Trackday/TT FD Build - EFR, Megasquirt, & more!
 

I've been meaning to start a build thread on my car for quite some time, but haven't gotten around to it until now. I'm around 1.5 years deep in this build, so I'll try to break up the work I've already completed into several replies that I'll post in the coming weeks.

Intro

My previous RX-7 was an '87 Turbo II. I bought this car in '07 with a blown engine (blown coolant seal which snowballed into trashed irons) and rebuilt it for my high school senior project. I kept it pretty close to stock throughout college, mainly just using it as a weekend toy. I was heavily involved in Formula SAE in college, so I didn't spend as much time as I otherwise would have building the FC (but I didn't have the funds to do so in college anyway). After starting my full time job, I started looking for ways to get more into road course driving/racing. To do that, I wanted a car to dedicate to the cause:
Yeah, I chose option 3.

So, I started looking around for an FD. I favored a stock-ish car so I could have a blank sheet of paper to start with. Also, sunroof cars were out, and I secretly preferred a red car (even if I told everyone I didn't care about the color). I knew it would be a long search, so I spent some time doing an "experimental" build on my FC and tooling up my shop (basement/garage). I got a 7670 EFR turbo and built a custom V-mount. In the process, I bought a TIG welder for my shop (gives you superpowers), and I already had a mini-mill and mini-lathe, along with the other standard shop tools. So I put all of this together with a Megasquirt MS3X and had a pretty nice FC mountain carver:

https://cimg6.ibsrv.net/gimg/www.rx7...b28b2a2e84.jpg

https://cimg1.ibsrv.net/gimg/www.rx7...eadb1755d7.jpg

Anyway, back to the FD search. At this point, I had written some XML code that would scan every Craigslist in a 500 mile radius and dump the results into an excel sheet that I was keeping (RX7 or RX-7 or RX 7, 93, 94, 95... you get the idea). I would run this a few times a day to look for new listings. Finally, in April 2015, this popped up:

https://cimg7.ibsrv.net/gimg/www.rx7...a57911af17.jpg

1993 R1, 36K miles, all stock, $16K. I found the ad within an hour of it being posted and called him immediately. It turns out, an older guy from Tennessee had bought this car in 2008 and parked it in his garage under a car cover. He died in 2015, and it was inherited by the guy who posted the ad. He didn't really know anything about FD's, he just had a bunch of inherited stuff and was looking to unload it. So, with that said, I drove out to pick it up that weekend. The car was now at the NC outer banks, which was about a 6 hour drive from my house. I opted to rent a uhaul trailer and pull it back, rather than try to road trip in a car that hasn't really been driven since 2008. The transaction went pretty well. The car wasn't "showroom perfect" like I had secretly hoped, it had spent the early part of its life in Florida, so there was a bit of corrosion here and there. Still, no rust at all, and pretty fantastic mechanically, super quick start-up. I ended up getting it for $15.5K.

Arrived to pick up the car the night before. This is still the only time I've been to the beach in the last 3 years, heh.
https://cimg5.ibsrv.net/gimg/www.rx7...0935f2bc88.jpg

The drive home:
https://cimg8.ibsrv.net/gimg/www.rx7...e7da1ad604.jpg

Finally home:
https://cimg3.ibsrv.net/gimg/www.rx7...b89aa1199b.jpg

https://cimg7.ibsrv.net/gimg/www.rx7...ca6cbfeea5.jpg

https://cimg3.ibsrv.net/gimg/www.rx7...7022a52235.jpg

Initial plans were just to play around with it for a bit in stock-ish form. The stock twins had always intrigued me, and this was probably one of the last examples with a working stock sequential system. In the meantime, I parted out my FC and started buying parts to build the FD.

Anyway, that's all for now. Like I said, I'm 1.5 years into the build, so there's a lot more to come just to get up to the present state of the car. :nod:

I can't wait to see the rest of this thread. I always love track car builds.

Killer deal. Can't wait till you get this build thread current. It's like binge watching a whole season of a TV show. Can't wait.

Post up some pics of your home shop too! I'm psyching myself up to get a mill and a lathe.

Tall driver mods

The first step was to shoehorn myself into the car. The stock seat wasn't comfortable for me to sit in as I had to lean really far back to get head clearance. Additionally, the stock steering wheel felt too large in diameter and would hit my knees.

Luckily, I had a knockoff Bride Low-max (Vios III?) seat and 350mm Momo Monte Carlo steering wheel left-over from my FC. I put the wheel on and spent some time trying different seating positions with the seat "loose" in the car. The location I ended up favoring was: seat bottom on the floor, maybe ~30 degrees reclined. This felt really comfortable and still gave just enough headroom to clear the headliner with a helmet on. The Momo wheel hub and a 1" spacer put the steering wheel around 2" closer to me than the stock wheel, which made the slightly reclined position work well.

