Welcome to our resource for MG Car Information.



MG parts spares and accessories are available for MG T Series (TA, MG TB, MG TC, MG TD, MG TF), Magnette, MGA, Twin cam, MGB, MGBGT, MGC, MGC GT, MG Midget, Sprite and other MG models from British car spares company LBCarCo.

MG Midget and Sprite Technical - ARB - stupid question

When recommending an upgrade to the ARB, people seem keen on the 11/16ths. Why not 3/4"? What's in a 1/16th?

Moss supply both, MGOC only seem to do the 3/4".
Stuart K

Beats me. Both my 60 and my 72 are running on 3/4" ARBs and I see no reason to regret it. Perhaps the 11/16" would be a trifle less likely to rip out the bottom of the chassis legs, but I doubt it.
David "anyone make a tubular ARB for the Spridget?" Lieb
David Lieb

Doesn't speedwellengineering have a tubular one?
Trevor Jessie

"keen on the 11/16ths. Why not 3/4"? What's in a 1/16th?"

Stiffness is proportional to the square of the radius or diameter. So, 121/144, or roughly speaking, the 16th is worth about 18% in torsional stiffness, plus some for bending resistance outboard of the support points.

FR Millmore

You are right. I never noticed the word "tubular" in there before.
David "needing a seeing-eye cat" Lieb
David Lieb

FRM is correct here, the calculations in the effectiveness of an ARB are very complex and
"What's in a 1/16th?" really is a significant amount.

When working with racing Alfa GTAs, we used ARBs with only a few thou difference to give different handling.

M T Boldry

The Tubular bars are about an inch diameter and no-one to my knowledge makes a 3/4 or 11/16 tubular bar.
Daniel Thirteen-Twelve

<<Made from 4120 Chromemoly 7/8" tubing, this bar is 7% stiffer than a 3/4" solid bar and weighs 4 pounds less.>>

From the speedwellengineering website - however it gives a price of $274.50, so those 4 pounds are lost with a rather large loss of .

Why does no-one do an ARB to mount on top of the chassis legs? Seen on racers, where the ground clearance issue is critical. Presumably Peter May can make them up, since his race cars use them, but his website makes no apparent mention.


('Seeing-eye cat' - lol David)
Tom Coulthard

I remember seeing a Sprite built with an ARB mounted under the chassis rails, but with the ends turned up above the lower wishbone pan and short drop links pointing upwards. It looked quite neat, but I am not sure that it was any great improvement to ground clearance where it matters.

Guy Weller

midget triallists mount them on top, too. I've never looked closely enough.

There is a definite desirability to having the legs of the ARB as level as possible with the vehicle at rest so that the effective leverage is as constant and consistent as possible. Achieving this tends to be non-trivial in my limited Spridget experience. I really do not like the standard fitment with the link pointing down from the wishbones.

There is a British poet, currently living in Wales, named Les Barker. With a name like that, you would be correct in assuming that many of his poems relate to dogs (like his classic "Dachshunds With Erections Can't Climb Stairs"). He does have a priceless poem, however, entitled "Guide Cats for the Blind" from which I cribbed the phrase. He has a website at
David "credit where credit is due" Lieb
David Lieb

Some say 11/16 for Aseries Midgets and 3/4 for 1500s (assuming fast-road use).

I use 11/16 for road, track-days and competition (for my RWA that was Aseries and now K).

Increasing ARB stiffness increases understeer.

Anthony Cutler

Here is a photo of ARB on top chassis, I have been running this for about 20yrs with no probs. Its supported on two simple brackets that are attached to the two bolts that locate the towing eyes. The bar itself is a shortened A40 Farina one, the midget arms are too short as the brackets are slightly forward of the original. Also the links are cut and rewelded at a different orientation.
Ian Webb '73 GAN5

Cant get the photo uploaded, I can email if anyone interested.
Ian Webb '73 GAN5

Yes please! news56 at hotmail dot com

You may need to shrink the pic to upload it to the bbs.
Mick - trying to sort the wiring, please.
David "the plagiarist" Lieb
David Lieb


I've done similar but used a sprite 5/8" ARB although fitted in A40 farina ARB bushes and outriggers on the chassis legs to improve the effectiveness of the ARB. The lower front edge of the radiator stantion just above the chassis leg was notched to suit. I've used the car for about 10 years like this without problems.

David Billington

David L /Mick you have mail.

My set up is similar to David B's but mine doesnt modify the chassis in anyway, as the mounting brackets are bolted using the towing eye bolts, and my radiator mounts are not notched, as my bar is in front of them (hence the need for a bar with longer arms) Also I was aiming to have the links to the wishbone vertical not leaning forward as in DB;s set up.

It looks a bit easier on a frog!

I wanted it to be completely reversible and removable, the only cutting is one hole in either side of the front shroud panel, the bar has to be threaded through this

Ian Webb '73 GAN5

Putting the supports further outboard as David B has done effectively increases the rate, since more of the bar is in pure torsion.
Longer arms decreases the rate, since the same vertical movement results in less twist.
Tubular bars are less stiff than solid ones at the same OD - but much lighter.
Equivalent tubular bars will be larger OD than solid ones - and still lighter.
Deflection in mountings has a huge effect.
You cannot directly compare bars by size if any other parameter is different: material - alum/steel/titanium, etc, hollow/solid, length of arms, length of arm in bending rather than torsion, support bearings.
The ideal bar is tubular, light metal, supported by real bearings at the ends of the torsion section, has very stiff arms in bending.
Look up standard sprint car suspension - these hold the car up, but the principle is the same.

