn in but will make it lessstable in a straight line. Also toe will change dynamically so toe out will increase as you turn the wheel.

You could try zero toe, I used this setting on my AMG and it made the car turn much better without effecting straight line stability noticeably.

 

Always love home made solutions. What a great thread.

Not exactly wheel alignment, but built my last LR chassis on exactly the same basis. Couple of straight edges, spirit level, T square and reference measurements, and chalk. (comp bloke came and bought it after 1 outing)

But I did fab some quick length templates as a jig, perhaps you could easily do something along those lines for future use.

Still waiting for the full build thread, no pressure, as it were.

 

That's a great improvement on the usual direct measurement. The magnification greatly improves accuracy. An alternative is to use a movable screen as the target which avoids moving the car itself and allows you to use distances which are even units for convenience.

The site is a great find because most tables and calculators are oriented towards one known measurement or another instead of the "solve against any given known" that the site takes.

As for toe-in versus toe-out, most production cars have toe-in for the purpose of tracking stability and to allow for eventual suspension wear. As posted above, toe-out is usually only found in track cars that need quick turn-in. Probably things like autocross and slalom as opposed to Nascar.

 

Would the guys be interested in an account of what I did? If so, I'll get to it.

Thanks for the encouragement re: alignment. The 'moving the board' idea is great, but I did not have anything suitable, but I intend to buy one, as it would make it much easier and quicker to check the toe. I also revisited my geometry calcs and did them more precisely from a logical point of view. In case anyone is interested, this is the revised version of my original idea. Same measurements but less approximation. Apologies if this is too long and boring! In sum on an xjs with metric track rod ends, by calculation from measurements (see 18 below) 1 turn of track rod changes a 200mm radius wheel toe by 2.1mm or 0.61°

1. Centre rack and put tape at 12 o clock on steering wheel, so rack can be re-centred by eye.
2. Park car about 3 yards from the barn door.
3. Take laser level and place across the wheel centre. If needed tape it to a bit of square ally tubing cut to the right length so it fits exactly across the wheel rim edges.
4. Shine the laser on the barn door. Mark door. Repeat other side.
5. Measure exact distance from wheel centre to the door. Measure exact distance between marks on door. These are measurements 1
6. Drive car as close to door as possible (about 1 yard from wheel centre to door). Repeat measurements. These are measurements 2
7. Results: Measurement 1:
a. Distance 1, wheel centre to door: 3780mm
Width 1, between marks: 1734 mm
8. Results: Measurement 2:
a. Distance 2 wheel centre to door: 1060mm
Width 2 between marks: 1772mm
9. Therefore the car had toe in, as the more distant marks were closer together than the nearer ones. How much toe in?
10. To calculate the toe in, first draw out the rhomboid shape given by the two width measurements and by the difference between the two measurements of distance: This gives a rhombus with a longest side of 1772 and the next longest side of 1734, with a distance between the two long sides of: 3780*-1060 = 2720mm. A right angle triangle can be drawn at the end of each long side of this rhombus with the following dimensions:
a. Side 1 = 1772 – 1734 = 38 / 2 = 19
b. Included angle = 90°
c. Side 2 = 2720
11. Next go to this useful trigonometry site: TRIANGLE CALCULATOR Solves 4 Cases of Triangles - SSS - SAS - ASA - AAA. You can input any combination of triangle side lengths and angles and, if logically possible, get all the other dimensions and angles calculated exactly. A right angle triangle of dimensions as shown in 10 above has an angle of 0.4° of toe in with the hypotenuse calculated at 2720.1mm.
12. The angle of toe in calculated at 0.4° needs to be changed to an angle of 0.1432° toe out (which is 0.5mm each side at wheel rim compared with wheel centre) so they needed to be adjusted by a total of minus 0.4° to plus 0.14° = 0.54°.
13. So, not knowing how much one turn of the track rod would change the angle of toe, I adjusted the toe by 1.25 turns of the track rod each side.
14. The measurement process described above in1 and 2 above in 7 and 8 were repeated with the following results:
a. measurement 1: Distance 1, wheel centre to door: 3745mm
measurement 1: Width 1, between marks: 1833mm
b. measurement 2 istance 2 wheel centre to door: 1100mm
measurement 2: Width 2 between marks: 1800mm
15. Because the width is 33mm greater at the more distant measurement, there is toe out, how much?
16. To calculate the revised toe, repeat steps made in 10 above. That gives a rhombus of dimensions: longest side of 1833, the next longest side of 1800, with a distance between the two long sides of 2645mm. A right angle triangle can be drawn at the end of each long side of this rhombus with the following dimensions:
a. Side 1 = 1833-1800 = 33 / 2 = 16.5
b. Included angle = 90°
c. Side 2 = 2645
17. A right angle triangle with the dimensions 2645mm, 90° and 16.5mm has an angle of toe out of 0.36° with the hypotenuse being 2645.1mm. 0.36° corresponds to a toe out in mm (at 200mm rim radius) of 1.25mm
18. Therefore the 1.25 turns of the track rod had changed the toe by a total of 0.4° toe in to 0.36° toe out, ie a total change of 0.76°. Therefore 1 turn of the track rod gives a change of toe of 76/1.25x1=0.61°. This corresponds to a change of toe on a 200mm radius rim of 2.1mm per turn of the track rod.
19. Therefore 0.33 of a turn of the track rod gives a change of 0.2° of toe, or 0.7mm.
20. Step 17 showed that the angle of toe out was 0.36° and 0.14 is required. A one third turn was given to the track rod end thus the toe was set at: 0.36 – 0.2 = 0.16° toe out each side. In mm this corresponds to 1.25mm – 0.7mm = 0.55mm toe out at the wheel rim. This is quite close enough to the desired 0.5mm toe out!


