I'm working on swapping an FC subframe into my race car FB. Like many before me, the reasons for the swap are more precise rack and pinion steering, more modern and HD hubs, better brakes - and a project to occupy my time and imagination.

Preamble
I read all of the old threads on the topic and the issues of the change in roll centers. Some bullets for the uninitiated:
- the FC frame rails are higher in the car than the FB
- the FC subframe is taller than the FB subframe in order to accommodate the higher frame rails
- if you bolt up the FC subframe to the FB frame rails, you effectively raise the body of the car and change the roll centers and a few other things that I probably haven't fully grasped.
- when you install the FC subframe, you should try and position it so that the wheels are somewhat centered. This usually means that you shift it back about 1" from the OE studs in the FB frame rails. This seems like it might introduce an issue with decreased caster but I haven't gotten there yet... That might or might not be an 'issue'.

The 'better' way to do the FC subframe swap is to account for the differences in height and either:

1. cut the FB frame rails so you can mount the FC subframe at a height that addresses the roll center issue. Some Champ Car endurance racers did this: https://forum.champcar.org/topic/161...4-swap/page/3/

2. cut the FC subframe so it is shorter and mounts the FC subframe at a height that addresses the roll center issue (user economiser did this)

I'm going to do #1 cuz the subframe is oddly shaped and it would be a biatch to try and cut that accurately (for me). I also briefly considered hacking a Miata subframe and use the LCA and steering rack from that - or just making my own subframe but both are more work. Let's start with measuring stuff so we /I understand the differences and how to approach this.

Measurements
Let's talk about how I did my measurements so that hopefully someone can correct my understanding before I start making mistakes or reinforce that I'm on the right track.

I spent the afternoon with both subframes measuring differences (height, width etc) and the height that the lower control arms mounted compared to the top of the subframe (as a proxy for the bottom of the frame rail). As I understand it, if I cut the FB frame rail by the difference between the FB and FC lower control arm mounting pivots, that will preserve correct roll centers.

Here's a pic of the FB subframe with me trying to take measurements and picktors simultaneously. It wasn't easy. I later put the phone away so I could use both hands to measure stuff as accurately as possible. The pics were more for reference (and to share with the rest of the class).
by https://www.flickr.com/photos/26005589@N06/, on Flickr

It's hard to judge by looking at the Square Rule and taking into account any changes in perspective from the camera but when I measured with the Square Rule and some accurate machinist rulers, it looks like the pivot for the front lower control arm on the FB is 72 mm below the top of the subframe - i.e. 72mm below the FB frame rail. By my measurements it is 72mm both on the front and rear of the LCA bolt. I measured both sides to see if there was any difference in mounting angle.

Moving onto the FC subframe... It's taller. Duh. The black line is the FB. The mounting pivot for the LCA in the FC subframe is 98mm below the top of the subframe (again, acting as a proxy for the frame rail).

by https://www.flickr.com/photos/26005589@N06/, on Flickr

Conclusion - If I'm understanding this correctly, these measurements tell me that the FC subframe needs to be mounted 26mm higher than the bottom of the FB frame rail to maintain roll center geometry. So I'd need to cut 26mm out of the bottom of the FB frame rail to mount the FC subframe. Someone pls correct me if I'm mistaken...

Let's move onto the measurements of the rear part of the front subframe in the next post. It's dinner time.

 

I have been meaning to update my now decade old FC subframe swap thread for a while with pretty much this exact info, so thank you for posting. As you suggest, putting the FC subframe on the FB rails places it too low and puts roll center somewhere below ground - not ideal. Stiff springs and a fat sway bar will offset much of this effect, but ultimately moving the subframe up is the better way to go.

