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 MGB GT V8 Factory Originals Technical - Three link rear suspension

I've neen looking at different suspension sets ups for the B. Weighing the pros and cons. I'm almost convinced that the three link set up is the best by miles in terms of results for effort expended.
IRS- very nice to have and proably the best as far as performance goes, but not better by very much and certainly the effort that goes into it doesn't generate the return. Something for the hobbyist. To buy one off the shelf (Hawk etc) is way too expensive.
. The problem with a four link is what do you do about locating the upper arms?. Bottom arms already done for you, top arms involves welding attachment points to the axle and lots of body welding and even then you're not really feeding force into the car properly. Then there's problems with interfering with the spring. The coil spring must go into the bump stop location. Costellos set up addresses this but it appears ellaborate and heavy and costs quillions.
All the problems pretty well go away when you look at a three arm set up. There appears to be more than enough room directly above the diff' for an arm. Even if there weren't you could go a little off centre. There is plenty of structure to bolt in an attachment point to the body of the car. Its a good spot to do this because forces would be fed into the tranny tunnel and along the body of the car. I would envisage bolting the other attachment point to the back of the diff'(as per the link below). In fact attaching there would also be a good oportunity to also make an attachment point for a Watts link, instead of using just a panhard rod.
It seems to me that the only tricky construction required is that bolt on bit at the back of the diff' which would not be hard (or time consuming) to make up. Just cut it out of a thick steel plate. Or maybe even a bolt on an attachment point straight into the top of the diff case. A short lenght of channel steel bolted in two places. Or even welded on. Then throwing away the leaf springs & substiuting a couple of short lenghts of square tubular steel, (for the bottom arms to attach) would almost be too easy. Keep the U bolts and hydrolic shockers.
A bit more work in a Watts linkage, but not much and there is still the pan hard rod to fall back on. Maybe a SD1 Diff' cover wouldfit? The arms of about the right length could be got from toyota. Springs would have to be made to measure (unless someone knows of some that would fit, maybe motacycle?)
All this seems very easy to do, and super cheap :).
What are peoples thoughts on this approach. Any suggestions would be greatfully received .
So long as the law allows me this is a go ahead, not just speculation. Any locals know what the position is on suspension swapping? It can't be too restrictive, there are plenty of hot rods around.
Spring ratings? any suggestions?
The best suspension arms? any suggestions? Should I have them made to order? Are rose joints better than bushings?
Is there any real advantage in making the upper arm adjustable? I've read a little about anti squat being controlled by this.
What are people thoughts on the notion of bolting straight into the diff case for arm attachment point? Any one know how thick the diff case is? I've never had mine apart.
Can the Rover diff' cover (with its central bolt) be made to fit an MGB diff'?

Forgot to ask.
I've read that the roll centre for panhard rods depends on how low you set the rod on the axle. Little is said about the other end. If this is so then a Watts linkage could do away with the need for a panhard rod tower. You could bolt your links up on the back corners of the boot. Does this sound reasonable?

Keep in mind that there are also triangulated 3-link setups available that would eliminate the space problem above the differential. I attached a link to show the type of link arm I am speaking of.
This can be used for axle location in a 3-link, 4-link or leaf spring setup and only requires tabs be welded to the axle and body.
Jeff Schlemmer

More questions-
This shows a couple of potential mounting points for a Watts location point/tope center arm plate. Those shallow holes either side of where the diff'cover goes. What are they for? could they be drilled deeper and tapped for a bolt?

Hi Jeff,
You appear to be using essentially a Satchell set up used on mustangs. I was most impressed with this and gave it close attention. There was even a varient could even be set up using "Ubolt on mounting points, a little like costellos.
It also means being able to get rid of the pan hard rod (or watts)However I suspect the battery boxes might be compromised (did this occure in your case?).I particularly want to keep this area unchanged as I intend to put twin fuel tanks there (someday!) and bolt the spare under the boot. Aside from achieving an empty boot ( :)) ), weight distribution improves. The other point is that you are still stuck with making upper mounting points for the upper arms. If you don't care about battery boxes it's a very good approach. I saw some amazing applications on a 4WD site. A kind of W varient.