Greg

 

Your rhombus could also be viewed as a truncated isosceles triangle

On a more practical note, how do you verify the parallelism of the laser to the wheel plane?

Do you just accept that the level is within specification?

One possibility is to adapt the calibration check used for spirit levels, which is to rotate the level about the center point on the same level calibration surface. A perfect level would centre the bubble in both orientations on both edges. That would be four possible positions where the bubble has to centre.

In this case, one would measure the horizontal displacement of the target points when the two different edges are against the spacer bar. This would be your parallelism error in the particular level you are using. In a perfect level, the target points would be identical without regard to which opposing edge was touching the spacer bar. The error is probably small on a good unit, but it is worth checking in case the particular unit in use has been knocked out of alignment. However, one can also make a compensating calculation if the magnitude of the error is known.

Finally, as you are measuring distance, it is not necessary to measure from the wheel center as long as you can measure the difference in position. A plumb bob hung from the center of the front bumper would do just as well and perhaps be less fiddly. Just measure between the two points that are indicated when far and near.

BTW, the movable barn door can also be just two pieces of cardboard fastened to two cinder blocks that you can move as required.

 

All excellent points for which many thanks. The laser pointer's accuracy as far as being parallel to its casing, I have accepted for the moment, but I realised it might not be. It claims to be, with the casing which is about 1 foot long, like a small spirit level, having a machined bottom edge to help ensure this.

FWIW, as (i) the measurements indicated a change from toe in to toe out with a very small adjustment of the track rods; and as
(ii) I started from the original professionally tracked point on the track rod threads albeit with (a) new track rod ends that were exactly the same length as the old ones, and (b) all new bushes and ball joints in the suspension;
I took comfort that the pointer must be reasonably parallell in its casing, as any gross error would be unlikely to have hit straight upon the in/out point.

Other than using a spirit level, which I am not sure I could do very accurately, I am a bit stumped over how to prove the parallelism of point to case bottom. Are there any other ways, other than resetting the toe to zero and checking fore and aft width readings are identical at 3 or 4 different distances?

Greg

 

You could validate the laser against a known straight line such as a string. Place the laser on a mount that is parallel to the string and measure the difference at the end point.

 

It is minus 12 here, so the brain is a bit slower than usual, plus it is too cold to mess about in the barn. But am I right in thinking that if the laser is parallel, if I did the procedure with the laser facing backwards, then I should be able to calculate a "heel in" of the same magnitude as the toe out?

