New Subaru Diff, Cv And Coil Over Conversion

[itiejim]

Thought I'd update the forum on the Subaru diff conversion that I've done to my '67 Spitfire.

 

There are various arguments over whether it's worth the hassle etc. However, for my car, fitted with a tweaked VVC 1800 K-Series engine and type 9 'box, I wanted to eliminate the weak drive shaft and get a stronger, cheaper, LSD in there. I've always been fairly happy with the swing spring suspension, however, whilst I was designing something it seemed sensible to see if I could reduce the camber change and provide some options for more easily changing ride height and spring rate.

 

I've been thinking about doing this for years and had been following Karl Dandridge's efforts quite closely. He has fitted a Subaru diff to the Herald Coupe he's got and had been promising a kit for ages.... I got bored with waiting. Karl has made some progress towards a kit and had some drive shafts made up with splines to suit MGF CVs on one end and Subaru CVs on the other. I was able to buy a pair of them of him which helped.

 

I didn't want to modify the chassis and chose to mount the diff in a 3mm steel three sided box, locating in in the standard mounting position. The diff bolts into this box on the sides by using the top three drive shaft seal retaining bolts. I have then welded a Z section from the front of the box forwards to meet the front mounting holes on the Subaru diff. From this I have used rectangular box section with the original Triumph diff "ear cups" welded on to mount the front part of the diff to the chassis. To hook the rear up I know that some have used a bespoke alloy casting. I see that Karl had one of these made up but I was not particularly convinced by the strength or quality. Therefore, I chose to have the original Subaru diff back plate machined down to the minimum possible to allow clearance for the crown wheel and welded a flat aluminium plate on the back. This allowed me to add a fourth side to my box (closing the rear part behind the diff back plate) and permitted me to make up some hangers from rectangular box section with tube welded in to take the standard rear diff mounts. This allows me to fit the diff in virtually the exact same position as the original one (it's probably 15mm further forward in terms of drive shaft position).

 

Vertical links are GT6 ones machined out to Nick's specification to take the MGF bearings. I've also ordered some of Nicks drum centering rings to ensure correct drum location.

 

I have used Canley Classics alloy Rotoflex wishbones and mounting brackets welded onto the chassis and am currently waiting for the arrival of a set of coil overs (12.5" extended length with 250lb 7" springs - 1.9" ID). The extended length ensures that the drive shafts don't hit the chassis on full droop whilst spring rate and length is a guesstimate at the moment. 1.9" ID springs are cheap enough to buy some more it I need them! With the type 9 'box and everything moved well back in the chassis I've ended up with a 50.8 cm prop shaft, which is on order from http://www.dandfltd.co.uk/

 

Upper arms are tubular steel with standard Triumph sized bushes fitted to tubes on the end. It's an arm with a single point on the inner and outer ends as the tie bar provides strong location and the whole thing is massively better located than the standard swing spring set up. The inner mount is welded to the box which gives a slightly shorter top arm compared to the transverse leaf, and it is also positioned lower down. I've yet to get a gauge on it, but there is no visible camber change between full droop and compression.

 

Now, I know that a photo says a thousand words, however, I've been busy in the garage and just realised that I haven't actually taken any of the kit out of the car. Ooops... Below (should be) a few photos of the diff and suspension as it sits in there now. I'll update further once the coil overs and prop get here and I can actually try it.

[pomwah]

Looks like a good solution, I'd really like to see more photos, I'm sure we all would, and particularly after the coilovers are fitted. Looking forward to the road test results!

[itiejim]

Thanks, tbh I thought I'd taken some photos of the box whilst everything was out of the car, but I can't find them. I'll post up some more and some first impressions as soon as the rest is fitted up.

[Nick Jones]

James,

 

Thanks for posting this - very interesting.