More questions-
Roll understeer, as i understand this is achieved by setting the trailing arms with an upward slope (that is at normal ride height) as the car goes into a corner, (a turn to the right for example) the rear rolls a little, the inside wheel (RHS) moves down and back a little backwards as it becomes more parrallel to the ground while the outside (LHS) wheel moves up & forward a tiny bit. The back axle steers the back of the car to the right relative to the front.
Now, is Roll understeer really a good thing?. Do practical considerations make it bad? I would imagine that once it goes past parrallel, roll understeer would then start to decrease,(ie increasing over steer) which might not be good at all.
If the concensus is that it's controlable and a good thing then how much Roll steer is good? Obviously you don't want the back doing too much steering all on its own! If you set up so that the arms are parrallel to the ground at max' suspension travel (if that's possible) then there's the problem of too much slope on the arms causeing the car to squat big time on accelleration. How may the top arm be used to control this? People must have worked out a compromise at some stage relative to power to weight ratio and suspension travel.
Any one wish to hazard an answer?


Peter, do nto get to involve in roll steer for the street. How much roll steer is needed? Well... it's like what type of underwear you use, everyone wears different types, but it serves the same purpose.
It all depends what you like in a suspension set-up
What I like you would not like.
The main point is to keep oversteer to a very minimum for safety on the street, all vehicles including the Z06 Vette have understeer build in on them for safety.
As for a 3 link(4 link including the panhard rod) it's a good set-up.
We are working on a 4 link Satchell for the B.
This is a very common set up for race cars and it can be taylor for the street. The roll center can be adjusted by were the bars are located, and yes you are correct, eliminates the panhard rod.
Mustangs had the upper arms angulated to the outside and were very short, it also use a couple of special calibrated shocks to eliminate the twist thus improving traction (wheel hope) on accelaeration and braking.

A 3 link + panhard rod is not an easy one to design, the top link has to be correctly located and strong end point to absorb the shock produce by the axle.
In a 3 link you need a very long panhard rod.
If in doubt, always keep the rear suspension softer than the front this will alowed you keep understeer.

If you are not in loved with your battery boxes, you can then really build a very nice 3 link or a satchell 4 link.
Just keep in mind that you are going to drive the car in the street 98% or 100% of the time.
Use thick wall tubing or Chromolly tubing for your rods.
Bill Guzman

Peter, I'm not actually using that setup yet. Its one of the many variations I am considering, but it looks promising. At this point, I'm considering cuting out the entire trunk floor and the battery boxes to build a real 4-link though. I just haven't had the guts to do it yet. I have too many other "tuning" projects in the works.
Jeff Schlemmer

OK Bill,(or anyone else who wishes to have a say) it sounds like designing in a whole lot of roll steer is not a good notion. I've done a whole lot more trawling around and come up mostly with people expressing some fairly intense and opposite points of view. Rollsteer; likly one of those notions that come into and and go out of fashion. I see allot of it in Catamaran racing where the current national champion has a rig set up so every one thinks it great....that year.
I'll try for dead neutral, lower trailing arms parallel to the ground. Actually there might be a tiny bit of roll OVERsteer in the current Hotchkiss set up with the outside wheel moving back a little as the spring is compressed. It couldn't be much but I think it would be a good thing to eliminate.
A solid mounting point to the body is not a problem. That area at the back of the transmission tunnel looks very strong. From what I can gather the top arm mostly is in tension so if I put a plate on the other side (inside) of the heel board that should feed stress evenly into the transmission tunnel, thence the whole car. I might even run a stringer along the top of the tunnel.That would be hidden under the carpet etc.
What has got me mildly perplexed it the notion of having the middle (top) arm shorter than the bottom arms. One article (bits quoted below)) suggested that the top arm be between 20 and 25% of the length of the bottom arms.

"Lensing says to avoid this common problem (roll steer), racers angle the upper arms, which helps induce forward bite through anti-squat but doesn’t create roll steer. “The more down angle you have at the front, the more anti-squat you have under acceleration. The front of your pinion tries to go up, which pulls on the upper link, forcing the rear tires down into the track.
“The downfall of this is that too much angle causes the car to get loose going into the turn,” Lensing says. “That’s because when you decelerate the pinion goes down, and with a lot of rear angle, pulls the rear up into the chassis. This makes the rear of the race car feel loose going into the turn.”