I will have to put a collection of pics together, but raising the FB frame rails 1" is what I ended up doing to get the FC subframe sitting happily in the FB. This is virtually the same as the 26mm you came up with. This puts roll center at a decent spot for moderately lowered cars while still keeping the engine below the hood and allowing the use of a strut tower brace etc. Again, I have many pics of this as I have done it on all 3 of my FB's (along with another FB which I later parted) and it comes out very nice - the difference is notable. I will attempt to get some pics together for you tonight, sorry if I hijacked your thread

 

@82transam , Thanks for chiming in. I read your threads and many of your posts on the topic so glad to see your confirmation. I didn't notice in your threads that you had gone this route so please feel free to share pics with the rest of the community.

Back onto measurements... I had read various people having different measurements for the mounting of the rear of the subframe. 1/2", 5/8" and ~1" were common references. The inconsistency had me puzzled though so I measured mine. I placed a level across the subframe front to rear mimicing the subframe mounted to a frame rail and measured. By my measurements, it looks like the rear mounting point of the FC subframe is ~22 mm lower than the front.

So rather than a straight cut of the frame rail at 26mm, I can cut the frame rail 4mm shallower in the rear stepping it down. This is the direction that I intend to take. Alternatively, I can cut the frame rails straight back for the length of the subframe at 26mm and add a spacer to the rear of the subframe like many others have done.

I'm puzzled by the old measurements of 1/2" and 5/8" so if anyone has insights on how they got to that number, please share.

Metric:
by https://www.flickr.com/photos/26005589@N06/, on Flickr

INches: (Excuse the odd tilt in the ruler. I went back after taking these pics to get more accurate measurements. Hard to hold the camera and hold the stuff to take measurements). It's also difficult to overcome the change in perspective from taking the pic vs the actual measurement. Camera should have been a bit lower in this pic.
by https://www.flickr.com/photos/26005589@N06/, on Flickr

Pic of me using a level held flat to where the front of the subframe bolts up acting as a proxy for the frame rail. I could have used my big *** square rule but this was easier.

 

Yeah I really need to update that old thread, most of that info is very outdated now. I ran the car like that for many years and loved it, but knew I could do better. I came up with similar measurements as you – needing to move the subframe up 1” and around 6 years ago went ahead and gave it a try. The early prototype was rather ugly but worked and proved my measurements were correct. I refined it a bit on one of my other cars. Basically I made a U shaped frame rail out of 1/8” steel which was then grafted into the stock FB frame rails (which were cut to allow it slip up in there that 1”). This worked out pretty well, and one my FB’s is still setup that way – but again it’s not super pretty.

Fast forward to about 2 years ago, and my initial FC subframe swapped car was in need of some rust repair, in the frame rails, and I figured it was time to do it better. I decided the cleanest way to take care of the rust and get things looking tidy was to completely replace the frame rails, from the firewall, all the way up to the radiator support area. I used 2x4, 1/8” wall rectangle tube and by doing pie cuts, formed it to the FC subframe’s contour. I used 1/4 wall tubing to make sleeves and have grade 8 bolts holding the subframe in. After much measuring and such I cut out the entire FB frame rail and grafted these in. I’m extremely happy with how they came out – so much so that I have done a slightly more involved version (I like metal work, if that isn’t apparent lol) on another FB I am in the process of building.

The following pics should show what I’m talking about pretty well – any questions please ask as there is a lot going on. I will try to find my notes with measurements and such as I’ve tried several different (albeit subtle) locations for the subframe for/aft over the years and think I’ve come up with the best setup, at least for me

 

You got mad skillz @82transam !

If you can figure out how you got your measurements for the drop at the rear attachment point of the FC subframe, that would be helpful. I think our measurements are pretty close overall. I used mm since it is easier on my eyes and so I don't have to convert fractions which I can't do quickly.


The next thing I need to figure out is how much further back to place the FC subframe relative to the front-most stud of the FB's subframe attachment point. Many people say about an inch back so that the wheel is mostly centered in the wheel well. Given the purpose of this specific car is to be an ugly race car, I'm not too worried about how centered the wheel is in the wheel well visually. I'd rather place it a bit forward so that I have additional caster - but then I worry about what that will do engine mount placement...

What sort of max caster numbers were you able to achieve with your various approaches?