FR Millmore

Can you please explain how "putting the supports further outboard as David B has done, effectively increases the rate, since more of the bar is in pure torsion."
I fail understand this as I always thought the arms twist (torsion) against each other and the mountings are simply a support bearing. I can see it helps the stability of the bar a little. Also a frequent upgrade of our type of arangment is to fit a third mount in the centre of the bar, so I can't see how that fits in with your first statement. I'm more than happy to accept different ideas..... just cant get my head 'round that one .
All the rest spot on.

I did think about building a tubular bar with blade type arms when I first made my setup, I just didnt know what material to use for the tube,and also how to calculate the specification/rating foe our cars....sounds like something DB could do??
Ian Webb '73 GAN5


Imagine the mounts are close to the centre of the car; some of the twisting motion of the bar is lost as the bar itself (outboard of the mounts) bends under the force of the suspension movement. This reduces the apparant movement of the suspension (from the view of the part of the bar under torsion). This bending is not 'seen' by the bar at the other end (like the torsional bending is seen).

So the bar will appear stiffer if the mounts are placed near the ends of the bar (reduced bar bending).

In the design of the car, the chassis mounts are the convenient place to mount the arb; not the OPTIMUM place. The bar stiffness will have been calculated to take into account the above bending effect.

Anthony Cutler

Are you saying that the portion of the bar outboard of the mounting, bends before the bar itself twists ??

Still trying to get my 'round it !!

Ian Webb '73 GAN5


What Anthony said is correct. The standard mounts, being well inboard of the arms, allow the main torsion section of the ARB to bend into a shallow S shape so losing some of the benefit of the bar.
David Billington

Yes; sort of!

It doesn't bend BEFORE the bar moves. Rather, each part of the system will bend/twist according to its stiffness.

The strain is shared by all the components in the system, incl the bushes, according to their strength. All components deform together; but the deformation of each component in line in the system reduces the amount of movement and therefore force that can be applied to the next element in the system.

So any bending of the bar is motion lost for the twisting of the shaft; and only the twist is conveyed to the other end.

(And of course the other end of the shaft will bend too, which looses vertical motion applied to the suspension on the other side.)

Maybe I'm getting too hung up on 'before' - and your description/understanding is good.

Anthony Cutler

yep - that's better!

Anthony Cutler

Anthony has it.
Consider a bar bent in a V shape and supported at the apex. When one end of the bar is moved with the other stationary, both sides of the V bend, and there is no useful torsion "section" whatever, just a tiny bit at the apex.
If you make the bar U shaped, there will be some torsion in some ill defined portion of the center section.
Make the bar the usual shape and the whole "center" section is in torsion.
But what exactly is the center, as the forces see it?
With a straight bar, rigid bearings at the end, and arms attached square and solid, the entire bar is "the center"
As soon as you let the bearings distort, or introduce bent ends where bending changes to torsion "somewhere", life gets fuzzy. It's easier for the bar to relax by bending than by torsion, so it will do that - just like the rest of us lazy buggers.
A center bearing keeps the middle of the bar from bending, and acts as a cantilever to make the "center" part outboard of the normal bearings work in torsion.

It can be quite instructive to model this stuff in plastic tubing etc. Much more fun that going to school for eleventeen years and still not understanding it!

FR Millmore

am i right in thinking a centre support is benificial & doesn't reduce the rate by half?
Brad (Sprite IV 1380)

Center support will slightly increase the rate, probably most noticeable on relatively small diameter bars, since they are more prone to bending. Fitting stiffer bushes will have more effect. Note that if snug & solid bearings are used, bending can cause binding, which will make the rate unpredictably and unrepeatably greater.

FR Millmore

Thanks guys I have it now, I hadnt really considered the S shaped bending before,(only the twisting) that would be why the thicker ones will tend to rip the mounts of a less than perfect chassis. It also must accentuated by arms not being at 90 degrees to the bar some of the original designs have arms at 45 degrees or so.
This site may be of interest, it has a table that compares different thicknesses of the same design a/r bar:
For example it tells us that a 3/4" (19mm) bar is about 200% stiffer than the standard 9/16 (14.3mm) !! not surprising it rips the standard mounts out the chassis.
Actually from your information I could easily move my mountings more outboard to good effect, they are already almost 2" wider than standard anyway, another reason perhaps why I have always felt my bar performs better than it its only a 5/8 and with approx 1" longer arms (cut and shortened also)

Ian Webb '73 GAN5

As I showed above, you don't need a table, just square the bar size (in common units) to compare. 3/4=12/16, so 12sq/9sq = 144/81.
For hollow bars, the stiffness is the difference between the tube as if it were solid, and the invisible bar that is the hole. Same for weight.

Note that another major factor is where along the suspension the bar is connected. Closer to the wheel = more effective. Since it is easy and accessible to drill extra holes and move the links, this can be a handy tuning aid, if the bar is such that the links are long enough to stay fairly close to right angles with bar and control arm. You get the same effect by sliding the links along the bar arms to effectively change the arm length, a common means of on-track adjustment, shorter arm giving greater stiffness.

FR Millmore

This thread was discussed between 15/07/2009 and 20/07/2009

MG Midget and Sprite Technical index

This thread is from the archive. The Live MG Midget and Sprite Technical BBS is active now.