Hey guys, looking for some information on the stock 20b setup, hoping you guys can help me. I have an unusual car, a 1959 Edsel Corsair. For those unfamiliar, here's what it looks like:



The drivetrain is dead, and I'm looking outside the box for replacements. A fuelie V8 is the obvious choice, but I'm a Mazda guy at heart and a rotary Edsel would be... unique. With the Big Block, the car is about 3700lbs and I have TONS of space for motor mounts, intercoolers, plumbing, etc. I've done a fair amount of searching, but most of that is geared towards high power setups. I'd be interested in rock stock, auto trans and all.

I see a ton of threads that say to rebuild before install. Do the motors usually have problems, or is this just insurance for the 700hp monsters most people seem to be running?

Most of the info also points to aftermarket ECUs. Other than the difficulty of tracking down the stock coil packs, is there any reason not to keep the factory computer?

Any pictures of the bottom of the motor? I'm trying to get an idea if the oil pan will work with the Edsel's crossmember or if I'd have to mod.

Any pics of the linkage on the auto trans? Is the trans controller seperate from the ECU in case I go aftermarket later? How about the output? Same bolt pattern as an RX7?

Thanks guys, I haven't decided if I really want to go this way, but this info will help.

 

Its your car and project but I'd make sure the car doesn't have collector value first.

If your willing to use the 20b tranny, you can get nearly everything you need from a Eunos Cosmo front clip.

 

The collector value of an Edsel is debatable. They sell very cheap in anything other than Concours condition, and this one is way too far gone to do a 100% original resto. It looks good from 10 feet, but the interior and mechanicals are pretty incomplete.

So the stock ECU will work? Is it seperate from the transmission controller? Any of the trans/oil pan pics?

Thanks again!

 

I sent you a PM with a link to some oil pan pics. Maybe you can leave the stock pan on?

 

It is mostly for insurance in general, regardless of horsepower. What are the odds that a 14+ year old engine is going to work properly, especially after sitting outside in a junk yard for many years?

The factory computer does not work properly unless it is attached to all of its input devices. Therefore, to use the factory computer, you would either need to have all of the functioning auxiliary parts, or you would need to have the electrical background to make the computer think it is attached to the parts. Given the complexity and fraility of such a system, most people choose to simply install a standalone EMS and be done with it.

The oil pan is fairly large at the rear of the engine, and there is a small bump at the front of the engine for the oil pickup line. The small bump is what causes the fitment problems with the FC RX-7 sway bar.

The transmission is electrically controlled by a computer. The 20B engine bolts up to any RX-7 transmission.

 

Thanks guys! Great info!

Looks like the oil pan would be ok as is. The high portion in the middle should stradle the crossmember and steering linkage just fine. I may need to clearance the firewall a bit to set the motor that far back, but that isn't tough on this type of car.

The transmission looks like it would take some figuring out. The linkage is on the driver's side now, but it looks to be on the passenger side in those photos. I'm sure some digging in a JY could supply a solution there. Otherwise, I found plenty of info about using the TII trans. Hot rod suppliers make hydralic clutch pedals for old 50's Fords.

I understand the rational behind rebuilding and using aftermarket controls. At that point, I think I'd rather go n/a ported since 700hp would rip this car to shreds. I would think that it wouldn't be too chalenging to build a carb manifold for that motor. A 4bbl could definately do it, maybe a 3bbl IDA. Any options for retrofitting distributors? Maybe using Megasquirt to fire the plugs?

 

From what I understand, there are 4? solenoids that control the AT. That shouldn't be much of a problem to control, even with a ghetto analog setup. Hardly anybody wants the 20B AT, so you could probably nab a spare for cheap and stash it in your garage in case you need it later. BTW, I think the 20B-REW with the AT would be perfect for your car.

Based on my experience, I recommend that you simply use the stock turbos and EFI if you want an easy route to installing a 20B in your car. An NA setup would require custom headers and an entire custom/aftermarket intake system, plus it would be louder, have less power, and probably worse emissions than the stock 20B-REW setup. All you need for the stock 20B-REW is a custom downpipe, standalone EMS, and maybe some type of cheapie Ebay intercooler if your engine doesn't come with the stock intercooler. The stock turbos and fuel injectors will work great for your application. If you want something cheap, the Microtech MT-12S is proven to work with a 20B. I am not aware of anybody successfully running a 20B with a MegaSquirt. Maybe your situation is different, but I would rather spend a few hundred extra dollars on a proven EMS as opposed to wasting my time experiementing with an unproven EMS.

 

The TCU shouldn't really be *that* hard to use. I was looking at the linkage in the picture above, trying to devine how the trans is hooked to the shifter. If it's a mechanical link, I'm on easy street. The Edsel's shifter is a standard PRNDL1 arrangement with a rod linkage. If the 20B is electronically actuated (i.e. no mechnical link under the shifter) I can try and disect the Mazda mechanism and try to adapt the controls to a push button arrangement. What would be really cool would be to use the Select Shift wheel form a '58




Custom DP is easy, tons of room available. I want lakepipes on this car, so I can route one turbo directly into each pipe. Simple. Intercooler is easy enough too, I've got enough room to run an intercooler from a diesel truck.

Now as for ECU, I like the MS because a few friends have used it with great success. The computer is about $400 and includes boost control, excellent tuning software and its open source so you can teach it to work with all sorts of sensors. It's having friends that run it already that make me want to do it. One car is a high boost Volvo, the other is a fire breathing Stingray. Both run massive power like they're factory. That said, I'm willing to check out other stuff, but it better be pretty damn good if it starts at $1000 more. When I searched Megasquirt and 20B, I saw some carping that the poster couldn't get the spark timing right. Not sure why, as MS has been used on 2 rotor motors before. Something about not being able to get the computer to fire the trailing spark after the primary spark (?). Very skimpy thread though.