To make the "on the floor" seating position work, I rigid-mounted the seat rather than using sliding seat rails. I made some low-profile faux seat rails:

https://cimg9.ibsrv.net/gimg/www.rx7...7629dc0787.jpg

L-brackets connect the faux rails to the seat:

https://cimg6.ibsrv.net/gimg/www.rx7...2f84db00ae.jpg

https://cimg0.ibsrv.net/gimg/www.rx7...39577732ca.jpg

Also, since the seat belt buckle mounted to the stock seat, I needed a way to mount with the new seat. So, I designed this (also shown above):

https://cimg8.ibsrv.net/gimg/www.rx7...0f997fa32a.jpg

It mounts between the right rear seat mount bolt and the seatbelt anchor bolt in the floor pan. I forget how much load I applied to it in that FEA, but I remember being satisfied that the buckle would break before my mount would.

And voila, now a 95th percentile male fits comfortably with a helmet.

https://cimg0.ibsrv.net/gimg/www.rx7...65968c4673.jpg

And, per request, some shop pics:

https://cimg4.ibsrv.net/gimg/www.rx7...3097657c2c.jpg

https://cimg6.ibsrv.net/gimg/www.rx7...77e6e54f0a.jpg

https://cimg8.ibsrv.net/gimg/www.rx7...11398b4636.jpg

https://cimg0.ibsrv.net/gimg/www.rx7...99eae85065.jpg

Both the mill and lathe are from Harbor Freight, I bought both of them 4 years ago for ~$800 total. I have countless machining hours on both of them and they've been awesome. I upgraded the mill to a belt drive and an air spring and added a quick-change tool post to the lathe, all from Little Machine Shop. The lathe even has and auto-feed and can do threading. Yeah, they're small, but they're big enough for 95% of the parts I need to make for a car.

The TIG is an Eastwood TIG 200. It does AC and DC, HF start, pre and post flow control, pretty much everything you need. The pedal is a little bit non-linear, but it's easy enough to get used to.

Sweet setup!!

Random playing with the stock turbos

Not too much in this update, just some minor work on the car while it had the stock sequential system.

All of the vacuum hoses were original, and the few that I could get to easily were rock hard, so I decided to replace those just for good measure. Plus, it gave me a good excuse to dive into the engine bay a little bit.

There are a couple of turbos under there, somewhere...
https://cimg4.ibsrv.net/gimg/www.rx7...7bdbe339a4.jpg

https://cimg0.ibsrv.net/gimg/www.rx7...c6edf4630d.jpg

This picture says it all... One of the Mazda engineering managers in the early 90's must have had tons of stock in their vacuum solenoid/line supplier.
https://cimg4.ibsrv.net/gimg/www.rx7...e220b385ad.jpg

With the vacuum lines complete (one of the most difficult things I've done on a car), I installed a set of CXracing coilovers as a placeholder until I could decide on something better. For future reference, these aren't bad as far as a budget coilover goes. They're as good as any other "cheap" coilover I've used on other cars (BC racing, Stance, Megan).

https://cimg4.ibsrv.net/gimg/www.rx7...fc6371b35a.jpg

Also picked up some RPF1's from the RX7club classifieds. 17x9 +45 fronts, 17x9.5 +38 rears. Awesome fitment. (RPF1's are my go-to wheel choice, if you couldn't notice...)

https://cimg6.ibsrv.net/gimg/www.rx7...4c10b4b617.jpg

In the meantime, I kept tracking the RX-8. If you have an SCCA track night venue nearby, I highly recommend it, it's an awesome way to break up the work week. Here's a video from Atlanta Motorsports Park:

At this point, I was getting pretty excited to get the FD on track. I had spent too much time in the RX-8 at track days catching people in the corners and being held up, only to have them pull away from me on the straights and repeat the cycle in the next set of corners. More power would hopefully give me a bit more freedom to work through traffic, not to mention the other advantages. :nod:

So, the single turbo project begun. I briefly considered tracking with the stock twins, but I wanted a more reliable system (especially cooling).

The last pictures of the FD before being torn down for the transformation:

https://cimg3.ibsrv.net/gimg/www.rx7...a26922a5ee.jpg

https://cimg5.ibsrv.net/gimg/www.rx7...1df9bc7c41.jpg

And a preview of what's to come. Seems like a suitable replacement!

https://cimg6.ibsrv.net/gimg/www.rx7...be0642c44a.jpg

I too have an EFR7670 on my list. I'm hoping to get my project going soon too. Looking forward to the continuation of your build!

Is that the new dual AR housing EFR? I only saw pictures of those at SEMA.

