Low Engine C of G / Inverted Gearboxes

Not another gearbox question?! A bit long, but quite detailed. Please read carefully to make sure you understand what I'm asking, before posting replies, as I am NOT asking about ratios or torque capacities.
Having had a good look round the paddock at Goodwood on Monday, I decided that there was a lot of fresh air under many of the engines which I would like to avoid on mine, so I am looking at getting the engine/ box as low as possible. However I do not think I want to go to the extravagance of dry sumping. On that basis I am looking for combinations which will allow a wet sump engine to be placed as low as possible. Would I be correct in thinking that the shallowest wet sump for a 351 W will be about 6.5 inches deep? (Armando)(Also use Accusump) So therefore the problem is how to get the block low enough that the bottom of the block is 6.5 inches above the bottom of the chassis, and horizontal if possible.
Fitting a Renault 21 or Audi 016 the normal way up with this configuration will result in the gearbox probably dragging along the ground and the driveshafts trying to drive through an angle of about 30 degrees, so not particularly helpful. alternatively the assembly could be angled using a pan with a sump in its rear to reduce the angle on the driveshafts, leading to the distibutor poking through the bulkhead, and air traps in the water system.
My current understanding of the main gearboxes is the following.

ZF. The output shafts are ABOVE the input shaft (best combination, but most expensive)
Audi 016 and 01E, and Renault 21 and 25. The output shafts are BELOW the input shaft. This leads to the problem detailed above.
Porsche can be fitted either way up, but again much more expensive than Audi or Renault.

My question then is what is the best way to fit the engine as low as possible whilst still using Audi or Renault? Can they be inverted? If so who has actually done it? What is the best/ lowest compromise in terms of angle of assembly/ angle of driveshafts.

A bit of crude geometry might help reduce your concerns. You mentioned that you want to get the bottom of the block 6.5" from the bottom of the chassis. I assume the centreline of the crank is on the plane of the base of the block.
If I also assume that you plan to run about 5" ground clearance, that means your crank C.L. is about 11.5 inches above ground level. Assuming you are running about 26" dia tyres, this means that the centreline of your rear hubs will be about 13" above ground level (any neg camber will reduce that)

Your half shafts should be about 12" long and angling them down 7 degs should be no problem at all (look at the angle on some formula cars). This means that the inboard hubs can be about 11.5" above ground level.

The output shafts on a Renualt 30 g'box are about 1.5" below input CL or (11.5 -1.5) = 10 inches above ground level.
By canting the engine and g'box up at the back about 2.5 degs, the g'box o/p will be aligned with the halfshaft.
2.5 degs is very slight and if you could lower your sump depth to about 5" you could mount your engine and g'box completely flat.

My installation is dead flat using a R30 g'box.

Hope this helps a bit.


GT40s Supporter
Hi Lee

Dropping the engine is on my wish list so I have also been looking at other people's solutions over the years. I can only make a few observations.

None that I have seen or heard about have failed through drive shaft angle issues. If you take Roy Smarts car as an example, he has now had a lowered engine and gearbox for some years. With his high horsepower engine (500 hp plus) and monster torque, the drive shaft angle hasn't caused any issues I am aware of. And with his running the quaife 1st 2nd gear internals and LSD, the weak links in that drivetrain are the cv joints and drive shafts.

More of the UK competition guys have lowered engines than not. None have inverted the Renault box. However Kevin Farrington has. He is the only one I know of to have done this. He has not had any problems either but in his article in the GTD club mag an issue or two ago, he wrote up what was involved. It didn't sound like a five minute job!!!

Keeping the box the right way up does not mean it has to drag along the ground. Andrew Fordyce has recently had this work carried out on his car and uses a neat chassis solution to the problem. But still with a non inverted box.

On the sump side, I use a Canton racing sump which is about 7 inches deep. This is a saddle sump design so I can lower the engine more than enough without need for a new sump. The reason why I haven't done this job is firstly I can't weld very well so chassis reconstruction puts me off slightly but mainly because I would need a new set of exhausts which is big money.