So the stock ECU is no good? If I get a clip it seems that the stock computer would work better than just about anything. I know Mazda has built some crappy EFI setups in the past though, does it have inherent problems?


Thanks again guys. I think this would be a great motor for this car. The Edsel is a unique and quirky vehicle, and the 20B is likewise a very distinctive powerplant. My other options are rebuilding the stock motor and adding injection and a newer trans, swapping a big 460 in, or doing the rotary. I'm not sure which way to go yet, but the 20B is very tempting (if maybe a bit expensive). The rotary's biggest selling point is that it can belt out big block power with smoothness that a V8 just can't touch.

 

If you get a front clip from a running car its simply be a matter of pulling everything from the clip and installing it in the ford. Given how much room you have to work with this would be the easiest custom 20b swap imaginable.

If you think you'll want more than say 350bhp you need to be looking at using a different transmission and likely an aftermarket ECU. It only gets more expensive from there.

 

I got my 20B to run NA with ignition on a MegaSquirt. Leading only, though (but late leading, with proper tach signal). This was back in '05. I think I was the first. Please don't ask because I can't recall the specifics.

You say you have friends that know the MS? Good. Use them lol.

 

I do have to question how much get up and go the car will have since it is so heavy. Im not sure at the same time if that is important to you or if you are just going for something different.

 

I think a 20b can move the Edsel around fairly well. This car weighs 3700lbs as it sits, and that includes a 700lb all iron big block, and an iron cased automajik trans. For comparison, the stock Mazda Cosmo checks in at 3450lbs. It's actually possible the Edsel may be a tick lighter than the Cosmo once that big lunker is out from under the hood.

I'm not terribly concerned with how fast this thing ends up as long as it can move with reasonable authority. The emphasis is definately on different. Now that I'm seeing the costs involved, I'm even wondering how a 13B-REW might do. They're shockingly cheap these days, and with the stock twins it might be a reasonable option. Particularly with a stick. Keep in mind that this car normally only uses 2 speeds. First gear in an FD box is more than twice as short as a Fordomatic's 1.47:1 starting gear. Torque multiplication helps alot.

Now what I do wonder is this. Using either the REW or the 20B I wonder if anyone has tried running the turbos in series. It's usually only done on high output diesels (like 18 wheeler high output). Basically, you blow the output of one turbo into the input of the other the boost multiplication is exponential. Turbos work on pressure ratios, so if each turbo is nominally creating .6 bar (<9psi), the end result is 1.56 bar (~23psi) of positive pressure. In theory you could get that boost with stockish lag, spooling one turbo then the other. Just a thought since a custom ECU and rebuild is needed anyway.... Obviously some fab would be needed, and the intake temps before intercooling would be tremendous. Good thing I have tons of room in the nose

 

What you're referring to is called Compound Turbocharging. With both the REW and 20B the secondary turbo isn't large enough to due this, nor are the turbos plumbed together in a such a way that you could try this as is. AFAIK this hasn't been tried on a rotary engine yet. At the moment there is a thread on Compound Turbocharging in the turbo forum.

You could use the REW but I would suggest keeping the stock sequential turbos so you actually have some torque off idle. Also the Eunos Cosmo used a 4.30 diff with the REW and a 3.90 with 20B.

 

The turbos are mounted on a log-type manifold that has only one exhaust outlet. However, if you ditch the stock collector pipe, I am pretty sure you could fabricate a split collector that routes the exhaust to one sidepipe and the wastegate discharge to another sidepipe.

It is a big advantage to use an EMS that has good local support.

Typical aftermarket standalone EMS: about 30 good new wires.

Front clip: More crusty old wires than I can count:
http://www.agt.net/public/nowback/rx7club/20b/2/

How many front clips come from a running car, and how many front clips include the exhaust solenoids that are mounted above the differential?

+1

 

Firstly, I've never gone looking at front clips. I figure you can get them in any state of condition, running or not, depending on whats available at that time.

Secondly, if their only solenoids for the exhaust, do you absolutely need them? If so I wouldn't think it would take much to remedy unless there's something unique/special about them. Fix it or trick the ECU into thinking they are there and functioning. It seems like a relatively small thing compared to the rest of the swap.

Is there anything else that would pose a problem? If there is don't leave us in the dark.

quattro

 

Yes, they are in various states of condition, but I have never known anybody to cut up a perfectly good car, so I think it is overly-wishful thinking to expect everything to work perfectly.

The stock ECU needs them for input to work perfectly. If you don't have them, it will not work perfectly. Yes, you can jury-rig the ECU to work without them, but my point is that the stock ECU will not work perfectly without all of its auxiliary components. If any of the auxiliary components fail, then good luck finding a replacement. Only one person on this forum took the time to try using the stock ECU, and even he recommends against it.

The stock ECU also requires the stock ignition coil packs, which commonly have a badly pitted L3 socket. I am not aware of any other coils that will work with the stock ECU, but maybe you can check the AUS or NZ forums to see if they know of any.

The most important things to keep in mind:
- A 20B conversion is NOT cheap.
- A 20B conversion is NOT easy.
- A 20B conversion is a CUSTOM process, not a plug & play item.
If at any time you think differently, then think again. This advice comes from experience.

 

I understand. I know of only two or three members who've done a 20b conversion for less than $10k. I'm saying it should be cheaper and easier for him mainly because he's not looking for more power than stock (no changes to the engine or turbos) and the car would have a huge engine bay (compared to an FC or FD) to work with.

 

v
v
v

 

Alright.... I'll shut up now.

 

just drop a 302 out of a late 80's model mustang and call it a day