 

I really like the idea of the "diff in a box". I've toyed with something similar (not got further than day-dreaming plus the odd sketch so far though), first in the context of a Sierra diff in a small chassis Triumph and more recently in the context of a Granada or BMW diff in a saloon. The BMW E30 is a sort of halfway diff in a box solution itself.

 

So really good to see one that has actually been done and looking very promising.

 

I have a design observation on your top link replacing the spring. I made the same comment on KDs original design and I thought he changed it, though I doubt that had anything to do with my comments. My concern is that while the OE leaf spring is pretty rigid front to back and will give reasonable fore/aft location of the top of the vertical link, the single arm with a single, narrow bush located well inboard will be comparatively flexible as even a small mount of deflection of the inner bush will be amplified by the arm length and allow the top of the vertical link to move around quite a bit. This is reasonably easily fixed by making the bush a lot longer or, better IMO, making it more of an A-arm with two inboard bushes spaced as far apart as you can get them on your box.

 

Keep up the good work!

 

Cheers

 

Nick

[itiejim]

Karl did start with a single link before going to a triangulated one but, having chatted with him about it extensively I know that he then settled on producing a single arm for his (still pending) kit (though I don't know why)

 

I'm not aware that he took any professional advise on it, however I did!

 

I trialed this set up on the rear part of an old chassis before thinking about fitting it to my car and, as I wasn't sure how to go with the top arms, I took the whole lot to a local rally car preparation specialist who has experience designing bespoke suspensions for him to review. It was on his advice that I went for a single point inner and outer top link and he actually produced the top links and fitted the pick up points for me. I'm no expert on this, however, I understand that the re-located tie bar used on the rotoflex suspension (when compared to the non-rotoflex) is positioned in such a way that the whole thing - lower arm and tie bar, works as a single wishbone (therefore with two inner pick up points). This should provide more than enough stability.

 

In any event, compared to the non rotoflex suspension, the stability of the vertical link is massively increased. Have you ever tried trying to rotate the the jacked up wheel of a non rotoflex car with the handbrake on? The amount of movement is incredible.

 

I can't say that I expect to bolt this up and for it to just work perfectly out of the box, however, having located the diff securely (which acts as the platform for the upper inner links) if I need to revise the wishbones then it's no great issue. I'll keep this thread updated with how it actually works on the car.

[pomwah]

Thought about this a bit more.....

 

Don't see the radius arm in the images, are you using it? How long are the upper and lower arms. If the upper is longer than the lower there must be some camber change, from positive through neutral and back to positive, might not be a lot if the difference is small. If you start with some negative camber the change will be towards positive.

 

Do like the idea of the diff mounted,top arm it's something I hadn't considered and it would be relatively easy to move the mounting point, or points even, further outboard.

 

Really interesting.

[itiejim]

Yes, definitely need the radius arm - I only got around to fitting them today.

 

The top arm is longer than the bottom arm, though not as long as the transverse leaf. The inboard mounting point is also lower than the transverse leaf which, clearly, is on top of the diff rather than to the side of it.

 

I've just made some measurements and camber change goes from +3 degrees at maximum compression, to 0 degrees at the mid point, and back to +3 degrees at maximum droop- so exactly as you predicted!

 

I could try moving the inner lower arm pivot point up to the top hole and see if that improves things. Either way, as a starting point I'm not too unhappy with a total change of 3 degrees.

[itiejim]

Hi folks, thought it was about time I updated this thread with some feedback after having done a couple of thousand miles. I've also had the opportunity to drop the diff out again, so took some photos if anyone ever wants to replicate (or improve on) this.

 

Firstly, driving impressions. In short, it's a huge improvement. Now, I know that, as the creator of this set up I'm likely to be biased, however, in defence of my objectivity, I only ever took this job on in order to get a solid axle with an LSD in as I was always happy enough with the swing spring rear suspension. I do have to say though, that the far improved axle location is immediately apparent, in particular in transition from power on to trailing throttle. The car is far more predictable and it's now easy to drift (if that's your thing) - something I always struggled to do smoothly before.

