110 replies to this topic

[hydra]

Hey all,
I've got this project I'm working on where I want to install a Subaru EJ2x series engine in an old Beetle. So far so good, its been done successfully many times before and there aren't any real difficulties with the actual engine swap. Here's where I intend to take the project a little further. I would like to install a Subaru transmission as well instead of the commonly used built up VW 5-speed or Porsche 5 speed alternatives. I'm not going to get into the why's, as this isn't the place for it. In any case, installing the said tranny presents the problem of ending up with 5 reverse gears and 1 forward gear, and is clearly unacceptable. Flipping the ring and pinion gears is also not possible, due to the design of the tranny housing. Thus I am left with two options, inverting the tranny and fabbing my own spacer/adapter kit which I'm not all that keen on doing, or I could invert the rotation of the engine!

To do so, I would need to make different cams, and invert the polarity of the hyperdrive flux converter - more commonly known as the starter (sorry, that was a Star Wars reference I couldn't resist ) There are also a couple of misc items like the toothed gear for the crank sensor and cam angle sensors. I guess I should add that I will be running a Mega-Squirt II setup, so I can alter any parameters at will should I need to. Is there anything else I haven't thought of? The advantage of having to make different cams is that I can make them more performance oriented than the stock cams (whereas a tranny spacer wouldn't add any performance by itself) which is a plus I guess. What I'd like to know is if there are any deleterious long-term effects to running the engine in reverse... For instance, the oil holes in the crank are optimally "timed", running the engine in reverse would change all of that. Same thing goes if there is any piston wristpin offset, although that can easily be overcome by installing the pistons in the opposite direction. What about water/oil pumps? I'd like some more thoughts on this, and please please please, I don't want anybody to try and talk me out of the idea, I just want valid engineering arguments as to why I SHOULDN'T go through with this. Let us assume that I've got free access to CNC equipment and 3D drafting programs, and that any initial costs could be amortized over several units, and so aren't THAT big a deal...

[soubriquet]

How about the cam drive belt? Will you be working the belt directly against the tensioner rather than having at the end of the line to take up the slack? I'll look at my books tomorrow;

[hydra]

I don't think that would be a problem, but then again I'm not sure... Can anybody care to comment?

[hydra]

Actually, supposing belt slack IS a problem. Couldn't one just extend the tensioner so much that all the slack is taken up and becomes tension?

[soubriquet]

It's nothing to do with belt slack. The tensioner's job is to tension the belt, not to transmit power to the cams. Placing it first in line is Russian roulette.

[hydra]

Yes, I know that. Maybe I wasn't too clear with what I meant... What I meant is that you could adjust the stock tensioners to take up the additional slack. In any case it doesn't look like that's much of a problem anyway with the Subaru's DOHC drive setup. I'll try to attach a pic...

[sblick]

Don't you also have the problem of the water pump vanes not spinning in the right direction also is the oil pump gears going to have problems. Is it an electric fan? If the water pump is spinning backwards so is the fan. I thought that all the main bearing caps have the hole drilled in them to spit out oil in the direction the piston is turning.

[sblick]

Now I see that this is probably an air cooled engine sorry about the water pump thing

[bobdar]

Rotation reversal was a common deal for circle track racers, you might take to someone with that experience. Offset-pin pistons (if so equipped) should not be a problem, since the street car would have the offset on the thrust side of the piston ("noise offset"). In reverse rotation, the offset would be in the "power offset" mode. Small block Chevy tuners would swap pistons across the banks to achieve this small power advantage.