Lefthander Chassis recommends 5 to 7 degrees of downward angle at the front. “Also, your upper link should be about 20 to 25 percent the length of your lowers,” Lensing says. “This makes about the smoothest transition for traction. Too short, and the traction is abrupt and then goes away; too long and it’s harder to get forward traction.”

That is one very short arm; surely the guy means 20 to 25 % shorter than the lowers, not of the lowers. Even then that's pretty short, what about pinion angle? All the four arm set ups take great pains to get the uppers the same length as the lowers to ovoid changing the pinion angle. Why would this not be a problem in three arm set ups? My instinct was to make the upper arm the same length as the lower. Is pinion angle not really a worry after all? With the Hotchkiss set up you do get small changes in pinion angle.
Why is a short arm better at putting down force on the tyres than a long arm? Is it something to do with the angle? Shorter=>angle. That doesn't really make sense. However the Mustang also had short uppers.
Incidentally if this is true about the upper arm being short then this is ideal for MGB's with the problem caused by the sloping heel board. Also 20 to 25 % of the lower is about the same length as a front hydraulic shock absorber which is food for thought. Especially as I have a spare.

I liked the Satchell set up but upper angled arms/links would have to move the roll centre up surely and if you put them on the bottom you've still got a problem about where you put the upper parallel links. I would imagine (like Costello) you would bolt a structure to the shock absorber bolt location. Run forward to the heel board for the upper link attachment point and back a little for coil overs’s mounting point. Also if you don't want a whole lot of squat the point of attachment to the axle of the bottom angled arms needs to be about 4 or 5 inches below the axle (if you're running them to the leaf spring mount). Plus of course you lose the back corner of the battery box, or in my case, potential fuel tank!
Three arm is the way for me I think. It's just so easy to do. My main concern is working out why they run a short upper arms. I did notice that on toyotas (that all seem to use four link), they all have short uppers as well.
Now there is a possiblility that might take the worry out of this. Assume that Toyota got it right and simply use the links straight off a toyota (or similiar). The bottom ones are about the right lenght (18 inches). Put a pair of shorter toyota upper links close together either side of the diff' much as Jeffs suggestion. No clearance problems at all then, even for much larger diff's. There are a couple of flanges either side of the diff that look like a good place to bolt to.
I could put the Body attachment points either side of the top of the transmission tunnel with reinforcing stringers running up the inside. An even better distribution of forces.
This would be a kind of cross between a three and four link.

I'm using this thread to collect relevant info' on 3 link implementation. Useful to me as well as others.
I am finding that it might be a good idea to offset the upper link to the right. From a discussion between Beath and Billy Shope on

"There is a simple way to equalise tyre loading with a live rear axle using a torque tube attached to the body at a point offset to the right. To find the location of that point draw a line in plan view through the diff centre towards the front of the car and at an angle to the centerline with a tan equal to the diff ratio; ie if the ratio is 3:1 the line would pass through a point three units forward and one unit to the right of the car centerline, attaching the torque tube anywhere on this line will give equal loading on the rear tyres ( draw adiagram and think about it)"
then Bill "Your description of an offset, or asymmetrical, trailing link is the best solution for equalization of rear tire loads. This was used by Jaguar in their early C-Types. (Late C-Types had IRS.) The Jaguar design, however, used 3 links. Two were symmetrically placed below and the third was above, offset to the right (Your suggestion will work, of course. It's just that the 3link is more easily implemented.) "
For an 18 inch bar/link & a 3.9 diff'), thats 4.5 inches to the right. Should just be able to fit it up against the battery box. Could bolt the axle mount to the RHS of the diff', drill out and tap bigger boltholes. The following middle axle mount is very elaborate, however also doesn't seem to support haveing a shorter bar at the top. It also demonstrates that MGB's are far better set up for a three link!