Turbo stuff

A few details about my turbo selection: Power goals for this project were somewhere between 300 and 350whp. I feel like this is a good, reliable power level for a 2700lb track car on street tires. At some point in the not-so-distant future, I would like to build the car for NASA TT3 -- which means a power goal of ~320whp without aero or ~305whp with aero. This narrowed it down to an EFR 7670 or 8374, so let's dig into the details to choose between the two:

7670:
https://cimg6.ibsrv.net/gimg/www.rx7...1a222485e2.png

8374:
https://cimg9.ibsrv.net/gimg/www.rx7...ecd01c681d.png

Using Matchbot, I plotted a 12psi, 2000-8000rpm flow curve on both both compressor maps. The points fit the 7670 compressor map fairly well, with the peak efficiency islands sitting in the middle of the powerband and no major drop-offs at the high or low flow points. The 8374 map looks like a bit of a mismatch at first glance, but the high RPM compressor efficiency is pretty outstanding. The surge line looks to be in a bad spot, but the 8374 cars I've seen on the forums don't seem to have low RPM surge issues. Without getting too lost in efficiency comparisons, it's a little more helpful to look at the resulting compressor outlet temperatures, also a (raw) output from Matchbot:

https://cimg6.ibsrv.net/gimg/www.rx7...a67dc5057e.png

The 8374 has a bit of a temperature spike at low-mid RPM, but it falls back down once it hits its efficiency sweet spot. I don't think the temp spike is a huge deal; it's at low RPM (low mass flow) and the intercooler can soak up the extra heat pretty easily. The 7670 is definitely at a slight disadvantage here, with a climbing temperature towards redline and ~20F hotter temps at 8000rpm. In the grand scheme of things, 20F is only worth ~10hp, and the intercooler should help to bring that 20F delta down to, say, 10F. So the steady-state performance advantage goes to the 8374, but the 7670 isn't at a huge disadvantage.

Another goal of the build was super-fast transient response. Being mainly a track car, every tenth of a second of turbo lag makes the car more difficult to drive. EFR turbos are well-known for low lag in general, but the 7670 should have a slight advantage here due to the lower inertia.

At the time I had been making this decision, I had also been following Cartmill's 8374 build thread. He seemed to have a minimum power level of 335whp at the limit of boost creep (12-13psi) -- I would like the option to run lower power than this. Forum feedback seemed to suggest that the 7670 could run sub-10psi boost level, so the advantage went to the 7670 here.

So, all things considered, I chose the 7670. Now, somebody has already noticed the dual actuators in the picture in my previous post. I have a prototype version of the Variable Turbine Volute (VTV) turbine housing. I got it to provide some "field feedback," but since the development is still ongoing, I won't be sharing any VTV-specific results in this thread. Just pretend it's the standard 0.92 A/R IWG housing (which, as long as the valve is kept open, it is). ;) It was shown at SEMA and you should be able to find some details on Google.

I like your thought process on your selection. I am personally leaning toward the 7670 for my car. I'm still trying to make up my mind if I want to run the low or high compression rotors.

The guys at Turblown have tons of experience with the EFRs and usually steer people into the 8374, simply because the spool characteristics are not "that" different. I have seen loads of videos on both turbos and they spool lightening fast.

I will simply pretend I didn't see the second actuator. When you are able to share about it, please let us know. I really dig the potential of that turbine housing. With that housing, you could feasibly run a much larger turbo and get even better low end torque, like say a 8380 hybrid (if they ever make one).

NASA TT3 -- which means a power goal of ~320whp without aero or ~305whp with aero

Well, NASA dyno classing.
You chose correctly since you want to restrict horsepower, but torque, turbo response and boost are unrestricted.

7163 and higher boost might have been even better for 320rwhp if you think you could keep that combination running long enough (higher exhaust manifold pressure/less reliability).

I think EFR 7163 super core in stock '89-91 RX-7 manifold and turbo housing would be great for max power under 5,250rpm while limiting power after 5,250rpm.

I don't agree with your inputs into matchbot for a rotary. Those flow numbers look low to me unless you are plotting for ~250rwhp there and a big pressure drop on the inlet and outlet of the turbo for the PRs.

The 76 series consistently seems too small for "large" track scenarios(not autocross). But, that's just the view from my armchair, watching all you nuts actually make it happen.

I keep seeing Elliot with Turblown say that the 8374 spools just about the same and has much better midrange. On the other hand, you already have a turbo with the VSV setup. The response very well could be amazingly better.