[JwS]

The timing belt tensioner on the SOHC engines is oil fed and hydraulicly locks, continued use under load may tend to push it back anyway though, which could be a problem. Also using the tensioner side of the belt (normally the 'loose' side) as the driving side of the belt will allow the cam timing to vary depending on the tensioner position. At the very least this will make it tough to set the cam timing right
I think alot of the engines that are reversed use a gear or short chain drive, american v6s and v8s and such, which don't really care which way they turn, or in some cases the drive can be reversed by using an aftermarket gear set.
Piston pin offset shouldn't be a problem I think (if it has any), the water pump may be a pain, and the oil pump is located directly on the nose of the crank, so that may cause some headache too.
If you are going to the trouble to transplant the engine/tranny why not put it ahead of the axle like a Porsche 550?
It is a decent little engine though, I have a 99 impreza, reasonable power, pretty smooth, it has been pretty reliable for the last 130K or so. The oil pump seal failed (lousy seal material) but other than that nothing. Mine burns a little oil, but it has done it since I got it so I ignore it.

JwS

[hydra]

JwS,
The DOHC engine has 2 roller pulleys per bank, in additional to the said tensioner. so I don't think that belt slack will be that much of a problem. In any case, I can always fab up some sort of supplemental tensioner setup to minimize the problem.

Regarding the water pump, does anybody has any solutions short of casting a mirror-image impeller or using an electric water pump? What about the oil pump? I was thinking I could reroute the internal passageways or maybe even casting/fabbing a new oil pump housing.

Going with a mid-engined setup is naturally the easy way out, but its not an option for anybody wishing to keep their Bug practical (i.e. seats 4) and stock-looking, so that is NOT AN OPTION for the purposes of this exercise...

And yeah, the EJ2x series of engines are very well designed units...

[hydra]

Another point I'd like to bring up,
Would the tranny care at all if thrust loading is in the opposite direction?
What about the crankshaft oil-hole "timing"?

[McGuire]

I think the easiest way would be to just turn the driver's seat around.

[Ray Bell]

I was going to do this once with a Peugeot 304 engine to fit it into a Clubman...

They run the reverse direction to normal, so I was 'normalising' it. But I was going to have all sorts of things drive off the timing chain, not retaining much of the original at all. Water pump, oil pumps, all of that stuff was going to be changed.

An interesting mental exercise... best advice I can offer is 'don't do it!'

[hydra]

Oh nonsense...
Come on guys, think positive! When there's a will there's a way! You might not think its worth it, but believe me when I tell you that it is!

[McGuire]

How about this: construct a pair of slimline gearboxes to mount on each side of the transaxle at the output flanges. Everything then runs in the original direction except the half-axles. You could also fiddle the gear ratios to under/overdrive the final ratio as required. And maybe you could use the geartrain offset to move the entire drivetrain mass forward in the chassis a bit...

[Ray Bell]

Hey! I was thinking something like that, but at the wheel end... something like the old VW Pickup reduction gears...

[McGuire]

Originally posted by Ray Bell
Hey! I was thinking something like that, but at the wheel end... something like the old VW Pickup reduction gears...

That was exactly my train of thought too...then I figured why not put the gears on the unsprung side of the halfshafts.

[ciaoduc1]

You could probably fix the water pump problem by re-routing the serpentine belt so the back side of the belt turns it "backward" and using a smooth pulley on the pump. And since you're fiddling with the belts, add another pulley to the front of the crank to run an external oil pump.

[soubriquet]

The oil pump is directly driven off the front of the crankshaft, so it will have to run backwards. You will either have to swap the input/output plumbing, or do something cunning like re-jigging the outer rotor so that oil is pumped across the top, rather than the bottom of the pump.

How about fitting the engine into a Subaru; that way you'd have a really good car ?

Don't some Honda engines run in "reverse"? Transverse fwd with the gearbox on the left when you pop the bonnet and look in? Why would Honda do this?

[Ray Bell]

Yeah, you're right... the Honda Coupe 7 and Coupe 9 engine ran backwards...

It was an inline 4-cylinder air cooled engine with a big fan arrangement. No idea why they did it, though.

Maybe the VW would be better off with an engine you don't have to modify... maybe a Tatra?

[bobqzzi]

Goodness, what a project. Having been inside all 3 trannys, I can't imagine why you'd want to use the Subaru one- it seems plainly inferior.

