2ZZ-GE Swap With No Cuts-- Discussion

[Santa Barbarian]

After considerable delay, and with several False Starts, we're just getting ready to put a 2ZZ into an X-roller. And the question is: Is the native 2ZZ small enough that it can occupy a Bertone engine bay with no cuts to the firewall, trunk-- or relocation of tank and spare??

This began over in a discussion about another car with a K-swap in it, so this is a placeholder for what I hope turns into a serious build.

The subject engine is pictured below. Svelte, slender, light-weight, buff and powerful. How much baggage does he need to bring along?

 

[kmead]

Roobus's K20A2-Powered LBZ on BaT

@Santa Barbarian, have you looked seriously at a 2ZZ? It is a compact engine. It falls a bit short of your desired power, but I have driven a 2ZZ swapped third gen (hairdresser's) MR2. That MR2 is heavier than an X, and I can say that the 2ZZ moved it spiritedly. It might be worth considering if...

xwebforums.com

For some additional reference.

As an aside, this swap is what Steve Hoelscher suggested as a good alternate to the K swap approach. It has been on my radar as a subject for some time but work and other things in life have pushed this back but is high on my list of things to do.

 

[Santa Barbarian]

Roobus's K20A2-Powered LBZ on BaT

@Santa Barbarian, have you looked seriously at a 2ZZ? It is a compact engine. It falls a bit short of your desired power, but I have driven a 2ZZ swapped third gen (hairdresser's) MR2. That MR2 is heavier than an X, and I can say that the 2ZZ moved it spiritedly. It might be worth considering if...

xwebforums.com

For some additional reference.

As an aside, this swap is what Steve Hoelscher suggested as a good alternate to the K swap approach. It has been on my radar as a subject for some time but work and other things in life have pushed this back but is high on my list of things to do.

I'd add right away that the stock Celica-GTS intake is huge, so our thoughts are to attempt this with the inclusion of some not-inconsiderable fab work on the intake and exhaust side. The intake plenum, runners and throttle assemblies are real deal-killers here.

On the other hand, Toyota (and Yamaha) also threw us a few bones along the way. The water pump housing assembly is nicely placed. And the injectors (for anyone staying fuel-injected) are integrated into the heads. They also use nice, compact coil packs in a Coil-on-plug set-up.

 

[kmead]

https://danstengineering.co.uk/index.php?route=product/product&product_id=1290

Perhaps a few clues here though it would be better to go with an IDF pattern ITB if going down the ITB path. A plenum which is up high and directly above the intake ports versus offset to the drivers side would likely do well.

 

[Santa Barbarian]

Perhaps a few clues here though it would be better to go with an IDF pattern ITB if going down the ITB path. A plenum which is up high and directly above the intake ports versus offset to the drivers side would likely do well.

We've been looking at their 2ZZ components a bit-- they're doing interesting work on the platform.

 

[kmead]

The thought I had was to flip this part of the manifold so what we see here was on the bottom (ie rotate it @180° on its long axis) and make new runners to mate to it and try to keep the runner length the same or close to OE. Then you could use the standard throttle body, sensors etc.

 

[Santa Barbarian]

The thought I had was to flip this part of the manifold so what we see here was on the bottom (ie rotate it @180° on its long axis) and make new runners to mate to it and try to keep the runner length the same or close to OE. Then you could use the standard throttle body, sensors etc.

We were looking at it... but it's sure a shame Toyota didn't make the mounting face and fasteners symmetric.

That thing also splits, which is nice. But I've spent no time trying to work with the component pieces in that way.

It's actually an interesting departure for Yamaha-- their trickeration usually includes some sort of variable-runner-length mechanism.
Both the early SHO designs incorporated that, as did the Volvo B8444s. All Yamaha designs.
But I don't think the 2ZZ has such provision.

 

[kmead]

We were looking at it... but it's sure a shame Toyota didn't make the mounting face and fasteners symmetric.

That thing also splits, which is nice. But I've spent no time trying to work with the component pieces in that way.

It's actually an interesting departure for Yamaha-- their trickeration usually includes some sort of variable-runner-length mechanism.
Both the early SHO designs incorporated that, as did the Volvo B8444s. All Yamaha designs.
But I don't think the 2ZZ has such provision.

Indeed, unfortunate they weren’t thinking of us when they designed it

You are correct as I understand it, it is just a straight shot without secondary short circuits. Lotus did some tuning on the electronics side to fatten the torque curve @6k where there is a dip as it flips over to the VVTi portion of the program.

 

[Santa Barbarian]

You are correct as I understand it, it is just a straight shot without secondary short circuits. Lotus did some tuning on the electronics side to fatten the torque curve @6k where there is a dip as it flips over to the VVTi portion of the program.