Unfortunately, the FD is pretty much restricted to the TT# (dyno) classes. Ideally, I would like to run a TT letter class, but it's not really possible unless you have a near-stock FD, as the TTC+14 starting point is pretty restricting in terms of mods to stay in TTB. In theory, I could tune the powerband to make 320hp everywhere, so I could ramp the boost really aggressively at low RPM and make tons of torque like you're saying. The 7163 is a pretty attractive turbo in general, somebody needs to jump on that bandwagon and give it a try on an RX7. I do like the idea of using an S5 TII turbine housing -- unfortunately the volute shapes wouldn't be ideal for that turbine wheel. The 63 turbine is a mixed flow wheel, so the turbine housing is designed to introduce flow to the wheel with an axial component; which the S5 TII housing wouldn't do. Still, just because it's not ideal doesn't mean it wouldn't work.

I think the Matchbot inputs are pretty realistic. Take 7000rpm for example, that's 0.15 cubic meters per second of volume. NA with 100% VE and 70F IAT (1.18 kg/m^3) would be 0.18kg/s. 12psi with 100% (relative) VE and 70F IAT (2.16 kg/m^3) would be 0.33kg/s. The 70F IAT is unrealistic for a track day, I estimate something like 120F, and that bumps the mass flow at 100% VE, 12psi, 7000rpm to 0.30 kg/s. I'm not sure what the VE numbers actuality are for a stock-port REW, but I tuned them until the power curve looked realistic. I ended up somewhere around 93% at 7000rpm, so 0.28kg/s.

I do like the idea of using an S5 TII turbine housing -- unfortunately the volute shapes wouldn't be ideal for that turbine wheel. The 63 turbine is a mixed flow wheel, so the turbine housing is designed to introduce flow to the wheel with an axial component; which the S5 TII housing wouldn't do. Still, just because it's not ideal doesn't mean it wouldn't work.

I don't know...

----------
Story time-

I accidentally got a 60-1 P trim S5 stock hybrid with mixed flow wheel when the stock exhaust housing was bored deep so one volute slot was entering the turbine wheel from behind and the other volute slot was entering the turbine wheel "correctly" in the tangental fashion.

That turbo spooled like a mofo and would leave black marks just engaging the clutch from idle. 20psi boost by 3,000rpm with wastegates open since 5psi.

Yes, problem was I could not control the boost creep. Even after adding an external 60mm wastegate. Finally I had to almost double the manifold and housing exhaust runner volume up to the wastegate passages (heavily favoring wastegate flow) and double both volute slots width (Divergent AR now) to get boost under control.

That killed the response and the boost creep all in one swoop, now it behaved as a well set-up T4 60-1 P trim does on a rotary. Finally when I wanted to raise the boost to 18psi on race gas on that turbo I had to anti-lag the turbo into spooling in the upper rpms with AFRs burried richer than 10:1. Crazy rich for 2psi more boost netted me 5rwhp peak and around 30rwhp max under the curve.

I didn't feel anything like that turbo response (un-boost-creep-fixed) until I put the EFR 7670 on the car.
----------

Basically, if you skip the story time- I had a stock hybrid mixed flow turbo by accident. But I think it could be replicated with the EFR 7163 by boring the exhaust housing deeper when adapting the supercore to the stock exhaust housing.

Boost would probably have to be controlled with exhaust restriction.

Turbo stuff - Part 2

So, with the turbo on hand, it was time to make a manifold. As you know, I originally had this turbo on my FC for a short time, so most of the "up-front" work on the manifold was for the FC. First I spent some time with the turbo hanging from an engine hoist in the engine bay to find the best position for the turbo. I wanted it to be mounted as low as possible, and as far rearward as possible. Once I found that spot, I bolted a blank T4 flange to the turbo (from ebay) and a blank flange to the engine (from Racing Beat). I measured the relative position between these flanges as best I could, and I re-created that positioning in CAD:

https://cimg8.ibsrv.net/gimg/www.rx7...e4984da29d.jpg

I connected the flanges using 1.5" schedule 10 stainless tubing.

Based on the model, I made a jig and put everything together:

https://cimg0.ibsrv.net/gimg/www.rx7...ca0fa2e821.jpg

At this point, I had planned on having someone else weld it -- but I got ambitious and decided that this was a good time to buy a TIG welder and teach myself how to weld. It ended up going pretty well, although if I had to do it over again I could do a better job, but the end result wasn't too bad:

https://cimg5.ibsrv.net/gimg/www.rx7...71e8514b13.jpg

So, that was for the FC, and it worked quite well on that car (DIY manifold tip: Run it on the car and get it hot to stress relieve it, then take it off and surface the flanges). When I started this project on the FD, I tried fitting this manifold as-is, but it didn't quite fit. The FD has a narrower engine bay than an FC because of the extra wheel well space needed for the UCA's; this caused my wastegate actuator to hit the side of the engine bay. After some (less-scientific) fumbling in the engine bay, I realized that I could likely modify my FC manifold to fit the FD. I ended up cutting the T4 flange off of the manifold, taking ~15 degree pie slice out of both runners, and welding a new T4 flange onto the manifold:

https://cimg4.ibsrv.net/gimg/www.rx7...f6bb4fed97.jpg

The result on the car:

https://cimg5.ibsrv.net/gimg/www.rx7...654acf4aed.jpg

I'm pretty happy with the positioning; it's really low, it's far back enough that I can fit a comp inlet pipe for an eventual CAI, there's enough clearance from the actuator to the engine bay, and it's still fairly easy to access the bolts.

