Dumb Idea? Engine Building Strategy.

alfadriver said:

In reply to DaewooOfDeath :

Back 15 years ago, F did our first DI engine that had 10.5:1 compression and was capable of over 400 hp from 3.5l. It was limited by the trans, as you originally noted- so only 365hp  

Just as I retired, I was working on its truck cousin engine originally released in 2012 and we had to push exhaust temps to 950C  because it had to meet very modern CO rules  

It's been a long time with high compression turbo engines. 
 

For the most part, knock is just as hard at high speed as low now. Low speed combustion still sucks, though. 
 

The one thing I personally like researched is how far open valve PFI could be pushed to be a cheap version of DI. The spray models have gotten really good thanks to DI. 

This is the engine in the Taurus, right?

Also, what sucks about low speed compression? My understanding is that it was mostly to do with knock, ie, the slower speeds literally giving the mixture more time to pre-ignite.

Your point about the very high exhaust temps are interesting as well. I kind of assumed that part of the reason for all the modern turbo engines was the exhaust turbine driving up EGT. This, I assumed, helped fire up the cats sooner. 

In reply to DaewooOfDeath :

It was supposed to be combustion. I think I typed too fast and autocorrect changed it. To get decent flow at high flow, lower flow turbulence is not enough sometimes, and combustion variability goes up a lot. 
 

For a BP Miata, this removes some concerns while adding specific constraints. For example, a mild 175 hp or so turbo build is not going to need a built drivetrain. While running it stoich under boost will cost more in aluminum than you save on gasoline.

I'm honestly not understanding the ask in the original post.  Is there a specific design target for an engine you want to build for a vehicle you want to build?  Or is this all just rampant speculation? And if so, is there a goal with the speculation?

This has been a really fascinating thread so far. 

Still not sure what we are looking to do, but I'm enjoying all the science flying around. 

Further fodder for the thinking folks-

1972 Mercury Comet, modified Ford 289 V-8 (cam, long tube headers, ported heads), Holley 350 cfm 2 bbl carb, 5 speed OD trans, and 3.08 gears in the axle.

My uncle regularly achieved 20+ mpg while cruising. 

Gearing and transmission counts for a lot; it determines how you use the power the engine is making. 

In reply to DaewooOfDeath :

Since you're in Japan and motor swaps are a no-go, have you considered swapping platforms? 
 

I really enjoyed my old NA Miata's. I also really enjoyed my old NB's, my NC's and my S2000's. I would add the ND to that list if I could get comfortable in one. 
 

If you're attached to your NA and just jonsing for more power, I get it. But at a certain point, buying a stock S2000 starts to make a lot of sense- it essentially accomplishes what you're describing in a reliable OEM package. My old 2004 Mazdaspeed Miata also had quite a bit of lag, but quite a bit of character without having to swap to heavy factory components. Even my NC2 feels quite a bit peppier than my old NA's (even if it's nowhere near 220hp). 

If you fit, an ND2 feels like a rocket ship compared to an NA, without the loss of that light, Miata feeling. It also gets outstanding fuel economy to boot. 
 

If you're going down a money pitt rabbit hole, trying to figure out how to make more power out of an NA without heavy drivetrain components, from a pragmatic perspective, it might make more sense to just buy a similar car, with lightweight drivetrain components that's already substantially faster- namely the NC, ND or S2000. Admittedly, from an efficiency standpoint, my AP2 S2000's got poor fuel economy- spinning 4000rpm on the freeway and regularly revving it out to 8k surely didn't help- but the powertrain was a gem. If you want similar acceleration in a lighter package, with far better fuel economy, the ND2 exists. 

Good point on using gearing to aid efficiency. If yours happens to be a 6-speed NB, finding a set of 3.636 gears for the diff nets you similar ratios as the 5-speed in the first 5 gears, and turns 6 into a more meaningful overdrive.

DaewooOfDeath said:

What's your opinion of the Nissan variable compression? I saw a teardown of one and it looks like a lot of extra moving parts in the highest stressed parts of the engine.

