A different idea for performance ECU

[younkin1]

Hi guy's,

I'm a new guy here and this is my first post. I'm starting the restoration of my 82 Spider, its a 5-speed car with 72000 miles on it. I'm getting ready to pull the engine and trans out of it for a complete refresh and reseal of the engine and trans, clutch replacement and 5th gear synchro replacement. While its out I want to make same improvement. First my goal- Understanding that the trans and clutch assemblies are good for about 150HP before you start having problems. I want to make the most driveable, reliable, efficient car possible by using all of the engines potential without going forced induction while keeping the costs modest as possible. So in short, 140-150HP, keep costs down.

Here's a little bit about me. I just finished turbo charging my 1993 1.6 Miata and it took about 7 months of evenings and weekends to complete, so I know this Fiat project may take a while. Doing the research for my Miata project led me to some interesting ideas I would have never thought about. In the end I took the 1.6 stock engine from 116HP to 265HP at just over 12psi with stock internals. I think that is pretty good for stock internals. Now I want to take what I learned during that project and do the same to the Fiat, minus the turbo.

Here is what I want to change- Replace the AFM with a more efficient MAP system, install a distributor-less ignition system using COP's (coil over plug), and change the fuel injection from batch to sequential firing with more modern/efficient fuel injectors. I'll also replace the 2.0 head with a 1.8 head to help increase the compression ratio. Once again, most likely leaving the stock internal (depending on their condition upon inspection). One more thing, my girlfriend hates the clutch pedal feel, so I need to fab up a hydraulic clutch system too (thats another discussion)

I started putting together list and pricing out the needed items.I found I'd be poor all summer long if I did it by buying new aftermarket parts. So I remember seeing during my research on the Miata project this guy (tom's turbo garage) who when building his turbo Miata used a DSM (Mitsubishi) ECU to run has engine. In fact, he used the entire wiring harness and sensors in the thing and it ran great. You can google it and see what he did. He's a super smart cat. This started me thinking about the same thing. What car has all of these items I want. I came up with a 1999 Mitsubishi Eclipse NA DOHC 4G63. It has everything I listed except the COP ignition. But I'm sure I can figure that out.

Now we come to the point where most people ask why go through all of that work for such a small gain? My answer- I love challenges and I like being different. Also I think I can do all of the ECU/Injection/Ignition mods for about $500 using used Mitsubishi parts, maybe even less...

Ok guy's, what do you think?

[So Cal Mark]

have fun and keep posting as the project evolves. One note, I don't think I've ever seen a bad 5th gear synchro in one of these trans

[younkin1]

Hi Mark,

Thanks for your response. Not to get off topic, but about that synchro. The transmission symptom is, it will pop out of gear when initially putting it into 5th sometimes you will get a slight grind when re-engaging. I did notice when I initially got the car, the previous owner had the wrong fluid in the transmission. The previous owner had some type of a 80-90w in it I think. I then replaced it with SAE 90 gl-1 lube and the symptoms got better. the grinding for the most part went away but it will still occasionally pop out of 5th on initial engagement. What else do you think it could be? Thanks again

[TimpanogosSlim]

I'm no automotive engineer but for distributorless COP you will need a crank position sensor minimum and often a cam position sensor as well.

You'll notice that our spark plugs aren't in wells and angle toward the firewall. It appears that Ford/Motorcraft DG508 "mustang" coils have a bend in them that will allow #4 to fit. These are "dumb" coils that need an external igniter.

There may be other candidates.

Miller's MMule makes crank pulleys with 36-1 sprockets for a cps.

[younkin1]

Hi Timpanogosslim,

You are correct, it will need cam position sensor. When I did my turbo Miata I found it used a Mitsubishi sensor, this same one that's in a 4g63 engine. I'll have to machine the housing down and fab up a new shafts drive for it. The COP'so are going to be a little more difficult because as you stated the plug positions. My idea would be to create a mounting bracket between the cam box's.

[TimpanogosSlim]

Indeed, looks like that engine has crank position and cam angle sensors. Also, people do convert them to COP using what looks like dumb coil packs to me.

[TimpanogosSlim]

I don't mean to be crapping on your thread or idea. My brain just chews on stuff.

The mitsubishi engine you're looking at has a shorter stroke than the 2.0 TC. I am unsure what implications that will have for your conversion.

[v6spider]

I'll be curious to see what you come up with for a hydraulic clutch system. I've done some measuring and there isn't much room there. My spider is a custom with Chevy Vortec 4.3L v6 with over 300hp.. I'm using a cable pull bell housing from a v8 Monza with a 10" clutch. I'd love to go hydraulic to lighten up the pedal stiffness..

Cheers!
Rob

[younkin1]

TimpanogosSlim,

Don't worry, you're not crapping on the thread. That's what this is all about, discussing an idea. One that may or may not work and I'm okay with that. I love exploring the ideas and seeing what other people think. after discussing this with people like you and others it may end up going a different direction. I may try it and it may fail. But it's not always about the destination but about the journey.

Now to what we were saying about stroke. Technically I don't see anything other than the overall size or how the size is calculated to be a problem. I don't think different borders or strokes make a difference in the calculation other than its a 2.0 engine. But it's funny you say that because I was pondering the thought of using my old Miata ECU and harness to do the project now but its a 1.6 engine. But all were really speaking about our values. Values that the ECU reads. If I increase the injector size to increase fuel and the o2 sensor still reads correctly would it not run? Maybe I would have to increase the injector pulse width? Thanks again for your input.