Also made a downpipe (similar situation to the manifold, I had originally made this for the FC but modified it for the FD):

https://cimg2.ibsrv.net/gimg/www.rx7...c601790080.jpg

Making a downpipe is especially not-fun, I spent quite some time under the car holding up 3" tubing with my forehead while I marked where to make cuts... For the rest of the exhaust, I used an RX7 store resonated midpipe and a Racing Beat single exit muffler.

Picture of everything together:

https://cimg2.ibsrv.net/gimg/www.rx7...834572808d.jpg

Other random notes: I used -6 hardline to 3/8 hose for the water port connections (turned out really well). The oil inlet is -4an SS braided line with 90 degree hose ends (I wrapped it in a thermal sleeve to protect it from the turbine housing heat). The oil drain is just 3/4" aluminum tubing welded to a 2-bolt flange, one on the turbo and one on the engine, connected by hose elbow. I used a Pineapple racing "BT" idler pulley. And notice the small LIM heatshield.

The next update will get into some Megasquirt/electronics details.

keep it up love the build

ECU & Electronics

When it came time to choose an ECU, I initially gravitated toward Haltech, simply because of the amount of experience with that platform in the RX-7 world. My knowledge of Megasquirt at the time was from the MS1-era, and I saw it as a super-cheap system held together with Radioshack parts and bubble gum... However, around that time, Aaron Cake came out with his MS3 write-up. After doing some more MS3 research, I realized how capable of an ECU it is; sequential injection, direct fire ignition, auto-tune, even advanced features like launch and traction control. So, the choice was easy; why spend thousands of dollars on an ECU when a Megasquirt has all of the functionality at a fraction of the cost? So, I went with an MS3X. Still, it presented a few minor challenges on an FD:

- Electric OMP control: Not supported by Megasquirt currently. Choices were to either premix or build a separate controller for the OMP stepper motor. I ended up premixing, but not without an attempt at building a controller:

https://cimg6.ibsrv.net/gimg/www.rx7...b3391580fe.jpg

I used an Arduino to read an 0-5V input, read the 0-5V position signal from the OMP, and step the OMP position up or down until the OMP position matched the input (within a tolerance). The picture shows me testing it with a potentiometer as the input, and in this case it worked perfectly. My plan was to use one of the generic PWM outputs on the Megasquirt, and convert that to a 0-5V signal using a voltage divider and an RC circuit for the input to the Arduino. However, when I tested it on the Megasquirt, I had trouble getting a consistent analog signal out of my RC circuit... Anyway, after that I just decided to premix and never ended up investigating it deeper.

-Tach output: The FD get's its signal for the stock tach from the ECU. Luckily, this is easy with a Megasquirt. My instructions from another thread: "Find the yellow plugs for the stock ECU. On the smaller of these two yellow plugs, there is a yellow wire with a blue stripe. Cut this wire and connect it to the tacho out wire from the Megasquirt. Proper "Tacho output" settings in Tunerstudio are "Fixed" and "Normal" speed. Your tach should work now."

- Mounting: I wanted to mount the ECU in the stock location, so I made a bracket:

https://cimg8.ibsrv.net/gimg/www.rx7...fdf3f03ab4.jpg

https://cimg1.ibsrv.net/gimg/www.rx7...ec0e651266.jpg

https://cimg5.ibsrv.net/gimg/www.rx7...036a9f23ec.jpg

- Other notes I made in another thread:
For coils, I wanted something powerful and reliable, so I chose the IGN-1A Mercury Marine coils (aka AEM smart coils without a cool sticker). I had already removed the A/C, so I was left with a nice hole under the PS pump and 4 M8 threaded holes, all in-plane. It made for a nice mounting location for the coils (again, with some more bracketry and some aluminum spacers I made on my lathe):

https://cimg4.ibsrv.net/gimg/www.rx7...b585ed4af5.jpg

https://cimg1.ibsrv.net/gimg/www.rx7...61e236be4a.jpg

https://cimg3.ibsrv.net/gimg/www.rx7...d767c83083.jpg

The plug wires aren't pictured, but I ended up ordering custom length wires from Sakebomb (I considered making my own from a "kit," but it's not really any more expensive to just have Sakebomb do the work).