It's definitely complicated.  I don't know a lot about them other than I've heard mention of them being about as (un)reliable as you'd expect from Nissan.

DaewooOfDeath said:

rslifkin said:

In reply to DaewooOfDeath :

Higher compression with boost is already possible if you have an efficient chamber design that doesn't need a ton of spark advance, variable valve timing, good boost control, etc.  For example, on the turbo version of the Mazda Skyactiv 2.5, they drop compression from 13:1 to 10.5:1, but that's still a fairly high compression engine.  And the new turbo I6 in that family is listed as 12:1 compression.  Honda's 1.5 liter turbo in the Civic, CR-V, etc. is also over 10:1 compression.  And then you've got stuff like Nissan's variable compression thing. 

What's your opinion of the Nissan variable compression? I saw a teardown of one and it looks like a lot of extra moving parts in the highest stressed parts of the engine.

It's a lot cheaper to get variable compression from variable cam timing. And has some other added benefits. 

TravisTheHuman said:

I'm honestly not understanding the ask in the original post.  Is there a specific design target for an engine you want to build for a vehicle you want to build?  Or is this all just rampant speculation? And if so, is there a goal with the speculation?

It started as rampant speculation with the goal of making a very mild base engine high performance capable with good putzing fuel economy and low torque.

Now I'm thinking it might be interesting to put mild cams in a supercharger BP Miata. 

roninsoldier83 said:

In reply to DaewooOfDeath :

Since you're in Japan and motor swaps are a no-go, have you considered swapping platforms? 
 

I really enjoyed my old NA Miata's. I also really enjoyed my old NB's, my NC's and my S2000's. I would add the ND to that list if I could get comfortable in one. 
 

If you're attached to your NA and just jonsing for more power, I get it. But at a certain point, buying a stock S2000 starts to make a lot of sense- it essentially accomplishes what you're describing in a reliable OEM package. My old 2004 Mazdaspeed Miata also had quite a bit of lag, but quite a bit of character without having to swap to heavy factory components. Even my NC2 feels quite a bit peppier than my old NA's (even if it's nowhere near 220hp). 

 

If you fit, an ND2 feels like a rocket ship compared to an NA, without the loss of that light, Miata feeling. It also gets outstanding fuel economy to boot. 
 

If you're going down a money pitt rabbit hole, trying to figure out how to make more power out of an NA without heavy drivetrain components, from a pragmatic perspective, it might make more sense to just buy a similar car, with lightweight drivetrain components that's already substantially faster- namely the NC, ND or S2000. Admittedly, from an efficiency standpoint, my AP2 S2000's got poor fuel economy- spinning 4000rpm on the freeway and regularly revving it out to 8k surely didn't help- but the powertrain was a gem. If you want similar acceleration in a lighter package, with far better fuel economy, the ND2 exists. 

This is definitely an option. Indeed, I was initially shopping for an S2000. The prices are unfortunately at least 250% higher. 

I honestly would have preferred an NC but I got a good deal on my NB and it came with several quite pricey mods I wanted already installed - nice coilovers, cooling upgrades, etc.

I could swing an ND or an S2000 at those higher prices but I philosophically don't believe it's wise to draw down my no-strings attached cash reserve by more than 10% for a car ever. And I don't do payments, ever.

I don't break these rules for modifications either, so I tend to have cheap cars that get modified slowly over time, and which tend to emphasize creative, low cost solutions. Which is also more fun IMHO. 

In reply to DaewooOfDeath :

If you want fuel economy in a Miata, the gearing needs to change a lot. Cruising at 4000 rpm is a bigger knob to change than cams will. 

DaewooOfDeath said:

This is definitely an option. Indeed, I was initially shopping for an S2000. The prices are unfortunately at least 250% higher. 

I honestly would have preferred an NC but I got a good deal on my NB and it came with several quite pricey mods I wanted already installed - nice coilovers, cooling upgrades, etc.

I could swing an ND or an S2000 at those higher prices but I philosophically don't believe it's wise to draw down my no-strings attached cash reserve by more than 10% for a car ever. And I don't do payments, ever.