You are not the first person to think of drive shaft angles, but IMHO it isn't something to have sleepless nights over as it isn't actually a concern in practice. But that is IMHO!

Hope this helps


I've also been doing alot of research on this! Here's what I've learnt so far! please bear in mind this is only research & I haven't quite got round to cutting up my chassis!!

Inverting the gearbox is a very attractive idea, as it measures up like it was made to fit!

As Malcolm sias, there is only one person in the club that has done this.

As standard (with the R21T) the crank C/L is 70mm higher than the driveshaft output C/L. If you invert then you could retain the same driveshaft angles & lower the engine 140mm.

With Armandos GT40 sump at 165mm deep, you'll find the bottom of the sump would then be flush with the botttom of the chassis. Perfect!

FWIW inverting the trans was a common mod in Hillman Imps & there are a few about. I'm told some older forumla Renault single seaters used an inverted UN1, but I've never seen one.

Ultima use the Porsche G50/60 'box and or the 996 (getrag) 'box, both inverted. There is a breather mod they use but other than that I'm not sure exactly if there are any internal mods.

If you have even deeper pockets a Pantera ZF can be inverted by Roy Butfoy in CA.

I've thought long & hard about inverting the R21T, but do not have the confidence with the oil control, especially as I'm using DB's Quaiffe upgrade (V expensive!). Overall I'd have concerns as to how the lifespan of the trans may be affected by inverting. It may last for ever, or it may pop in 5 minutes.... I don't know!! An external transmission oil pump and/or cooler may help, but it's extra plumbing, complexity, weight, cost, etc.

So if you don't want to invert, the trade off is more extreeme driveshaft angles.

Basically, I've learnt that driveshafts can do 'high angle, low speed', or 'low angle & high speed', but not both!

The 'Off-road' 4X4 chaps use outrageous driveshaft angles when they're driving up & over mountains, etc. However it's all at relativeley low speed. In our application wheel speeds are much higher.

I found a company that can supply CV joints that (they say) will happily cope with 20 deg & more. These are pretty pricey, but are used in World chapmionship rally cars, where they need alot of suspension travel & high wheel speeds, combined with bucketloads of BHP.

So using some basic geo (as Trevor suggests) & a bit of trig, I calculated that allowing for full bump (75mm) putting the driveshaft at 20 deg, the crank C/L could be placed 195mm from the chassis underside, with the engine & trans dead level. I guess if you're happy with a bit less suspension travel, or a bit more driveshaft angle you could go lower. Also I guess it's V unlikeley you'll ever be at full droop whilst on full throttle!

As I mentioned, I haven't put any of this into practice yet. I'm getting the hacksaw out soon!!

Hope this helps!
OK. Now I've started on the Trig. The attached diagram shows the problem. The vertical dimensions are accurate, the horizontal ones only approximate but near enough.

What would happen if the assembly with a 165mm sump was installed parallel to the chassis with clearance of 10mm between bottom of sump and botom of chassis rail, and assuming ground clearance below bottom of chassis rail of 125mm using Renault box and 26 inch (660mm) tyres.

The result is a difference in static height between the CV joints of 100mm (4 inches). If we were to run with ground clearance of 100 mm (4 inches) them the difference would be 125mm (5 inches). So ground clearance plus CV difference = 225mm (9 inches)

A 4 inch difference results in a static angle of about 20 degrees, 5 inch about 25 degrees. OUCH!!!

This difference can be reduced by tilting the assembly. Keeping the centre of the engine at the same height (by using a rear pan in the sump) the difference in height of the CV joints can be reduced by approx 8mm per degree of tilt. So with ground clearance of 5 inches, the difference between CVs can be brought down to 2 inches by tilting the assembly by about 6 degrees. If running with 4 inches of ground clearance, this would require about 9 degrees.

Please tell me I've got it wrong. Is this problem the same with the Audi box? What is the dimension between Input and Output shafts


HI Lee,

Nice drawing, thanks for taking the time to post it!