More electronics pictures:

FFE 36-1 trigger wheel
https://cimg1.ibsrv.net/gimg/www.rx7...282ff76476.jpg

Hall sensor. I chose hall over VR just for piece of mind. I've always been paranoid about running a weak electrical signal through an engine bay, even if the wire is shielded. Hall doesn't even need a shielded wire as it just outputs a 0-5 or 0-12V square wave.
https://cimg5.ibsrv.net/gimg/www.rx7...0dcff256eb.jpg

In the middle of wiring:
https://cimg8.ibsrv.net/gimg/www.rx7...94ad4f240e.jpg

My MSDroid interface:
https://www.rx7club.com/attachments/...hy-not-msd.png

Fuel system

I wanted new EV14-style fuel injectors rather than the old pintle-style fuel injectors. The EV14's have a much finer spray pattern and would allow me to get rid of the plastic diffusers in the intake. ID injectors are EV14's, but I've always felt like they're overpriced to pay for their polished website and "tech" articles that seem to have been written by a Gawker blogger... The EV14 injectors that I ended up going with were the FIC fuel injectors sold by IRPerformance. I went with 4x 1000cc injectors. I was expecting to get a datasheet with the injectors, but all I got was a label on the bag that said "1000cc/min @ 43.5 psi. Matched"... No dead time info or actual flow rates. Not a huge deal though, I made an educated guess at the dead time and trusted that the injectors were actually matched. If I bought injectors over again, I would probably just suck it up and by ID's so I could have a datasheet.

For fuel rails, I bought FFE primary and secondard rails. Worth every penny. I used an Aeromotive 13129 fuel pressure regulator, which works really well screwed onto the end of the secondary rail on the firewall side (I eventually swapped the mechanical fuel pressure gauge for a pressure transducer that I log to the Megasquirt. And yeah, at the time I didn't realize that the vacuum port I used was just a pass-through. After some confusion I switched it to an actual vacuum port).
https://cimg6.ibsrv.net/gimg/www.rx7...006b5a8d2d.jpg

I linked the rails with a 3/8 SS hardline:
https://cimg5.ibsrv.net/gimg/www.rx7...93901bc5d9.jpg

https://cimg7.ibsrv.net/gimg/www.rx7...9427fb4155.jpg

The rest of the connections are just 5/16 hose (-6 ORB hose barbs off of the fuel rails). I did some quick hand calcs and saw no reason to use larger fuel line, so I kept it nice and simple. I added a Subaru pulsation damper just before the primary fuel rail for good measure. Fuel pump is a Supra TT.

That's all for now. The next update will get into details about the custom V-mount.

i'm curious how the ms3 and efr7670 work out for you.

and yes, i did have boost control issues with my 8374 and the turblown shorty manifold, but my engine had a medium-large street port, and elliot posted a few times some not-so-scientific justification for the lack of boost control.

that was two FDs ago, (efr setup was on my red car), and on my last FD (white car) i had an ebay manifold/wastegate and a $650 borg warner s360, which worked *surprisingly well* for costing $1200 total including the downpipe. it was literally 1/3rd the price of my EFR setup, and it did very well on the track.

i'm officially "against" having a $2200 turbo on a mostly-track FD now, as the reality is you will most likely throw an apex tip through the turbine wheel at some point (obviously different situations and tuning will have different results -- and i never blew an engine in 7+ track days including 95*+ heat). that's undoubtedly a great turbo, but if i make another turbo setup for track-only (ish) i will go with another "budget" borg warner s series turbo and chinese manifold/wastegate.

i'm also interested in this project and your choice of turbo because boost response isn't crucial on the track like it is at a different type of driving (like autocross for example). i presume it will perform very well and be very responsive, and it should be a great all-around car.

i like what's going on here. post some more updates! :)

Just a thought on your towing set-up, looks like too much weight on the front of the trailer unweighting the front of your truck.

We use a similar style towing vehicle - I can think of an instance going to a race meeting, downhill in the rain, with fronts locking and trying to modulate, rear end of another car getting larger and not much retardation happening...very vigilant on load position now!

The UHaul trailers have the stops and straps at the front of the trailer. Unless you were to strap up further back with your own straps, you are limited to what you can do.

Now, if you do strap the vehicle down on a UHaul trailer in a different manner than they provision for, they will not cover you under their insurance if something happens. The trailers are designed to have the front tires in the chocks and the web over the top, even if the design is not the best.

I used to work for UHaul and there were instances of corporate denying insurance claims for a variety of reasons. #1 was not hooking up the trailer in the way UHaul wanted you to. Worst case, the trailer hops off the truck or a wreck happens and they see the car in the middle of the trailer, they will instantly deny any liability.

You are better off buying your own trailer.

Yeah, you nailed it. The Uhaul was just to get it back to the house when I bought the car. Since the FD is fairly short, and the wheel straps are at the front of the trailer; the car was too far forward and I ended up with way too much tongue weight. It pitched the truck around like crazy with every bump. If I start towing in the future, yeah, I'll just buy my own trailer.