Roll bind and loading- Four link set ups, and even sachel, all score a bit of roll bind Looks like I'll be leaving the toyota uppers alone. One very strong heim jointed link is the go, or rather, because of road noise, a heim joint at one end and a rubber bush at the other.
More from Bill on squat- Can anyone explain this?
"With horizontal trailing links, your friend will experience a considerable amount of squat. If the intersection of lines through the links, as viewed from the side, falls on a line passing through the rear tire patch and intersecting two other lines, one a horizontal line through the center of gravity and the other a vertical line through the front tire patch, the car will neither squat nor rise."

He is obviously speaking of familiar points of reference, which I don't have. (ie doh)

Peter, you can not make a rear suspension perform as you may like it right after the install.

First erase the racing tuning from your head.
Second, build it strong and with the right pinion angle.
Thirth, TEST AND ADJUST, then TEST AND ADJUST, then.....get the idea.
What is good for you in suspension set up may not be the right set up for me or anyone else.

Start with parralell links to the ground, then TEST with a stop watch until the car and same driver will not get better times. Use a slalom course. On a street car it's all about transition, safe transition with out sliding the rear wheels (oversteer)

Imagine driving your car in the fwy and one day you may have to do a evation manuver (transition)you would want your car to be able to perform corectly without loosing control, that is what you want.

Yes, you can adjust the bars to produce more traction on side more than the other.
This was a common practice when posi traction was out of reach. You would elevate one link (opposite to the one living the black mark) until both tires would live two black marks. This was great for straigh line, which is what you would on the dragstrip to launch the car straitgh, but it will not handle correctly on transition for a road car.

Bill Guzman

Thanks Bill,

Got under there today with a ruler and had a very close look. In so far as centre upper link goes there is even more room than I thought. At normal ride height I have at least between 12 and 15 cm clearance above the diff'. 15 if I go off centre to the right, which would help equilize tyre load. I've got 5 cm more upward before the suspension is on the bumpstops, Plenty of room for a link bush. The diff even has a flat bit on the top right hand side. It is also effectivly square on the back diff cover on the RHS. I figure that I can bolt an unsidedown L shape to attach the centre mount to. Will enlarge and retapp the relevant diff' cover bolts. Mind you I think it might be much easier to just bolt straight through the top with two big bolts and a plate underneath. Is the top of the diff' strong enough do you think?
With respect to the shorter upper arm problem. In order to get a pinion angle change of > 3 degrees (2-3 dregrees is within bounds) the upper arm must be less than 68% of the lenght of the lower. That's for a suspension travel of 2 inches upward , ride height to bumpstop. More realistically if you were going along, bouncing up then and down 1 inche, 2 in total (a fairly bumpy ride) then that length gets down to 36% of the bottom arms. So it looks like there is no problem at all with the shorter upper arm necessitated by the MG's sloping heelboad :)

The springs are another matter. I seriously doubt I can get springs that will fit into the bump stop locations. They would have to go from fully compressed at about 2 inches, expand to about 10 inches for full bounce strap suspension travel (and ride at 4 inches. Any suggestions? Conceiviably I could put some sort of bolt in spring capturing disks into the perches. I don't know much about springs at all. Can they be got that small?. They would need to finish either end in a small coil to work. Has anyone fitted coil springs into the bump stop location before?

The only other option appears to be coil overs, which I really didn't want to do. Hughly expensive and everyone seems to think that at their softest they are too hard.

I have coils-over shocks installed into the location of the old wishbone shocks in my 77B. I didn't find out why you are trying to fit springs into the bump stop location. So maybe you already covered this option.


Hi Rick,
I'm avoiding coil overs in the back because although very good in the front, every thing I've read says that in the back they are too hard even at the softest setting, and they are too expensive for what I have in mind. I'm designing an entirely bolt in, bolt together trailing arm suspension any one can make, very cheaply. All you need is an angle grinder and a drill. The idea is that anyone can look at it and say "I can do that".

I've pretty well worked out all the details now, for anyone interested.

The springs will be 145mm ride height with 330lbs on each. That’s from the top of the axle to the base of the chrome bumper bump stop. Subtract 10mm for spring perch etc, so 135mm (330 lbs loaded) ride height. They need a free standing height of greater than 185mm. outside diameter of around three inches.
They may not have a compressed height much less than three inches. This is possible I believe. Obviously these need to be ordered from a spring maker. I might get very lucky and find an off the shelf pair that could be reset, infact odds on they ore out there somewhere!. Should be less that $150 to have made.