I've used almost identical dimensions for my sums! As you say the driveshaft angles are a bit steep, but I came up with a slightly different angle, here's my sums, please tell me what you think!!

What length are you using for the driveshafts? I've measured at roughly 400mm. From the diagram, one end of the driveshaft is 100mm higher than the other.

Using a right angle triangle, knowing one side is 100 high, the other 400 long & we don't know the length of the longest side.

From my trig this gives a driveshaft angle of 14.47 deg. (while at normal ride height)

If you go to full bump, (which I guess at 75mm) then the triangle become 175 high, 400 long, giving a driveshaft angle of 25.94 deg

it's probably worth mentioning that during the change from normal ride height to full bump, the drivesghaft length changes from 387mm to 359mm so a 'plunge' of 28mm would be required.

From the manufacturer I spoke to, this is not a problem with each CV having 16mm plunge.

I have'nt done the same sums with an inverted 'box, but I'd guess the angles would be much better, probably close to perfect! If you're starting from scratch I reckon it would be well worth looking at the Ultima setup.

Rick Merz

Lifetime Supporter
I modified my GTD frame to run a 351W with a Porsche G50-52 with a Quaif and my half shaft angles are less than 5° and the bottom of the oil pan is 4" from the ground. The only problem I have had is breaking half shafts and CV joints (due to excessive HP and torque), but Phat performance in California sold me CV joints and axles that can take almost anything that you can put to them. I also had my toolmakers at work make up adapters (the bigger CV joints have (6) 10mm bolts on a larger bolt circle, same as the Porsche out put flange), which will bolt the bigger CV joints to the stub axles. I was then shearing the (6) 8mm bolts in the adapters so I had to have the adapters indicated to the stub axles and welded, since the adapters were welded I haven't had any driveline problems since. BTW I have made many different 1 off components for my GTD and I can supply prints and 3D renderings for anything that I have modified or made for my car. I have a lot of experience with AutoCAD and AutoDesk Inventor 6 and I would be happy to help anyone with their project.



That's a really nice installation, superb... great work!

I'm assuming that to fit a G50, the trans *has* to be inverted, as in the porsche the engine sits behind the trans.

Is this correct & if so, I guess the crank centerline must be below the driveshaft output centerline. Do you know by how much?

I assume the ally pot with the 6 oil lines is some form of transmission oil cooler/pump?

As Julian said, nice drawing. Sorry, I didn't get back to you sooner, I've spent the last 3 days tearing down and rebuilding a friend's diesel injection pump; (lots of fun).

The differences between our two set-ups puzzled me so I looked a bit closer into the details. It might be worth checking one of the assumptions I made in my previous post; ie that the centreline of the crank is on the plane of the base of the block.

I went back and checked this and found that the crank CL is in fact about 1.5" above the top of the sump pan. I am not using a Ford block, so yours could be different. To me, this explains the differences, it also means that the R30 and R21 gearboxes have the same vertical offsets between input and output shafts (70mm).
Sorry if I mislead you a bit, but without having a Ford block in front of me, I had to make an assumption. The good news is that if the Ford block is similar, you will wind up with similar halfshaft angles.

Good Luck and sorry about the confusion if I caused any.


On my 302, the crank centerline is indeed on the same plane as the bottom of the block.

Good luck with the pump!
I hope these pictures are of interest. I've made a jig for lowering the engine/tansaxle. Ride height is 135mm & the crank centerline is 165mm higher than the chassis floor.

This picture shows the driveshaft angle (roughly) at zero bump.


Hmmm....puzzling. Just putting a protractor to the screen looks like your estimate of 14 degs at nominal ride height is about right.
I assume the centre line of the inboard cv was placed 70 mm below CL of crank. Those angles do look a bit marginal the way you show them.
As mentioned before, if you tipped the rear of the engine up a couple of degrees (around the motor mounts) you could buy a bit back. I don't think this would effect the lubrication efficiency of the engine.
Hi Trevor,

Yes, the inboard CV is placed 70mm below the crank C/L, & as you say the engine is level at the moment but could benefit from a tilt.