Thanks! Your build threads were the main thing that inspired me to get an FD and build it like this. Too many build threads on this forum focus on pointless dress-up parts and the owner doesn't end up using the car for anything exciting... Yours were like an R&D project; build it, take it to the track, improve what was lacking, repeat. It was awesome (and, spoiler alert, that's where my thread is going). Time for you to get another FD, btw.

kind words. thank you!

i will have my 5th FD this weekend, as a matter of fact :)

V-mount

Alright, cooling. I wanted a v-mount for obvious reasons; this being a track car, I needed to stuff as much heat exchanger surface area as I could between the grille opening and the engine. I'm not the type of person who pays the FD-tax for a pre-made v-mount kit, so I decided to make my own. CXracing's parts are too cheap to pass up, so I ordered a radiator from there (rad-016) with the intention of returning it if the quality wasn't up to my standards. Surprisingly, it's a really well-built radiator -- all of the welds are super nice, the fins are nice, and the pipe attachments are in a good place for the FD. After that, I went ahead and ordered one of their intercoolers too (IC0047-A).

The radiator came with two fans and "zip tie" mounts that go through the fins. I didn't like that mounting solution, so I opted to build a fan shroud. I used 3/4 angle aluminum to build a frame around the back of the radiator, welded aluminum sheet metal to the frame, cut out two fan-sized circles in the sheet metal, and bolted the fans to the shroud:

https://cimg6.ibsrv.net/gimg/www.rx7...1b837ca41e.jpg

Made some brackets to mount the front of the radiator to the frame rails (this is the driver's side):

https://cimg0.ibsrv.net/gimg/www.rx7...46abe3d8cc.jpg

I was too lazy to re-locate the battery to the rear of the car, so I decided to keep it up front. I've used a Deka ETX14 battery for my RX8, and it's been totally dependable for several years, despite its size. I used the same battery on the FD; I made a mount to located it on the driver's side of the engine bay, just beside the radiator:

https://cimg7.ibsrv.net/gimg/www.rx7...c54d603255.jpg

https://cimg3.ibsrv.net/gimg/www.rx7...46e7e3aa4d.jpg

In the process of adding the radiator, I had to ditch the stock s-shaped power steering cooler. I replaced it with a Derale 13249 heat sink that I mounted behind the radiator on a piece of tubing that goes between the frame-rails:

https://cimg4.ibsrv.net/gimg/www.rx7...43796f1fd4.jpg

I wanted to make an undertray that would do two things: 1. Protect the bottom edge of the radiator from bottoming and 2. Keep debris from bouncing up into the accessory belt area. To do this, I ended up making a "front subframe" of sorts:

https://cimg3.ibsrv.net/gimg/www.rx7...c682cd62b7.jpg

https://cimg6.ibsrv.net/gimg/www.rx7...098134b2d7.jpg

For the "anti-debris" part of the undertray: I found a cheap carbon fiber panel on ebay and riveted that to the undertray frame:

https://cimg0.ibsrv.net/gimg/www.rx7...5c109d3341.jpg

https://cimg5.ibsrv.net/gimg/www.rx7...6014d3f073.jpg

Also, you probably noticed a couple of pictures ago, the stock oil cooler hard lines had to go to make room for the radiator. I replaced those with nylon braided AN hose. I zip-tied those hoses to my PS cooler mount:

https://cimg4.ibsrv.net/gimg/www.rx7...39344d7270.jpg

So, since it's hard to show everything in one picture, the V-mount setup looks like this diagram:

https://cimg6.ibsrv.net/gimg/www.rx7...338239a012.jpg

With my new discovery that carbon fiber sheet is pretty cheap on ebay, I made some ducting for the v-mount out of carbon fiber:

https://cimg7.ibsrv.net/gimg/www.rx7...7f7c17d3b5.jpg

I cut out the flat pieces, epoxied thin angle aluminum to the inside edges, and gaffer taped the outside edges just to keep it from fraying. It resulted in a really strong and lightweight piece. Also, CF is soft enough on the edges that it doesn't rattle like sheet metal would.

https://cimg7.ibsrv.net/gimg/www.rx7...362aacc88c.jpg

I also made a small catch can that mounts to the rear of the intercooler:

https://cimg4.ibsrv.net/gimg/www.rx7...fe000099aa.jpg

And a picture of everything together:

https://cimg8.ibsrv.net/gimg/www.rx7...b933f3b417.jpg

I had a temporary hot air intake on the car at this point. Eventual plans were to run it down into the passenger side duct.

So, it's coming together! My next update will go into details about my initial testing.

This is a great build, and I appreciate that you can do so much yourself. Can't want to see more results.