[Wuzak]

I think most Honda fwd engines run backwards.

A former boss of mine had an S800 fitted with a Nissan motor. He ran it in Targa Tassie, but updated rules required a Honda motor. And the available Honda motors (in the correct years) ran reverse to the Nissan.

[Ray Bell]

Actually, the S800 (and S360, S500, S600... well, you get the idea...) engines all ran backwards. To stick the Nissan engine in would have required a reversal of drive.

I didn't think the Civic ran backwards, but I don't really think I ever looked at it...

[Wuzak]

It had a Nissan gearbox (IIRC) and a Holden diff. So the rotation was correct.

Don't ask me why it was thus.

[Ray Bell]

Holden diff? Probably a Torana housing...

The one I remember like that had a Cortina engine and some other diff, I don't remember which. They were a mighty narrow car.

[soubriquet]

Honda build an 1800cc flat six. Wonder if that runs in reverse? VW-Gold Wing anyone?

[Ray Bell]

Now that's more sensible!

[sblick]

Honda was the last manufacturer to switch to "normal" rotation. I think they did this in the late 80's or early 90's. They always said it was a performance advantage. If I remember right the GP engines also ran "backwards".

[hydra]

I like the reduction gearbox idea, I must say, I hadn't thought of that! I still prefer making the engine rotate in the opposite direction as it seems lighter, easier to package, and very possibly cheaper - not to mention its potential for added performance (bigger cams)

If the cams lobes are symmetrical, I can even make custom or adjustable cam sprockets to advance the cams by 360 minus lobe centerline ... Simpler, easier, and cheaper than making cams.

[Ray Bell]

Originally posted by sblick
Honda... always said it was a performance advantage.....

Depends on which side of the Equator you are...

[AS110]

Mitsubishi changed sides in the mid 90s,the engines used to be on the left,then they went over to the right hand side.Maybe it was the change in trans to the Tiptronic.Put an early 90s Diamante nose to nose with a 95 and the engines are next to each other,they look identical,just everything resersed in mirror image - but of course the motor still turns the same way.

[Kevin Thomas]

I saw a program once, in the US it was called Junkyard Wars. It was a british program. They had a drag race competition. The rear differential was wrong for the engine and transmission. They fixed it by spinning the differential 180 degrees around along the drive shaft axis.

Don't know if this would help here.

[ray b]

very common in paired boat engines
one is reversed to cancil tork reactions

most are pushrod V8s but volvo 4 bangers allso do it
and BMW i6s too
and a lot of diesels tooo

but why mess with a watercooled motor
when so much hotrod stuff is made for the stock bugs

[hydra]

Originally posted by ray b


but why mess with a watercooled motor
when so much hotrod stuff is made for the stock bugs

Hmm... Lemme see... I can either mess with a modern forced induction quad cam, EFI 4 valve engine, or I can go with an antiquated, leaky, and relatively unreliable pushrod 2 valve engine with carbs
Seems like an easy choice to me, and its not like there isn't a lot of hotrod stuff for the Suby engine either... Besides, I don't want to bother with the WHYs so much as the HOWs
I know very well I can stick with the air-cooled engine and soup it up, I just don't want to....

Its like you said - Question, wonder, and be weird!;)

[275 GTB-4]

This is all too easy......just start with a Commer Knocker....

[Ray Bell]

Originally posted by 275 GTB-4
This is all too easy......just start with a Commer Knocker....

Who's got that .gif file?

[AS110]

A guy who works for me has a Subaru in his Imp,he had to use the old pushrod motor because of width problems and it's mated to the Imp transaxle.I guess he would of had the same problems if he used the Subaru trans too,but the Imp always had a good trans.

[Catalina Park]

Originally posted by Ray Bell


Who's got that .gif file?

[Catalina Park]

Going back to the original problem of how to turn around a Subaru motor.....