It's worth mentioning that with the engine placed as in the picture, a great deal of chassis modification is required to get the Trans to fit.

Either the longitudinal chassis members that pickup the lower A-arms would have to be changed from box section to angle, or 'spread' the box section by 40mm each and shorten the A-arms accordingly.

The rear section of the chassis would also need to be modified, but as you know 'only' carries the trans mounts & the rear deck mounts.

Or, a third alternative is to set the crank centerline at 205mm & keep the A-arms as original.

Do you have an opinion as to the effect of shortening the A-arms by 40mm? Would this have a similar effect to moving the top pickup as per yours & Dereks studies?

Thanks for your input!

Ultima use the Porsche G50/60 'box and or the 996 (getrag) 'box, both inverted. There is a breather mod they use but other than that I'm not sure exactly if there are any internal mods.

[/ QUOTE ]

There are a few things that can be done. First is to install a new input shaft seal. When the transmission oil level is brought up enough to provide adequate oiling, the oil level is above the input shaft. A compromised seal will lead to obvious problems. The other, I just thought of, is to relocate the speed sensor up out of the oil. Not that the oil will attenuate the induced voltage, but that the magnet will be exposed to any suspended ferrous filings. Build up of this would attenuate the magnetic field. Lastly, one can relocate the full plug. However, I just made a dip stick out of a tie strap that is put into the old fill hole and held against the threads at the bottom of the hole to read the oil level. (Oil should be at least to the level of lower shaft in transmission.)


GT40s Supporter
"Do you have an opinion as to the effect of shortening the A-arms by 40mm? "

This is what Andrew has done and by his times this year he appears not to have suffered at all.

With regard to the speedo sensor, why not just block up the gearbox casing hole and use an electronic speedo with a pick up sensor working off the cv joint. Or even use one off the front wheels. This is where I have one for my telemetry sensor. Has the added benefit of being programmable to allow for different radius tyres as you chop and change through the life of the car.

Removing the sender to check for metal filings once in a while,
might not be a bad idea.
Kind of like having a magnetic oil drain plug.
BTW, how many quarts of gear oil are you using?
Great idea with the tie strap! /ubbthreads/images/graemlins/cool.gif

Malcolm, You are right; you can use a different sensor. Being, shall we say, thrifty, I used the existing one. It does go to an adjuster box first which performs two functions: 1) adjust the pulses per mile for the American made computer and speedo (although my Classic Instruments speedo can do the adjustment alone) and 2) acts as a repeater to enhance the signal strength for both as well.

Bill, it has been so long since I did it that my memory is a little fuzzy. I have it written down somewhere, but it seems to me that it was between 3 and 4 quarts of Redline MT90. Good point on checking for filings; just remember to drain a bit of oil out first since the orifice is below the full level /ubbthreads/images/graemlins/blush.gif Some will drill and tap a hole at the intersection of the webbing at the new "full" location, but I didn't think it was worth taking the chance of getting metal shavings in the transaxle.


I have between 3.2 and 3.5 quarts of Castrol 80W90 high performance LSD
gear lube in my G50. I just refilled the gearbox, and I haven't experienced a leak,
so getting a full mark on my new "dipstick" won't be a problem.
Thanks again for the idea! /ubbthreads/images/graemlins/grin.gif

When I designed and modified the rear chassis configurations on Andrew Fordyces car, I allowed a real time movement from static of 3 inches bump and 1 inch droop, these being figures calculated the old fashioned way on a drawing board, as I am not that computer literate to understand CAD systems. As it turns out, with the spring and shock setups that Andrew is running, this is just about right, and therefore the amount of castor/camber change on the rear suspension is negligably different with the lower A arms 40 mm shorter, and in my opinion better than the original. The new chassis members then allow the engine/gearbox unit to drop 75 mm, bringing the Canton 7.5 inch sump level with the bottom chassis rails, with the rear chassis running with 4.5 inch ground clearance. The car handles and performes very well indeed.