I recall one thread where someone grafted the upper strut towers from the FC into the FB so that they could maintain the setback measurements between the front of the subframe to the strut towers similar to the FC - presumably to maintain the full sweep of caster available in the FC chassis iirc...

 

This is what I'll be doing in a few months when I've got my new workshop built. The FC swap I did works ok, but it does introduce bumpsteer and when you move the frame 25mm back to centre the wheel, you end up with about 3 degrees of caster. Even less under severe braking.

It looks to me like the easier way to get caster back is to unstitch the strut towers from the inner guards and move them back about 25mm, and add new steel to fill in the gaps. Quite a lot of work though, so if anyone can confirm they've done this and it's improved the caster sufficiently I'd be interested to hear about it. I don't think modifying the ball joints is going to be likely on a street car subject to regular safety checks.

82transam, for those of us with solid frame rails in the front, is there a solution that involves keeping original rails and lifting the subframe up 25mm? That champcar link posted earlier is pretty thin on pictures. I can see how they've chopped it but there's not a lot of description on how they've reinforced the area they've removed.

 

The way it was done on mine was to extend the front of the subframe and redrill new mounting locations. At the rear, boxes can be welded on to correct the higher of the subframe. This way the proper front end geometry is restored. The most important aspect of this design is that the subframe uses the factory mounting points, without modifying the chassis rails, and is completely reversible. Where I live, this is also important, because modifying the chassis rails would make the car very difficult to compliance for driving on public roads. Not such a big issue when it's a race car though!

Caster is not an issue as this can satisfactorily adjusted with the ACBRON Motorsports coilovers I have. It is also a simple and elegant way of doing it.

Here's the general idea....

 

Yeah that's more or less exactly what I've done, but it doesn't restore proper front end geometry because your roll centre is too high when you place the FC subframe under the FB rails. The FC rails are a lot higher. Obviously this is an issue for you if you want to keep a street legal car, so you kind of have to live with it. If you've lowered it at all, you must be experiencing some kind of bumpsteer, maybe it's not so noticeable with narrower front rims/tyres but it's definitely there for me. Can you get under the front to take a photo when the car is resting on the ground and see the angle of the lower arms?

I might be able to add a bit more caster by positioning the BC Gold coilovers 45 degrees around and slanting the top hats back and inwards. Will have to investigate that as they don't have specific caster adjustment. I doubt I would get more than a degree or so doing it that way though, it would be nice to move the towers and aim for 6-7 degrees of caster.

 

Hi All, sorry for the delayed response, was away for a few days. Appreciate the compliments, it came out pretty well for using nothing more than a 4.5” angle grinder and an old mig welder for most of the work. I do not have any more measurements off the top of my head, but I can try to get some. I forget exactly what I got for caster in the various revisions, but camber plates with adjustability, or additional caster built in are a good idea. As you guys suggest – the best approach would probably be to use FC strut towers, or move/modify the FB ones but that was more than I was willing to do at that point.

The way Kypreo has his setup is how mine was for many years. Not ideal for roll center but it does work especially if you don’t have the car super low and/or have laws to abide by – I had probably 20k miles on that style setup and it drove quite nicely.

As for a solution that doesn’t involve invasive surgery like I did, you would probably want to chop up the mounting surface of the subframe so that it moves up farther, essentially lowering the tabs on the subframe. I had consider this route, but it is no small task and keeping everything straight was making me second guess how accurate I could be.

The pics I posted above are how that car sits now and I copied the setup on another 83 I am building. I believe this is as far as I’m willing to go with the FC subframe setup. It works very well and parts are still somewhat available. I will inevitably get the bug to change it eventually, but at that time I will likely move to something else entirely - Miata, Rx-8, full custom tube frame perhaps, not sure!