I used to think that "variable runner length" was "oh, it works on paper or the racetrack" thing-- but I have several AUDIs with the 4.2L, and they all seem to uniformly use some flavor of a 2-stage 0r 3-stage effective runner length. Better torque off idle but clearly improved power at higher RPM when it actuates. Of course, like many things "Audi", it worked fine at the factory, but the field's a different story; it's rare to find a D2 where at least one of those bearing-actuators isn't stuck. Or a vacuum pod blown out. It's kind of hard to feel the flat spot, UNTIL you fix the variable runner thing-- and it's demonstrably better.

The 2ZZ is interesting even with the VVTiL-- they don't turn it on until above 6000, despite lots of dyno curves clearly showing a torque-curve gap between 5200 and 5900. It's as clear as day, yet Toyota must have had some reason for not invoking it until you got over 6K.
Hilariously, there's numerous discussions on "tuning 2ZZ engines" where the claim is that it's a variable, PWM-modulate the control to get the effective variation in cam timing. Which I THINK is nonsense- the physical cam and the ledge-roller actuator to turn it on? Is either ON or OFF. I think.

 

[kmead]

I used to think that "variable runner length" was "oh, it works on paper or the racetrack" thing-- but I have several AUDIs with the 4.2L, and they all seem to uniformly use some flavor of a 2-stage 0r 3-stage effective runner length. Better torque off idle but clearly improved power at higher RPM when it actuates. Of course, like many things "Audi", it worked fine at the factory, but the field's a different story; it's rare to find a D2 where at least one of those bearing-actuators isn't stuck. Or a vacuum pod blown out. It's kind of hard to feel the flat spot, UNTIL you fix the variable runner thing-- and it's demonstrably better.

The 2ZZ is interesting even with the VVTiL-- they don't turn it on until above 6000, despite lots of dyno curves clearly showing a torque-curve gap between 5200 and 5900. It's as clear as day, yet Toyota must have had some reason for not invoking it until you got over 6K.
Hilariously, there's numerous discussions on "tuning 2ZZ engines" where the claim is that it's a variable, PWM-modulate the control to get the effective variation in cam timing. Which I THINK is nonsense- the physical cam and the ledge-roller actuator to turn it on? Is either ON or OFF. I think.

Indeed on all those points. The Germans in particular can create the most amazing solutions often through ludicrous complexity. For as long as it works it is freaking amazing but the reality is it won’t work forever and unexpected failures can cause a cataclysm of cascading failures.

I suspect Lotus actuated the system early while it is still in the torque band.

VVTi is not PWM but is continuously variable within its range of motion. As you aren’t jumping to a different cam but adjusting the duration of the existing cam by speeding up and slowing down the cam.
A couple of articles:

How Toyota's VVT-i System Works - Olathe Toyota Parts Center

We guarantee you'll learn something after reading our article about Toyota's VVT-i System. See for yourself.

parts.olathetoyota.com

Toyota Develops New VVT-i Engine Technology | Toyota Motor Corporation Official Global Website

TOYOTA MOTOR CORPORATION announced today the development of its new "Variable Valve Timing-intelligent" (VVT-i) technology, which increases performance and fuel economy.

global.toyota

 

[Santa Barbarian]

OH, it's definitely got two completely different profiles-- and there's no "analog partially-on" mode.
It's either on the low-lift profile or the high-lift. The sliding rocker arm pad is either in play or it isn't.

This is, of course, different from the cam phaser function, which works like it did in pre-1ZZ/2ZZ eras.
But the guys debating how to PWM drive the control to the hydraulic pad-slider are confused I think. The Toyota sources indicate that it's pretty important (especially since you're above 6000) that the pad SLIDEs RIGHT NOW, so you don't experience too many impacts as it's sliding to actuate the high lift profile.

Much like VW's early VVT efforts? Strikes me as a little Rube Goldberg, but as long as it works-- usually.

 

[Jonohhh]

I know little about this engine but.


It appears to be equipped with VVT and VVL- the two often get lumped into VVT for whatever reason.

Like said above, the VVT system is seemingly done via a cam phaser just as the industry has been using since the early, early 2000s. These are generally fairly reliable, and can be varied continuously between two end stops by varying the oil feed pressure. This is done by PWMing an oil solenoid at such a high frequency that the pressure sent to the phaser is steady and does not ripple.

The VVL aspect is not continuously variable- being a "cam changeover" type, it's
likely imperative that the system is switched over as quick as possible. This solenoid should absolutely not be PWM controlled...I imagine those trying to do so will figure out pretty quick

 

[Jonohhh]

Regarding variable runner length. It's fantastic if implemented properly - no clue how well it's done on other platforms. On my BMW, it's done with two electronic flaps in the intake manifold, resulting in 3 distinct resonance peaks.

The flaps fail every ~100k or so...of course. It seems some of the Japanese solutions involve less plastic and a simpler (and usually slower) pneumatic actuator, and probably last longer.

On flap based system, the flap should open or close as quick as possible. While there are continuously variable runner length systems (BMW DIVA), they're far more rare and more complex.

 

[Santa Barbarian]

Regarding variable runner length. It's fantastic if implemented properly - no clue how well it's done on other platforms. On my BMW, it's done with two electronic flaps in the intake manifold, resulting in 3 distinct resonance peaks.

