Caster and Camber angle moving around

3 years and 5K miles ago I did a complete rebuild of the front suspension including rubber bushings, new sealed lower ball joints, new upper ball joints (in original upper control arms), new wheel bearings, tie rod ends and shocks. 5K miles later and I was starting to feel some play in the steering. Turns out the tie rod ends were shot with visible sloppiness in the joint and there was play in the upper ball joint. I understand from some research that even with new upper ball joints the wear in the socket can create problems but only 5K miles on a tie rod end seems odd.

So I replaced the tie rods with SNG’s “new and improved” and re-shimmed the upper ball joint per the manual setup procedure and it all feels solid again with no obvious unnatural wheel movement. I’m using a DIY alignment kit from which covers camber/caster/toe. The caster angle measurement involves the +/-20 deg sweep of the wheels abd measure the delta in degrees (kit comes with a great wheel clamp and digital level gauge). I’m finding the caster angle is not very repeatable, Over the last week I have made the measurement over 20 times and can get readings on the same wheel from 0.5 to 3.3 degrees. The car is in the mid-laden position using the specified setting links. I have driven the car in and out of the garage multiple times to let the suspension settle but am still stuck with this non repeatability. Sometimes the caster reading will change without even rolling the car forward / back. To a lesser degree the Camber is also fluctuating by about 0.4 degrees.

I have tried wrenching on the wheel and the stub axle nut to induce movement but no obvious play is present. Any thoughts on what would cause this fluctuation? Worn upper ball joint socket?

Meant to include - 1968, OTS

Any play in the suspension system capable of producing those kind of variations should be immediately evident. I’m going to trust you are now well-practiced at identifying ball joint and tie rod play, so the next question would be your set-up.

What do you have the car sitting on, to reduce friction between the tires and the ground? I’m wondering if the tires are effectively ‘sticking’ then breaking free over and over again, and juddering your measurement equipment.

I use a large, heavy-duty trash bag, folded over double a couple of times to give me a near zero friction surface. My caster-camber gauge uses a bubble-level, but I can get very repeatable results.

Thanks for jumping in. I am using some home made turn plates consisting of two 1/4" oiled plastic plates with a center bolt and a series of 5 deg markers. It’s pretty low friction and I can easily turn the wheels. Both front wheels are sitting on the same plates. Some pics are attached, you can see the oil oozing out between the plates.

I am familiar with wheel alignment, but not all the nuances. But are your rims true, and could that be the issue?

As caster is a relative measurement between a +/-20 deg sweep of the wheel I don’t think a bent rim could induce a non repeatable reading.

When I took my car to a shop for alignment after the restoration, the shop owner found that I hadn’t properly adjusted the inner end of one of the tie rods (where it attaches to the rack). The spring loading on the inner “ball” of the rod masked my error when I put it together without doing the proper test described in the manual. Anyway, the slop became apparent when he grabbed the forward and trailing edge of the tire and muscled it back and forth. That doesn’t seem like it would impact camber, but who knows.

Pulled out all shims to get as much negative camber as possible without grinding on the mounting brackets. Moved my upper A arms as far forward as possible just clearing the shocks.
Then set toe in at 3/32.

That set-up is easily as good (likely better) than mine, so I don’t think that’s contributing.

How do your bushings look nowadays? I have about zero faith in repro rubber bushings.

Bushings look ok from the outside. This morning I’ve seen the caster angle range from about 0 to 3 deg and all I’m doing is rotating the wheel +/-20 deg. Car was stationary the whole time. Something odd going on. I’m going to break down the upper ball again and see if there’s anything obvious.

what would happen if you mounted one of your “spare” moderns cars on your setup and tested to see if your measurements were repeatable? This might isolate your testing equipment from car…

1 Like

I don’t see how your castor angle can change unless the threads are stripped on the upper fulcrum and the pinch bolts aren’t secure. That’s unlikely though isn’t it?