 

An aside... I did the 1" forward install so I could use as many original mounting points as possible, and used an FB power steering column. I made a steering shaft by extending the FC steering shaft.
I discovered that an NA6 steering shaft is almost exactly the correct length. The NA6 column is also very similar to Series 3 but different in a few ways that I've forgotten.
​

 

My car is in the workshop and I can't visit it (due to being in lockdowns). When I can go back I'll be sure to take some photos. I believe the angle of the lower arms has been addressed through my custom coilover bottom sections or anti-bump steer spacers, but I need to check.

I also told front vs rear roll centres is not really an issue either as I have a custom engineered panhard rod setup.

I will say that the shop engineering my setup knows what they are doing and have been designing suspension setups for IPRA racing in Australia for many years.

 

If you have a low panhard, it'd probably work well.

After driving my unmolested '81 around, its handling is sooooo much better, even after dropping the rear roll center 40mm when I built my 9" for the FC swapped car (it is at axle centerline now), I'm seriously considering binning the FC subframe.

Being a Pete project, I want to put a narrowed NA steering rack inside an FB crossmember, use Mazda3 (Mazda2?) ball joints to keep using the FC uprights, and move the tension rods to the rear, sort of like what Mk2 Escort guys do to decouple the front stabilizer bar from the suspension.

The thing I think is happening with the FC suspension is that the arms are too short, so the roll center is not just in a poor place, but it moves around a lot too.

I also want to disagree that the rack is more precise. My '81 has much more precise steering than the FC swapped car. There is a lot more involved than just the mechanical linkage between steering wheel and the uprights.

 

Its been quite some time since I've been active on here, but this is something I've been very slowly working on with my car as well. I made some new holes for the forward frame mount in the chassis so I could have the wheels to sit centered in the wells, as for the rear mounts I made up some some brackets from some 1/4" plate and re-used the factory steering box and pitman arm holes for bracket mounting, w/ 2 additional studs w/ spacers down through the frame rail. For struts, I had BC Racing mix and match a hybrid set of their stuff, building a coil over for me using the FB upper camber plate with their FC lower strut for a quick install. I'm still working on the steering shaft part to mate up the original shaft to the S4 manual rack, but things slow down when a mortgage and baby are involved

 

It has been a while since there have been posts on this topic but I wanted to revisit the concept of roll centre effects when installing an FC subframe into a 1st gen RX7.

I have been working a bit on my front end as part of my conversion and I have some constraints because my car will be street registered and needs to be engineered to comply with my local regulations. This means no cutting chassis rails or anything like that.

I think the preamble in the original post did a good job of summarising the issues.
* The FC front roll centre is already low. The short arms cause issues when lowering the car just in the FC.
* Compounding the problem is that the subframe has to sit around 20-25mm lower in a 1st gen than an FC by virtue of the lower chassis rails. If the car is lowered, the problem is compounded again. This is inevitable since most 1st gens running a shorter tyre height than the average FC.
* So, all things being equal, when using a FC subframe in 1st gen RX7, the roll center will be affected in the same way as if you were running a lowered FC, but 1" lower than that again.
* There is an additional issue in that the front vs rear roll centre differential is much greater in a 1st gen due to the factory rear roll centre being quite high.

That last problem I think it actually not so great a problem for most people, and there are fixes to at least partially address it. In my case, I have a custom adjustable panhard rod that sits quite low and moves the rear roll centre much lower than the factory position - see below:



In a traditional setup the rear roll centre is essentially where the watts link pivots on the diff housing. As I understand it, with a panhard rod, the roll centre is effectively the lower diff mounting point assuming the panhard rod is kept horizontal. In the above photo, because the rear end is hanging the panhard rod is angled up, but on ground level it becomes horizontal. This point is considerably lower than the factory watts link pivot, lowering the roll centre substantially. As you can see, I can move adjust the rod position at both ends, making rear roll centre highly tunable.

i think peejay mentioning in other posts that he tried similar setups and they broke. I think mine should be more than strong enough for a street car. You'll see the mounts are very beefy and have an upper rod welded to the chassis rail to provide lateral reinforcement.

In summary, problem in the rear is solved.