I have come up with a solution that does not involve reversing the rotation of the crank.... Hydra says that he can't flip the gearbox due to

Flipping the ring and pinion gears is also not possible, due to the design of the tranny housing. Thus I am left with two options, inverting the tranny and fabbing my own spacer/adapter kit which I'm not all that keen on doing, or I could invert the rotation of the engine!

Why not flip the whole engine and gearbox! It must be easier to get the motor to run upsidedown than to reverse the rotation of the crank.

[hydra]

Sure, but then I'll have to include an anti-gravity generator to make sure the engine oils properly

As for difficulty of making the engine turn the other way around, I still don't see why its THAT hard?

[Ray Bell]

Originally posted by hydra
Sure, but then I'll have to include an anti-gravity generator to make sure the engine oils properly.....

It's called dry sumping...

.....As for difficulty of making the engine turn the other way around, I still don't see why its THAT hard?

We can tell.

[desmo]

Catalina Park, good job on saving that Double Knocker animation I thought it was gone forever.

[JwS]

Of course the option not yet mentioned is to turn the car over, the only problem there is that you'll need to build new inverted roadways to drive on...
JwS

[Ray Bell]

Originally posted by desmo
Catalina Park, good job on saving that Double Knocker animation I thought it was gone forever.

I don't think it's the same one...

I saved it too, from when you posted it before, but that hard drive is the one that's subject of a lot of discussion (and search for a controller card...). I feel sure that the one the Mountain Man has posted is running faster.

[Catalina Park]

Originally posted by hydra
Sure, but then I'll have to include an anti-gravity generator to make sure the engine oils properly

As for difficulty of making the engine turn the other way around, I still don't see why its THAT hard?

Lets see, to turn the motor upside-down...

1) Move oil pick-up.

To reverse rotation.

1) Make new cams
2) Make new oil pump
3) Make new water pump
4) Move tensioners on cam belt

The Subaru gearbox is not that good anyway. I would ditch it and go the VW box.

BTW. I have reversed the rotation on diesel engines.......

[hydra]

The Subaru tranny only has a bad rep because of all the high-rpm clutch drops and 4-wheel burnouts a lot of Subaru owners seem to enjoy doing. I don't think I will have a problem with a 2WD Legacy gearbox in a lightweight 2WD application with street tires. Besides, the VW Type I gearbox is pretty weak in stock form, and is only a 4 speed, and a 091 Bus tranny has REALLY short ratios, so the VW alternative isn't nearly as favorable as you make it sound. Porsche gearboxes are too expensive to buy and rebuild, and I would have to cut the rear torsion tube to make them fit.

The engine inversion idea is starting to grow on me, as it would allow me to comfortably package a turbo charger (I would have otherwise gone with an Eaton supercharger) in the tight confines of the VW's engine bay. My question is, can it really be done? I'm guessing I would need a complete dry sump system with POWERFUL scavenging... What about windage issues? After all, the engine wasn't designed to operate upside down...

[Ray Bell]

I think that turning it upside down would be, in reality, very difficult...

The inlets would then be under the heads... that will create problems. The exhausts might not be too much of an issue on top, but they'll create heat in the engine bay that wouldn't be there normally.

Then there's the oil situation. Scavenging (it would have to be dry-sumped) from what's now the top of the engine might well be a bit difficult to organise, but you might be able to use the windage to help. The fact that this is a one-off and there'll be a lot of experimentation to achieve the goal makes it just too hard, menthinks.

[Catalina Park]

Remember, it is a Subaru motor. It is not like the oil is going to run out the rocker cover if you tip it over! The big problem that I see it at oil change time when you have to roll the car on its roof to change the oil!

In Australia Subaru has problems with all gearboxes, not just in WRXs. They try to avoid warranty claims by calling it driver abuse. This is slowly backfiring on them with when it comes to repeat sales.

Do you want to know how I reversed the rotation of a diesel engine?

[soubriquet]

Do you want to know how I reversed the rotation of a diesel engine? [/B]

No, but I'd guess it was a 2 stroke.