Excellent idea! Thank you

I’ve been doing my own alignment on E Types for 20 plus years - started when I raced one, couldn’t afford to take it to a shop every time I did something to the suspension. First thing to be said is that it’s more difficult to do than you think. I’ve been asked to write something on how to do it - I’ve managed three drafts, all unsatisfactory. First chasing repeatability on a stock suspension is chasing a chimera. The suspension has rubber in it, and second it moves up and down, back and forth. Jaguar recommends camber be set at 1/4 positive plus or minus 1/2 of a degree. That’s a range of 1/4 negative to 3/4 degree positive, or 1 degree, and we use equipment that measures it to 2 decimal points? Similarly castor is plus or minus 1 degree. This means that one side at 1 1/2 degree and the other at 2 1/2 degrees is acceptable. As the car moves down the road the suspension moves up and down, drastically (I mean that word in it’s usual sense) changing camber on the wheels. at times in opposite directions, and the ball joints go back and forth on the rubber bushings changing toe and castor. The tires want to change direction because of this. It’s why your car doesn’t always feel stable.
What I’m trying to say is you don’t get repeatability unless you measure under exactly the same conditions, with the same stress on the bushings, and that’s impossible. You drive into your garage, stop and get out of the car, and measure camber. Then push it backwards, or forward, measure again and you’ll have a different reading. Camber lock is a real b…h. Minor changes in ride height will give you different camber.
Alignment shops have turntables that also move in and out with camber. The tech will lock down the car with hooks and chains so each side is at the appropriate ride height. The tables allow the wheels to move out, eliminating camber lock. He/she (it’s 2019) then sets castor and camber. We don’t have that luxury so every amateur suspension setting is subject to error. To get rid of the camber lock you need to drive around the block, not just roll the car backward and forward, then measure.
Andrew I don’t know why you are having this problem with castor. Unless your suspension is really wonky, and I doubt that, I suspect it’s a measuring problem. My castor and camber gauge from Longacre simply measures the difference in camber over a 20 degree swing of the steering. It cancels any errors associated with bushing deflection etc. I watched a video of the system you bought but they didn’t show castor measurement, so I can’t comment.

1 Like

My garage floor slopes slightly down toward the door. I trust that it is neither flat nor level.


Wherever space and time interact, there is information, and wherever information can be ordered into knowledge, and knowledge can be applied, there is intelligence.
Pavel Mirsky, mid 21st Century Russian General

I can appreciate the problems involved for the do-it-yerself guy.

I made a simple toe-in stick and do he measurement after rolling the car forward. That is because the pressure on the tires is in that direction (toward toe-out) when it rolls down the road. It isn’t perfect, but is close enough that the tires wear properly over the long run. So there are some cases where good enough is good enough.


Wherever space and time interact, there is information, and wherever information can be ordered into knowledge, and knowledge can be applied, there is intelligence.
Pavel Mirsky, mid 21st Century Russian General

Well said, and from actual experience. This comment also touches on the concept of “significant digits”, from the world of experimentation. Because of the design of our spindle carriers you can, and I do and did, measure caster directly from the flats machined on the carrier, as they are parallel to the ball joint axis. You only need to make up a simple gizmo to allow a precision level to “follow” the plane of the machined surface. I just don’t understand why all E-Type owners don’t do this. The ‘flats’ seem to be there in order to facilitate machining, and if you consider the likely tool paths, etc., the concept is logical.

Very nice plates Andrew !
But… I think your problem is the center bolt.
When you turn the wheels the turning axis does not meet the ground under the tire but further front, so the wheel must be able to move in a Arc.
With the bolt in place the rubber bushings will flex and your measurements will not be correct.
The same ofcourse applies and for camber measurements.



Andrew, do your plates allow friction free side to side and fore aft movement? (I’m agreeing with Aristides.)

I use soapy plastic sheets. They’re so slippery, I can push the car sideways with a finger.

Slick magazine covers also work pretty well.