I don't break these rules for modifications either, so I tend to have cheap cars that get modified slowly over time, and which tend to emphasize creative, low cost solutions. Which is also more fun IMHO. 

That's fair and understandable. I also loathe car payments. Admittedly, I've taken a small loan out on occasion, but typically only when a vehicle I'm looking for is just barely beyond what I was looking to remove from savings, with the intention of paying it off within the first few months of ownership. But I'm with you- I despise being in debt and respect your unwillingness to compromise your values for a hobby machine. 

I've also purchased a plethora of cheaper cars that I've modded over the years. I'll admit, it can be a pretty fun hobby, but there is a drawback at a certain point- or at least there has been for me. I've absolutely been guilty of buying a cheaper car and dumping so much money into it that I could've bought the more expensive car to begin with.

I do realize you used the words "low cost solutions"- if you can figure that out, more power to you, brother! 20 years ago, I remember coming back from Iraq, buying a new WRX and dropping a metric ton of cash into making it faster... what I ended up with was a car that I spent more on than an STI, but without the STI's 6-speed transmission, DCCD, Brembo's, LSD's, etc. I remember looking at the spreadsheet when I was done and thinking: I could have just bought an STI! Years later, you would think I learned my lesson- nope, I got carried away upgrading a DC2 Integra GS-R, to the point where I was starting to get dangerously close to Integra Type R money. For the record, I'm talking about fun street cars, not track cars built to a specific ruleset- they're going to cost you a fortune in parts anyway- but for a street car, I've found it's frequently cheaper long term to buy the car you wanted to begin with, especially when it comes time to sell; I can typically get my money back on the car, but I frequently lose my ass off on the parts I bought for it. If you can figure out how to play this game for less money than I can, again, more power to you, brother! 

roninsoldier83 said:

DaewooOfDeath said:

This is definitely an option. Indeed, I was initially shopping for an S2000. The prices are unfortunately at least 250% higher. 

I honestly would have preferred an NC but I got a good deal on my NB and it came with several quite pricey mods I wanted already installed - nice coilovers, cooling upgrades, etc.

I could swing an ND or an S2000 at those higher prices but I philosophically don't believe it's wise to draw down my no-strings attached cash reserve by more than 10% for a car ever. And I don't do payments, ever.

I don't break these rules for modifications either, so I tend to have cheap cars that get modified slowly over time, and which tend to emphasize creative, low cost solutions. Which is also more fun IMHO. 

That's fair and understandable. I also loathe car payments. Admittedly, I've taken a small loan out on occasion, but typically only when a vehicle I'm looking for is just barely beyond what I was looking to remove from savings, with the intention of paying it off within the first few months of ownership. But I'm with you- I despise being in debt and respect your unwillingness to compromise your values for a hobby machine. 

I've also purchased a plethora of cheaper cars that I've modded over the years. I'll admit, it can be a pretty fun hobby, but there is a drawback at a certain point- or at least there has been for me. I've absolutely been guilty of buying a cheaper car and dumping so much money into it that I could've bought the more expensive car to begin with.

I do realize you used the words "low cost solutions"- if you can figure that out, more power to you, brother! 20 years ago, I remember coming back from Iraq, buying a new WRX and dropping a metric ton of cash into making it faster... what I ended up with was a car that I spent more on than an STI, but without the STI's 6-speed transmission, DCCD, Brembo's, LSD's, etc. I remember looking at the spreadsheet when I was done and thinking: I could have just bought an STI! Years later, you would think I learned my lesson- nope, I got carried away upgrading a DC2 Integra GS-R, to the point where I was starting to get dangerously close to Integra Type R money. For the record, I'm talking about fun street cars, not track cars built to a specific ruleset- they're going to cost you a fortune in parts anyway- but for a street car, I've found it's frequently cheaper long term to buy the car you wanted to begin with, especially when it comes time to sell; I can typically get my money back on the car, but I frequently lose my ass off on the parts I bought for it. If you can figure out how to play this game for less money than I can, again, more power to you, brother! 