The flaps fail every ~100k or so...of course. It seems some of the Japanese solutions involve less plastic and a simpler (and usually slower) pneumatic actuator, and probably last longer.

On flap based system, the flap should open or close as quick as possible. While there are continuously variable runner length systems (BMW DIVA), they're far more rare and more complex.

The AUDI scheme on my preferred 4.2L's also has two separate flaps-- actuated by vacuum pods and lever arms that rotate an "axle" that runs the full length of the intake. I assume that what the Audi engineers "forgot" was that the inherent vacuum in the runner/plenum system must draw water, dust, grime, contaminants into the bearing surfaces-- which quickly deteriorate and seize up. It's a great feature, but a terrible real-world design.

 

[lookforjoe]

Some questions:

What transaxle is going with this? The sooner you can drop the drivetrain into the opening, the sooner you can start mapping out the overall plan for the physical placement. You will be fabricating the subframe & mounts? What's the plan for shifter layout, axles, exhaust & EMS? NA or turbo?

 

[kmead]

The MR2 site is down at the moment but this link provides some good clues to get one into a similar (but clearly not the same ) mid engine application.

http://www.mr2.com/forums/threads/313-How-to-swap-a-2zzge-into-a-Spyder

I would use the 6 speed Toyota trans.

Monkeywrench racing has some good info and parts:

MR2-S 2ZZ swap

So you’ve decided to replace your MR2 Spyder’s oil-thirsty underpowered stock 1ZZ-FE with toyota’s high revving 180hp 2ZZ-GE. This is an extremely popular swap and is pretty straightforward as engine swaps go. No custom work or fabrication is required since all of the critical swap parts are readily

www.monkeywrenchracing.com

 

[kmead]

Some questions:

What transaxle is going with this? The sooner you can drop the drivetrain into the opening, the sooner you can start mapping out the overall plan for the physical placement. You will be fabricating the subframe & mounts? What's the plan for shifter layout, axles, exhaust & EMS? NA or turbo?

All good questions. Agree whole heartedly with getting the assembly into place to see what crashes with what as early as possible.

 

[Santa Barbarian]

Some questions:

What transaxle is going with this? The sooner you can drop the drivetrain into the opening, the sooner you can start mapping out the overall plan for the physical placement. You will be fabricating the subframe & mounts? What's the plan for shifter layout, axles, exhaust & EMS? NA or turbo?

OH, we''ve had the roller up and down on the lift so many times checking "fit" that I've lost count.

LE5 with F23
LSJ with F35
WhiteBlock Volvo with M66
Several options with F40s
2ZZ with Lancia 5-speed

I wish it were possible to "build" the Fiat box up to the point of handling 200HP at 8200 RPM, but I'm assured it's as likely as Scarlett Johansson calling me for a date night this week. The Fiat box is uniquely styled to fit and obviously makes the axles-hubs-mounts discussion easier.
I"m happy to have someone educate me on what it would take to make the FIAT transaxle worthy here-- it would save a lot of grief.

The Toyota (Aisin?) 6-speed is the natural alternative-- but it's shockingly bulky. I've got a clutch assembly mounted on the 2ZZ ready to mount up the transmission when it comes in. Then the roller goes back on the lift and we'll see. I'm generally targeting, for frame rail considerations, less than 32". IN some ways, the 2ZZ is the easier part of the problem-- the shortest combo (non-FIAT) I've found so far is a Volvo M66 (although M56 also is short) mounted to the 2ZZ.

Mentally, I've already decided to punt on the shift linkages-- I have both Boxster and Volvo cable shifters here in the shop, leaning toward the Boxster set-up because it's smooth, it works and the mid-engine community seems to like it.

 

[Santa Barbarian]

What's the plan for shifter layout, axles, exhaust & EMS? NA or turbo?

By the way--

Shifter-- Boxster
Axles-- Custom (unless I can procure a Magic Fiat Box)
Exhaust-- TBD. At least the manifold is on the right side. Probably custom headers-- People are getting nice power bumps from exhaust mods
N/A-- The 2ZZ spins to 8200. I don't want to miss a note of the symphony. Plus 190HP seems like plenty with a flat torque curve and that wide a power band.
EMS-- Open issue. Megasquirt maybe. Or FuelTech more likely.

 

[lookforjoe]

OH, we''ve had the roller up and down on the lift so many times checking "fit" that I've lost count.

WhiteBlock Volvo with M66
) I've found so far is a Volvo M66 (although M56 also is short) mounted to the 2ZZ.

Just going to ask on this - being a Volvo guy, T5/M66 would have in many ways been my preferred option. Running a T5 to 9K rpm as I did in my old V70 is a glorious sound. Working in the driveway, I never even bothered trying to fit a Volvo drivetrain. I went K24 because the geometry was already figured out. Do you have any pics of the Volvo trial, or was it based on measurements?