Where the top arm attaches to the diff'- A piece of channel steel, two inches wide, set along the top of the diff', open end up, a little to the right side of the centre line. It hooks over the back of the diff' with a bolt on angular section that bolts to the diff' cover bolts. At the front it is held in place by a very large U bolt going around the front of the diff, (sourced from a truck axle possibly). All this is bolt together, right angles, so easily made, not too heavy and should be bullet proof as the only bolt under tensile is the huge U bolt. The others are in shear. The top arm/link (of the three link) locates into the U shaped channel steel on top of the diff. There is obviously considerable scope for adjusting the pinion angle.

Where the top arm attaches to the body- to the heel board just to right side of the transmission tunnel, up against the battery box. There is room for another piece of that two inch channel steel. Set vertically along entire height of the transmission tunnel (approx' 8 inches) and attaching to the heel board brace at the bottom. A long flat piece of steel (4mm?) will bolt to the vertical piece of channel and run up the inside of the tunnel. I might run this stringer at an angle down to the middle cross member, probably over kill though. There will be another stringer running from the top of the vertical piece of channel along the battery box. This simple set up would be very stiff and strong and evenly feed any stress into the entire car.

The top arm should slop downwards (from the diff' to body) a few degrees to prevent squatting on acceleration and wheel hop, according to the experts. A figure of one inch down at the front end was mentioned. Anyone have any thoughts on this?

The lower arms need to be around 500mm long. They will bolt into some short pieces of steel (that 2 inch channel steel again!) held in place by the old U

Shockers stay the same.

Would the chrome bumper style axle bump stops be suitiable as spring perches, anyone? I've never seen these, & the rubber bumper ones are too thin. Can coil spring perches be bought, or will I have to make my own?

I have reluctantly abandoned the Watts linkage idea, not enough clearance between the diff and the petrol tank.

Panhard rod-
The axle end of the panhard attachment point is to be on top of one of those short pieces of channel steel that the lower arms bolt to.

The body panhard attachment point, the 'tower'-
I will take a length of box steel, 3 or 4 mm wall thickness, and cut two of the sides, down a way. I will then push the remaining sides up over either side of the LH chassis member. One side goes through the left hand corner of the floor of the boot and the other up the wheel arch. I will then drill two holes through tight up against the upper web of the chassis member & bolt my new tower in place. A little like the front bumper irons bolts. The bottom of the tower will be cut so as to fit tightly up against the bottom of the chassis rail. I will probably put a third bolt at the bottom of the chassis member. This might also have to double as a bounce strap bolt as that may become a casualty since it’s in exactly the wrong spot. That tower location should triangulate up through the back of the boot and wheel arch and be very rigid.

Link/arm length-
Due to the location the upper arm can be offset to the back of the diff so can be exactly the same length as the lower arms if I want. This would make things a little easier/cheaper if I can get three all the same length. In fact the upper arm can be up to 50% shorter that the lowers before causing either significant pinion angle change (2-3 deg.) or breaking local laws.
Hopefully I will be able to source my 3 arms/links from other cars. An off the shelf deal.

The biggest expence will be getting the engineers stamp of aproval.

Well, anyone spot anything silly? I'd rather know now.
Anyone who's got the concentration to read through all the obove is certainly bright enough to spot any gaffs.
Can anyone suggest a potential spring?
Can anyone suggest a suitiable arm/link; strong, around the 450 to 500mm length?

Another thought,
If I drill a series of holes in both the "top-of-the-diff'-mounting-point-length-of-channel-steel" and the "heel-board-mounting-point-length-of-channel-steel" there is great flexibility for fine tuning the upper link with respect to antihop, anti squat and pinion angle. A good idea to do it before its all bolted in!.