What you are describing makes total sense, and I agree. Making a WRX into an STI is big money, and probably not the smart thing to do long term. Same thing making a GSR into a Type R. And your point on resale is right on. I've never not gotten hosed on resale. 

My experiences racing and building have always been different, though. The most indepth build I ever did was the epynomous Daewoo of Death, and that project included the following:

1. The $300 2.0L engine out of a Daewoo taxi to replace the original 1.5L. ($600 total since I blew up the first one.)
2. A flatbottom made out of scrap aluminum. 
3. A "brake kit" made from a hodgepodge of taxi stuff and Ssangyong (Google it here) minivan parts. 
4. Extensive cooling mods made from cut up metal desks I found in the trash, the oil cooler off of a Kia Bongo (Google it here), and stuff I bought at the gardening center. 

I ended up getting 5 years of frequent track abuse out of that poor turd of a Daewoo Nubira, and my total investment was about $6k. It was never pretty, it sure as hell wasn't going to impress the ladies, and it was about as pleasant to drive on the street as a gokart with concrete wheels, but that was half the fun. Whenever I had a dumb idea, I just tried it because "it's just a Daewoo, who cares?" Some of the stuff I tried failed or blew up but "it's just a Daewoo, who cares?" It also allowed me to do the build with a bunch of middle school kids, getting them interested in physics and engineering. They had no idea what they were doing, and so when they "helped" they broke a lot of stuff. "It's just a Daewoo, who cares?" None of this resulted in a nice car, but it sure resulted in a fun car.

Obviously, I'm going to treat my NB with more respect than I had for that poor Daewoo, but at the end of the day, NBs aren't exactly precious. If I put Kia Sephia cams in there, and strap on a supercharger (which I'm going to try to do with used OEM parts or used aftermarket parts), and blow the damn thing up, "it's just a BP, who cares?" If my small cams, rising boost idea just doesn't work at all and the car runs like crap, "it's just a BP, who cares?" I'm not exactly facing bankruptcy if I have to reinstall the stock NB cams.

This extends to a lot of other things I want to do with the NB. I had a ton of fun doing underbody aero on the Daewoo, and want to try a more sophisticated, outwashing version on the NB. Will it work? I don't know. If it doesn't, sheet aluminum and conveyor belt rubber just isn't that expensive. I want to try building a removable fastback roof, expanding on what I learned doing fiberglass on the Daewoo. My plan is to use a whole lot of craft foam, ceranwrap, and fiberglass repair kits for fishing boats. It's probably going to be ugly. I'm probably going to mess it up a couple times. I don't care. 

It also connects to my experience with shopping. Buying the over-engineered parts off of deeply unsexy vehicles - taxi cabs, minivans, miniature cabover trucks - was a huge money saver on the Daewoo. Minivan brakes, for example, are a goldmine. They aren't sexy, but minivans are usually derived from a company's sedans/hatchback platforms, and they are often 2 plus ton vans. Those brakes can handle heat, and there's a good chance they might swap onto your little car. And I think I can do this with MX5 stuff. Then engines came in a million boring sedans, station wagons, and mini trucks. Same things with a lot of the drivetrain components and little incidental stuff. I like the fact that a 205/50/15 is a lot of tire by Miata standards. I like the fact that almost nothing on a Miata is rare or special.   

I'm not sure I could say the same if I owned an S2000. Those F20C engines are absolute gems, and they're rare. I'm not sure I'd feel comfortable seeing how one runs after I swap in the cams from an 94 Accord DX. Type R Integras are incredible machines, not sure I'd feel comfortable mounting a homebrew underbody aero kit on one. Etc, etc. 

I like hoopties, and my enjoyment of the car hobby is more "rolling science experiment" than "nicest car." I also don't need a daily driver. I live near my job, and all the day to day stuff is within walking distance. I kind of wish you guys lived in Japan, I'd love to invite you all to take part in these experiments. 

alfadriver said:

In reply to DaewooOfDeath :

If you want fuel economy in a Miata, the gearing needs to change a lot. Cruising at 4000 rpm is a bigger knob to change than cams will. 