Greg

 

The information of interest is the deviation of the laser from the line formed by the edge of the level itself. It is likely that there is some deviation. What you want to know is how much, and can it be ignored as immaterial.

First, if the spacer bar is commercially made square tubing, it is likely within quite tight tolerances. So, that can be ignored. Although it would be best to always use the same edge against the wheel and the same end towards the front. That makes your systemic error consistent. And, as long as the error is consistent, your calculations can compensate for it. Masking tape is your friend here

Now, to determine the deviation. Mark one non-bearing surface of the level with masking tape as a reference mark. Use the same setup as originally described. Mark the resulting laser point against the target surface. Rotate the level along the longitudinal axis such that the other machined edge is against the spacer bar. Align the vertical height of the laser point with the previous mark height. Mark the resulting laser point against the target surface.

The resulting horizontal difference between the marks is the amount of deviation at that projected distance.

You can choose to either use the distance to as a direct adjustment against future distance measurements. Or, you can calculate the angular compensation to apply against future angular calculations.

nb. The target surface can be a horizontally aligned strip of masking tape at the height of the center of the wheel to allow for easy targetting and to keep all the marks at the same vertical height.

 

That was the drift of the isosceles triangle remark

Visualise the plan view of the car with a triangle superimposed. The base runs along the rear bumper in the case of toe-in.

But why do you keep on going after toe-out when most production cars are toe-in? Of course, the XJS may differ, but:

Unless Haynes specifically refers back to the supplement and declares the Jaguar document as containing errata on that topic, then the Jaguar document is likely to be correct and the Haynes document is the subject of bad transcription.

 

I know, old thread but ...

I arrived looking for the effect of each turn of the tie rod.

The specification for parallelism on the brand name consumer grade
laser levels is often in the range of 1/4" at a range of 50 feet.

The adjustment factor can be calculated by aiming the laser level at
a target, marking the spot, flipping the laser level, marking the new
spot, and measure the difference. Any offset of the beam from the
center at the laser level must be accounted for in the calculation.
That is, the laser may be mounted closer to one edge of the laser
than the other.

However, there is an easier way involving no deviation measurement
that is scientifically valid

It involves cancelling the effect of the deviation.

Let's take 25 feet in our example, then ...

- measure 25 feet from the centre of the hub on both sides to the front of
the car on both sides, put a target board there

- repeat 25 feet from the center of the hub on both sides to the rear of the car

- pick one edge of the laser level as the edge always to be adjacent to the
vehicle

- sight the laser from each wheel to the front and rear, always with the laser
exit pupil at the center of the hub

- measure at the front target and the rear target

- perform the calculation on a base of 50 feet (2x25)

The reason this corrects the error is that the errors cancel out
since the error is the same forwards as rearwards for the same
distance.

There are all kinds of "tutorials" out on the interwebs touting
the accuracy of lasers without any reference at all to the
introduction of errors by the laser to level interface.

The problem is the leap of logic from a laser producing a straight
line at the usual distances to a laser producing a straight line
parallel to the plane being measured.

 

BTW, after four years, how is the tire wear with the toe-out?

 

Well, following the earlier posts, I found out that the Haynes manual was incorrect and it should be parallel to toe in 0 to 0.5 degrees! So I readjusted for this using my "turns per degree" calculator. Long story short, I ended up with too much TI on one side, so I went down to the tyre place and got them to set it up on their laser.

My problem was, I think, besides correcting for the laser as you so clearly described, my inability to adjust the tie rod without moving and lifting the car, so I could never be sure to get the car back to the same position as before to re-measure it, plus at the time my floor was not flat! All in all a wonderful example of theory not working in practice. I have not yet had a do-able, affordable idea on how to overcome these problems, other than digging out a pit, which where I live would pretty fast fill up with water!

As for toe out, I had 1/4 of a degree toe out for years before my rebuild, and the car handled fine and tyre wear no different than 1/4 degree to in!
Greg

 

All I did was count the Threads (on the Track Rod Arm) of one of my other XJS's and then adjusted my other Car the same.

Not the way to do it I'm sure but She is driving perfectly straight, with no abnormal signs of any Tyre wear.