Just realized I've made a couple of errors.
Axle plus springs is 205lbs. Half on each side. This of course also doesn't include the wheels and tyres! These also have to be weighed and subtracted. So in Jeffs case thats roughly 1123 minus 205lbs minus whatever 3 wheels weigh. Divide by 2 for each spring.
Pauls "400lb to 450lb for a CB roadster, with rates of 105lb/inch (that is for a CB GT)
is starting to look good again"!.

Peter, I've been running a three link on my 74 B V-8 for two years, its great and not really that complicated, ask anyone that has driven it and they will tell you it handles like a dream. I don't have enough time to answer all your questions right now. But, shock placement will be behind the axel, panard rod will be in front of the axel, the big thing is KEEP THINGS SIMPLE, I'm running a 302 Ford, 323 hp @ the rear wheels and whigh 2200 lbs. Please e-mail me and I'll fill in the blanks

Steve Carrick

Steve, when you have time, please share your design with all of us! It would be a HUGE contribution to the archive!
Jeff Schlemmer

Terrific Steve, I've struck gold!
Very good to see your post. Three link has none of the problems of four link and makes so much sense on an MGB, given all that space above and infront of the diff'. Especially if you are running 323 BHP!
With that sort of hp I would guess that you're using a diffent diff' and had a top mount welded on?
I'm currently in the process of changing my internet provider (timing!)but can be got at work on

"Sherman, Peter" <>

Which is way up in central Australia, and I won't be back up there until the 7th. I'm currently using one of the computers at LaTrobe University.
I'll certainly email you from there(work), there is no substitute for experience.
Pan hard in front sounds pretty unique!


One thing you have to remember is that if you retain the leaf springs you already have 2 links.

I used to run a top link, Panhard for lateral location and two lower links beneath the leaf springs running from pick-ups attached to the lower spring plates forward to the front spring eyes, where I welded mounts right below the spring eye mounts on the frame (talking MGA here but similar on MGB).

One has to remember that the front half of the leaf spring, aside from just acting as a spring, also acts as a suspension arm, and if you simply installed arms below that, you'd have a situation where they would seriously bind as the diff moved up and down, with the two sets of arms fighting each other.

Solution is to either run coil springs if rules allow, or convert the leafs to a 'floating' configuration with spacer tubes between top and bottom spring plates so the springs are free to perform their springing function but no longer contribute to location.

That works very well, and was the system used by Huffaker on the SCCA MGBs and midgets.

As I run only vintage now and as they don't understand fancy suspension systems, my lower arms (which are very effective as traction bars) are hanging on the garage wall and I now run only the Panhard which is apparently known to them as a period mod, and the upper reaction arm from top of diff to frame, which acts as an anti-wind up arm, which they don't notice or perhaps don't care about.

I am glad I never went with the 4th arm and coil-overs - too much to retro-engineer for vintage when someone founf they didn't like them....

If I were going to do a street car with serious handling intentions, I'd certainly use a Panhard and at least upper link, and would think hard about a coil spring conversion with lower links/traction bars and the additional upper link. I would probably leave the lever shock in place as it has less unsprung weight than a tube and doesn't take up much room, plus you don't need to fabricate a couple more brackets for the tube shock and figure out what the heck rates you need in them.
Bill Spohn

Someone in this thread referenced my posts in, so I thought I'd add some further information. Since that exchange, I've come up with some alternative arrangements which will also dynamically cancel driveshaft torque with a beam axle RWD car. You can check them out at my blog:

The effects of roll under/oversteer are discussed in my blog and also in the student workbook which accompanies the Millikens' "Race Car Vehicle Dynamics." The significant point in these discussions is that neither roll understeer nor roll oversteer changes the wheel loadings in a turn, so vehicle stability is never a problem. But, when you consider the driver/car system, the presence of roll oversteer can be, at the least, unnerving, since the necessary steering corrections mimic true (dynamically unstable) oversteer.
Bill Shope

Thanks Bill,
Going to take a little time to get my head around all of that, but it certainly looks gospel.
Your calculator for the three link. I'm assuming it's set up for a link at the bottom? Getting some weird numbers!

This thread was discussed between 24/06/2006 and 05/07/2006

MG MGB GT V8 Factory Originals Technical index

This thread is from the archive. The Live MG MGB GT V8 Factory Originals Technical BBS is active now.