[TimpanogosSlim]

Yeah, a slight increase in injector size might be all you need to do.

but wtf do i know?

[younkin1]

have fun and keep posting as the project evolves. One note, I don't think I've ever seen a bad 5th gear synchro in one of these trans

Hi Mark,

Thanks for your response. Not to get off topic, but about that synchro. The transmission symptom is, it will pop out of gear when initially putting it into 5th sometimes you will get a slight grind when re-engaging. I did notice when I initially got the car, the previous owner had the wrong fluid in the transmission. The previous owner had some type of a 80-90w in it I think. I then replaced it with SAE 90 gl-1 lube and the symptoms got better. the grinding for the most part went away but it will still occasionally pop out of 5th on initial engagement. What else do you think it could be? Thanks again

[ClarkTheShark]

Well...

Your mind is in the right place. But why use a factory ECU when you can use an aftermarket ECU (megasquirt) that has 10x the tunability and bells/whistles.

Im using junkyard sensors from mitsubishis, fords, nissans even chevrolet. The beauty of megasquirt is that you can set parameters on all of your input sensors. So you can literally make anything work. anything. I would consider a microsquirt ms3. you can mount that bad boy anywhere and it will do everything the modern FI system needs to do.

[younkin1]

Well...

Your mind is in the right place. But why use a factory ECU when you can use an aftermarket ECU (megasquirt) that has 10x the tunability and bells/whistles.

Im using junkyard sensors from mitsubishis, fords, nissans even chevrolet. The beauty of megasquirt is that you can set parameters on all of your input sensors. So you can literally make anything work. anything. I would consider a microsquirt ms3. you can mount that bad boy anywhere and it will do everything the modern FI system needs to do.

Hi Clarktheshark,

I currently run a megasquirt and my turbo Miata and I love it. As I said in the beginning of this thread it was simply out of a challenge and the cost-effectiveness I'm not spending $1200 dollars on a megasquirt 3. I figured from the local yards around here I can buy everything I need for less $500 for sure, maybe even closer to $350.
But as I'm exploring this I'm finding problems myself.
If I had an endless supply of money I would definitely do a Megasquirt 3. But hey, thanks for your input.

[younkin1]

Well...

Your mind is in the right place. But why use a factory ECU when you can use an aftermarket ECU (megasquirt) that has 10x the tunability and bells/whistles.

Im using junkyard sensors from mitsubishis, fords, nissans even chevrolet. The beauty of megasquirt is that you can set parameters on all of your input sensors. So you can literally make anything work. anything. I would consider a microsquirt ms3. you can mount that bad boy anywhere and it will do everything the modern FI system needs to do.

Hi Clarktheshark,

I currently run a megasquirt and my turbo Miata and I love it. As I said in the beginning of this thread it was simply out of a challenge and the cost-effectiveness I'm not spending $1200 dollars on a megasquirt 3. I figured from the local yards around here I can buy everything I need for less $500 for sure, maybe even closer to $350.
But as I'm exploring this I'm finding problems myself.
If I had an endless supply of money I would definitely do a Megasquirt 3. But hey, thanks for your input.

[TimpanogosSlim]

Well...

Your mind is in the right place. But why use a factory ECU when you can use an aftermarket ECU (megasquirt) that has 10x the tunability and bells/whistles.

Im using junkyard sensors from mitsubishis, fords, nissans even chevrolet. The beauty of megasquirt is that you can set parameters on all of your input sensors. So you can literally make anything work. anything. I would consider a microsquirt ms3. you can mount that bad boy anywhere and it will do everything the modern FI system needs to do.

Hi Clarktheshark,

I currently run a megasquirt and my turbo Miata and I love it. As I said in the beginning of this thread it was simply out of a challenge and the cost-effectiveness I'm not spending $1200 dollars on a megasquirt 3. I figured from the local yards around here I can buy everything I need for less $500 for sure, maybe even closer to $350.
But as I'm exploring this I'm finding problems myself.
If I had an endless supply of money I would definitely do a Megasquirt 3. But hey, thanks for your input.

I think he meant microsquirt v3, possibly running msextra code. Less than $400 with an 8 foot harness.

it'll only do batch injection and batch ignition. There are guys who have repurposed the led lines to drive logic coils though.

That's my vague plan at the moment, but you probably saw my thread anyway. uSquirt, edis, etc.

What a modern ecu "needs" to do is debatable. Current OEM ECUs clobber ms3 in terms of cpu speed, ram, io, etc. but they are not very friendly to programmers. On the other hand, they do a lot of diagnostics on the fly that MS simply doesn't do. I understand that the now nearly 10-year-old ECU in my GTI has a pretty good idea of the condition of the oil, for example. I've been assured by tuners that it is not simply a service time. It's also extremely sensitive to fueling and ignition changes and takes advantage on the fly. People tune their water/methanol injection systems by graphing the dynamic timing -- once you have 0 degrees retarded with a 100 octane tune, wmi, and 93 octane, you're done tuning.

I think your plan has the potential to work. I think I've seen people do similar stuff. I don't have a clear idea of the hurdles before you.

It's worth a shot anyway