 

Problems: firstly, the springing was too soft, initially I fitted 6" 250lb springs, however, they were just too soft and needed winding up to get the ride height and compressed too much under load. I swapped them out for 7" 325 springs which gave the right spring rate but the increased length meant that the spring platforms interfered with the CV boots... In the end a 6.75" 325 lb spring seems to have done the job.

 

The only other issue I had was some flex in the diff box which I wasn't happy with, hence the reason for dropping it out for a minor reinforcement.

 

All in all, well worth the effort.

 

 

[Spit131]

Any updates James ???

[itiejim]

Updates...

 

Well, the suspension and rear drive train are fine. Spring rate and length is great now, the car is predictable and finds traction well. The bad news is that I've destroyed my engine

 

For some reason the (remotely mounted) oil filter blow off the other day, effectively pulling the threaded pipe out of the allow housing. Needless to say, I didn't notice it immediately, and by the time I'd got 1/2 mile down the road and noticed the "odd engine braking" it was too late.

 

So, whilst I've got to have the engine out I thought I'd go the whole hog and lift the body. I've had the car for over 20 years and it's been fairly comprehensively cut about in that time, so I guess it's probably fair to lavish a bit of love on the old girl.

 

Now, just got to find the last thing holding that body on (no - it's not the earth strap)...

[pomwah]

Bugger. Handbrake cable?

[GT6 Nick]

The two bolts in the floor pan behind the seats, hidden under a layer of tar. Triumph's wee practical joke.

[itiejim]

Thanks for the feedback chaps. I got the two bolts out from the rear "seat" area but think it must be the handbrake. I took the cable out from the handbrake and pushes it through the hole in the floor. Thinking about it now though, isn't there the pivot point attached to the body which it would still be hooked up on? Have to take a gander in the week when I can get some more garage time.

[cdn12345]

i have a subaru limited slip diff in the loft for this project one day..... i was waiting for the one on the saloon to explode, but as it refurbished it may take a while. As crown wheels are getting hard to find its got to be worth doing, I t would be usefull if a scale drawing of the carriers could be made for different applications.

[GT6MK3]

James.

 

If you're able, I'd love to see some photo's, measurements etc with the body off.

 

My GT6 swing spring has a set of Nick's CV conversion axels waiting to go in, but no one has written the book on converting swing to non-roto cv, and I'm curious...

 

Can you help?

 

C.

[Nick Jones]

There have been various writings on various forums about converting swing-spring to roto style.

 

There are some "gotchas"

 

Obvious:

- You need to weld on the chassis brackets to attach the wishbone to

 

Not quite so obvious:

- The radius arms AND their attachment points to the tub are different.

- The brake pipes take a different route (including hard lines)

- The handbrake cable run is different and the swing axle cable run gives significant variations in tension with suspension position (basically the brakes come on as suspension compresses).

- The damper pick-up points on the roto VL don't always line up very well with the chassis brackets.  This varies from car to car and can usually be overcome with washers as shims.  The roto cars used longer dampers and upper brackets welded to the tub in the wheelarch.

- If you have a wide track swing spring setup the roto arrangement reduces the track by 1" per side.

 

Opinions vary as to whether you can use the swing spring with the roto suspension.  I don't know the answer!

 

Nick

[motov8id]

gt6MK3 do you want to install  CV joints on a swing axle rear suspension? That's not possible without a lower control arm

[itiejim]

Thought I'd update this thread with more photos since the body is off!

 

 

[Nick Jones]

I like that.  Would have a wider bush on the inner end of the upper arm...... but that's all.

 

For those luddites (me included probably) who like the leaf spring, do you think the concept could be adjusted to retain it?

 

Thanks for posting the pics

 

Nick

[itiejim]

Thanks Nick. The inner bush is a front wishbone one, the outer is the original leaf spring one (or perhaps a Spitfire tie bar one?)- I tried to keep things as straightforward as possible!