I very strongly suspect there's some factory mix and match with the drivetrain in my particular NB. The vin tag says it was originally a 5 speed, but there are very much 6 gears in the box. It's also supposed to have an open differential, but it sure drives like it has a torsen. I'm also not sure what combination of 6 speed and rear end gear results in a 6th gear cruise of 2800~ rpm at 60 mph, but that's what I'm doing. 

I do think the gears are a little short for the factory engine, and will probably be much too short with boost. 

What do you think the Sephia cams would do to low speed torque? The car as it sits currently comes on cam pretty noticeably at 4500 rpm. 

You keep talking about these cams saying things like little and tiny, but I haven't heard any numbers yet.

What is the difference in lift and duration between them and how would they phase in the Mazda head? You need that information. With what manufacturers are doing now it wouldn't surprise me if the Mazda cams are already milder. Modern 4 cyl engines are making power to 6000+ RPM on as little as 170 degrees intake duration. 
Also, what you're feeling when it's coming on the cams at 4500 is the VVT, and you should be able to alter that. 
 

As far as how an engine drives, for an NA engine I've become a big fan of engines with good flowing heads (but not massive ports) and a fairly mild duration cam but with plenty of lift and a relatively quick lift rate.  Think LS motor.  In other words, get power via optimization, not by brute force. 

I went with a combo like that in the Jeep and it picked up quite a bit of power compared to stock as well as revving significantly higher (limiter is at 6100 and it's falling off, but not sharply by that point).  Stock the limiter was at 5300 and by 5000 it had absolutely fallen on its face.  The new combo uses much better flowing heads (with some port work done), but they have only slightly larger ports so velocity is still good at low RPM.  The cam has more lift than stock (and I added higher ratio rockers for even more lift and faster lift rate) but it's only a few degrees longer in duration than stock and sounds nearly stock at idle.  Intake and exhaust manifolds were changed for better flowing stuff too (dual plane air gap intake and tri-y headers). 

The end result of that combo was only a tiny loss of low end grunt in exchange for dramatically better top end.  And surprisingly, after I was done fussing with the tuning, it got better fuel economy on the highway than it did stock.  I'd say the fuel economy improvement was in the ballpark of 10%. 

Peabody said:

You keep talking about these cams saying things like little and tiny, but I haven't heard any numbers yet.

What is the difference in lift and duration between them and how would they phase in the Mazda head? You need that information. With what manufacturers are doing now it wouldn't surprise me if the Mazda cams are already milder. Modern 4 cyl engines are making power to 6000+ RPM on as little as 170 degrees intake duration. 
Also, what you're feeling when it's coming on the cams at 4500 is the VVT, and you should be able to alter that. 
 

Mine is a 98 NB, which I think is equivalent to a 99 in USDM. No VVT, though I wish it did.

According to Google, a stock 99 1.8 BP has  intake duration of 237, exhaust of 242. lnlet lift at 110.5, exhaust at 112. 

The contemporary Kias seems to use hydraulic lifters, so no go on that idea.

rslifkin said:

As far as how an engine drives, for an NA engine I've become a big fan of engines with good flowing heads (but not massive ports) and a fairly mild duration cam but with plenty of lift and a relatively quick lift rate.  Think LS motor.  In other words, get power via optimization, not by brute force. 

I went with a combo like that in the Jeep and it picked up quite a bit of power compared to stock as well as revving significantly higher (limiter is at 6100 and it's falling off, but not sharply by that point).  Stock the limiter was at 5300 and by 5000 it had absolutely fallen on its face.  The new combo uses much better flowing heads (with some port work done), but they have only slightly larger ports so velocity is still good at low RPM.  The cam has more lift than stock (and I added higher ratio rockers for even more lift and faster lift rate) but it's only a few degrees longer in duration than stock and sounds nearly stock at idle.  Intake and exhaust manifolds were changed for better flowing stuff too (dual plane air gap intake and tri-y headers). 