Now to the front, I think there is a much more elegant solution than anyone else has raised in the 1st gen context, and it's something that FC owners have known about and have been doing for a while, which is to:

(a) lower the ball joint attachment point relative to the lower control arm pivot points. You cannot easily raise the lower control arm mount points on the subframe. The front arm could be redrilled, but the rear point would require some fairly substantial modification and I've never seen it done effectively on an FC. The better solution to correct the lower control arm angle and restore it closer to the geometry at factory FC ride height is to attack the problem from the other end, but lowering the attachment point to the knuckle. There is a limit to how low the ball joint can move it based on brake and wheel configuration (especially if running smaller diameter wheels), but you can lower the increase control arm angle and thereby raise the roll centre using this method. A number of people have used spacers or extensions that sit between the ball joint and the knuckle attachment point. Or you can modify the knuckle itself (which I'll get to below).

(b) move the tie rod attachment lower to suit. The problem with changing the lower control arm angle is that the tie rods will then sit at a different relative angle, introducing bump steer. You can also address this. There are plenty of aftermarket tie rod ends that feature some kind of bump steer correction to have the tie end pivot sit lower. Or you can modify the mounting point on the knuckle itself.

An elegant way to combat both problems is at the knuckle. I prefer this method because I hate the idea of a ball joint spacer under the knuckle (from a safety perspective). I also don't want aftermarket heim joint style tie rod ends because of NVH, longevity and legality reasons (ball joints with dust seals last longer and will be easier for me to engineer for a street car).

And there is an off the shelf solution. Forged replacement knuckles from Japanese tuning house Parts Shop Max shown below:



As you can see, both the ball joint attachment point AND the tie rod attachment point sit much lower, relative to the axle/spindle, compared to the factory FC. In fact they sit 45mm lower. This provides 45mm of simultaneous roll centre and bump steer correction. In the context of a FC subframe in a mildly lowered 1st gen RX7, this 45mm difference effectively restores the LCA/tie rod geometry to what it would be in a factory FC. That is, you make up the 25mm lower position of the subframe when it is installed in an unmodified 1st gen chassis, then another 20mm on top of that to partially account for lowering. Even accounting for a 1st gen sitting on shorter tyre height, it will restore the geometry to close to what it would be like in a mildly lowered FC, which although not the best in the world, is going to be more than acceptable for most people and overall probably way better than a factory 1st gen, factoring in the vastly improved steering.

The added advantage of the Parts Shop Max knuckle is that the tie rod pickup point is also altered slightly to improve steering angle, steering response and optimised ackerman, to address other common complaints about the factory FC setup.

I am yet to do this and install in my car, but I've seen these installed in FCs and they work. You can modify a knuckle in the same way, but you want to make sure you know what you're doing. These ones aren't that expensive and are forged steel.

It is not the perfect solution, but I think it has the potential to at least partially address most of the perceived issues with the FC subframe setup in a straightforward and elegant way. Keen for people's thoughts on this.

 

I should add to the above, the caster issue has also come up in this topic when installing the subframe forward to correct wheel position relative to the strut tower.

I haven't yet gotten an alignment so I can't tell you what caster to see what it comes out in my car.

Contrary to what I thought, I do not have caster adjustment on the strut tops but there is a bit of extra caster built in by virtue of the offset of the strut mount in the plate.

If you stick to the factory lower control arms, it is also possible to gain caster by using a rear bushing with offset mounting. This pushes to the rear of the LCA forward, adding an extra 1 degree of positive caster. SuperPro makes these, part number SPF1743AK.

Further caster gains are also possible with aftermarket lower control arms or other knuckle modifications. To my mind, this is a far superior solution to modifying strut towers.

That said, from all accounts and my research, you don't need much more than 4.5 degrees in an FC due to the nature of its camber curve - it seems extra caster will change the feel but necessarily make the car any faster. The offset bushes and my strut top design should hopefully be more than enough for me to restore close to factory +4.5 degree caster, if not more.