As for retaining the transverse leaf, why not? It would be easy enough to make a mount on top of the diff box. Personally I preferred to run coil overs for the greater spring choice but understand why others may not wish to.

[esxefi]

hi,itiejim

 

unless you have modelled that set up on a computer for camber change to me it looks like you will induce massive positive camber on roll.

 

the top arm is nearly twice the length of the lower wishbone.

for a given arc the top arm will prescribe a larger radius than the wishbones' one resulting in positive camber at the wheel.

 

you need to swap the lengths or at least have the top arm fulcrum centre distance equal to the lower wishbones one.

 

an easy way round it is to weld some 1 1/2" id box steel from where the top arm connects to the diff cage and extend it out to match the wishbone fulcrum length and shorten the top arms to suit.

 

i'm only going by the pics so if i'm talking b*ll*cks disregard this post.

[Urge]

It's no different in length to a leaf spring?????

 

Make what you want of that observation.

 

One thing I would draw yr attention to.

 

Imaging what goes on to the hub and vlink when the brakes are used.

 

What's it doing to the set up.

 

Now ask yr self. Is my set up better or worse in this instant than a leaf sprung one

[esxefi]

true about the leafspring but it doesn't follow a true arc like the top arm does,the further you go towards the spring eye the more movement there is so making a theoretical short top link.

that's how the original triumph set up manages to get very small negative to zero camber change on bump and use the short lower wishbone.

[Spit131]

To my mind , having twisting movement in the VL is undesirable and unnecessary .

No point fitting more stable calipers to a VL if the VL is free to twist even more .

Although cost effective and not disastrous the weeknes of the Triumph set up is primarily the fact that the hub is only held stable on one side at the bottom , the single arm wishbone combined with the tie rod will hold the leading bottom leg of the VL securely but the only thing preventing the rear leg from twisting rotationaly under braking as Pierro points out is the leaf spring . I don't think the leaf spring does too bad a job of this but to make a real improvement one needs to fully stabilise either the top of the VL or the bottom of the VL or in an ideal world both ( Jango ) as sharing the load between both wishbones/ ' or alternative arrangement ' will dilute loads further .

If you fit a solid ( which does not have to mean heavy ) bottom wishbone which is attached to both legs of the VL and also has two mounting points on the chassis your VL CAN NOT twist under braking . Then the top link be it a leaf spring or rod affair is only there to set distance between the diff and the VL ( all though I understand the principle of wishbone lengths playing an important role I do not pretend to know the effect of the leaf spring compared to itijim's arrangement above as esxefi has questioned , need some heavy maths/engineering know how to answer that , not in me ) If this top ' member ' assists with the stability all the better but if the bottom wishbone is stable enough either of these upper type of connection is unlikely to play a part as there will be more give than in the lower wishbone ( which will have hopefully almost none ) so their stability benefits will actually never come into play . As stated above this will put additional strain on the lower wishbone bushes/rod ends but ' IT WILL BE STABLE ' . Don't forget that torque caused by braking will not change ( unless you upgrade the rear brakes ) what ever system is used , ITS THERE ALL READY .

Pierro is right that people often forget braking effect on the VL and its suspending components .

My advice is minimum of one very stable wishbone to start with , attached to both legs of the VL not just the front and as wide as possible double mounts on the chassis , then one has some freedom to play with the top spring/arm or whatever , although I think as esxefi suggests the effects of camber change may not be as critical as people think . Prob depends a lot on tyres used .

Feel free to correct me as there are usually aspects I have overlooked , sure Pierro will find at least one !

Edited January 19, 2015 by Spit131

[pomwah]

Could be, .....more than likely really, my thinking is wrong but when trying to imagine how the leaf spring moves under compression I keep coming back to the idea that it extends as the arch of the spring flattens, increasing positive camber change.