The end result of that combo was only a tiny loss of low end grunt in exchange for dramatically better top end.  And surprisingly, after I was done fussing with the tuning, it got better fuel economy on the highway than it did stock.  I'd say the fuel economy improvement was in the ballpark of 10%. 

That's really interesting. I'll have to give it a shot.

DaewooOfDeath said:

What you are describing makes total sense, and I agree. Making a WRX into an STI is big money, and probably not the smart thing to do long term. Same thing making a GSR into a Type R. And your point on resale is right on. I've never not gotten hosed on resale. 
 

I'm not sure I could say the same if I owned an S2000. Those F20C engines are absolute gems, and they're rare. I'm not sure I'd feel comfortable seeing how one runs after I swap in the cams from an 94 Accord DX. Type R Integras are incredible machines, not sure I'd feel comfortable mounting a homebrew underbody aero kit on one. Etc, etc. 

I like hoopties, and my enjoyment of the car hobby is more "rolling science experiment" than "nicest car." I also don't need a daily driver. I live near my job, and all the day to day stuff is within walking distance. I kind of wish you guys lived in Japan, I'd love to invite you all to take part in these experiments. 

Twenty years ago, I wasn’t trying to turn a WRX into an STI—at least, not at first. It started with “just a couple of bolt-ons for a bit more power,” which quickly led to destroying 2nd gear in the transmission. That meant spending all my money on a hardened, cryo-treated JDM STI RA 5MT gearset (back when 6-speed swaps weren’t common or remotely affordable).

With a tougher gearbox, I figured I could push more power—so in went a bigger turbo, injectors, fuel pump, intercooler and inlet hose, plus the cost of dyno tuning time. More power meant I should probably upgrade the suspension—springs, shocks, sway bars. More grip? Might as well add wider wheels and stickier tires. And of course, upgraded brake lines and aggressive pads to keep it all in check.

What started as minor tinkering slowly spiraled into a less reliable, more fragile car—one that, in the end, cost about the same as just buying an STI in the first place. Except far more frustrating due to a blown transmission and eventually a cracked ringland. 


When I bought my GS-R, I wasn't intending to spend Type R money on it either. But it happened. I actually have a build thread for it here on GRM: https://grassrootsmotorsports.com/forum/build-projects-and-project-cars/the-teg-a-dc2-road-to-all-the-hondas-story/258401/page1/



But I'm picking up what you're putting down and I agree with your sentiments about not being comfortable throwing mods at an F20C. In fact, that's part of the reason I sold my last S2000 (although it was an AP2 with an F22C). I talked about supercharging it for 5 years, but never pulled the trigger. It was a beautiful Laguna Blue 2008 (tough to find), low mileage, no accidents, 10/10 matching VINs, in immaculate shape: https://grassrootsmotorsports.com/forum/build-projects-and-project-cars/ap2-s2000-a-love-letter-and-long-overdue-build-thread/272909/page1/

Wonderful car to drive! Unfortunately, as time went on, I started treating it more like an investment and less like a car. The last couple years I owned it, I only drove it 500-700 miles a year. I was afraid of getting a scratch on it. I did do some autocross with it, but I didn't want to put it on track, for fear of a mis-shift that would cost me all the money on a new F22C. I'm not much of a collector- I prefer both building and driving. So, in the end, it had to go. 

I now have an NC2 Miata. Last week I was pitching it sideways in the snow and laughing hysterically the whole time; which is something I never did in either of my S2000's, in all the time I owned them, because "special car". I like the more disposable nature of Miata's and figured if I ever blew a motor in my NC, 2.5L replacements are dirt cheap and plentiful. 

With you wanting to tinker for science, it sounds like an NB is a better fit. BP motors are still fairly affordable and easy to locate. I get it. My initial thoughts were if you wanted S2000 levels of power, without an engine swap, while maintaining S2000 levels of OEM reliability, buying an S2000 would likely be cheaper, long term; but if you are just looking for something to tinker with, I retract my previous statements- have fun with the NB!