You might want to rethink that "hot air" intake... here's what I did with a similar set-up. I got two cast tight-radius 4" elbows (connected by shortened 4" silicone couplers so they touch each other and the turbo to maximize space) to get the filter into the opening next to the IC... it's pretty open, eventually I'll make a cold air box, but with your fab skills this would be easy and more effective.

Also, if you're tracking I wouldn't waste the passenger side bumper opening (if that's what you're thinking), get a 2nd oil cooler in there.

ALSO: Maybe mods can figure out why every photo I attach from my phone comes in upside down?!

Yeah, the hot air intake wasn't ideal, but it was just temporary at this point. My eventual plans were to run it down into V through my carbon fiber duct on the passenger side of the intercooler. I'll get into that in a future update. And I already have a second oil cooler since it's an R1.

Nice engine bay btw!

Mind pointing us in the direction of these carbon fiber sheets? The only thing I found on eBay (in great quantity) was small 200mm x 300mm sheet. I assume you were using something a bit larger than that.

Also where did you source the CXRacing parts? eBay as well?

Thanks!! looking great BTW

Search "carbon fiber cutoff panel" on ebay. There's a company called elitemotoring that sells defective or excess panels from their production runs. .04" thickness is pretty sturdy and can be cut with sheet metal cutters.

And I bought the CXracing parts directly from their website.

awesome, thanks!

I see you've copied my former FD's (red car) intercooler setup and mount. i'm flattered.
i like what you've done here, EDIT: I see you're planning to run the CAI through your new CF box. I am interested to see how you do it!

great work with that front end subframe! it looks very nice.

here are some more pics of what I did, for reference. this setup worked amazingly, btw.

Attachment 752525

Attachment 752526

Attachment 752527

Attachment 752528

Initial Testing

With the physical build mostly together, it was time to start it up and and do some tuning. For starters I made a timing map based on some historical knowledge from the forums. My map is a combination of the arghx timing maps from the Rtek subforum and the standard PowerFC timing map. I made some educated guesses, used bits and pieces from those two maps, and tried to make the map as smooth as possible. This is the result:

https://cimg3.ibsrv.net/gimg/www.rx7...2a61a2667c.jpg

For fuel; I used some more of arghx's advice from the PowerFC forums as a starting point for my target AFR map. Also, much of the low-load stuff came from AaronCake's Megasquirt write-up:

https://cimg9.ibsrv.net/gimg/www.rx7...91153117f3.jpg

Then, I set the VE table to arbitrary values that I know to be too rich, turned on VE analyze (autotune), and started driving the car around. I would drive up and down my road and try to drive in as many different load/rpm situations as possible, take a quick look at the changes VE analyze wanted to make to the VE map, apply the changes, and repeat. Within ~30 minutes, the car was running almost perfectly. The nice thing about Megasquirt is; once the VE map is dialed in, changes to the AFR can be made directly from the target AFR table. So, for example, if I'm at the track and decide I want to lower my WOT AFR from 11.5 to 11.0; I just go to the target AFR table and change the 175 kPa row from 11.5 to 11.0. No other re-tuning necessary.

To continue fine tuning, I took the car to a local road (hwy 151) to generate more high-load/high-rpm data points:

https://cimg5.ibsrv.net/gimg/www.rx7...77869235a1.jpg

Sadly, I never took any video of this part. But don't worry, plenty of track videos to come in future updates.

So at this point, the car was running really well tuning-wise. However, I had a few lingering hardware issues I needed to fix; I had an oil pan leak, the clutch slipped at anything beyond 10 psi (I could run 8psi all the way to redline with the 7670, for reference), I wanted some stiffer motor mounts, and I wouldn't mind refreshing the 22 year-old engine before doing some hard track time. So I decided to go ahead and pull the engine to fix all of those issues.

https://cimg5.ibsrv.net/gimg/www.rx7...42a4e757e4.jpg

Still had the stock clutch with plenty of life left, just not enough for my power-level:
https://cimg6.ibsrv.net/gimg/www.rx7...67b011acb5.jpg

Exedy stage 1 HD:
https://cimg7.ibsrv.net/gimg/www.rx7...aae47a51b0.jpg

I replaced all of the soft seals, used RA super seals and new apex seal springs. The everything in this engine is super nice being a 35k mile engine, but it was nice to go through it to confirm that and put some fresh rubber in it. The RA super seals were just for a bit more safety on the track, as I trust the 2pc seals more than the stock 3pc seals. I didn't do any porting in the interest of keeping as much torque as possible. Also, my 7670 turbo is a better match for a non-ported engine.

https://cimg0.ibsrv.net/gimg/www.rx7...6830b79651.jpg

Banzai oil pan brace and Himni engine mounts. I used "the right stuff" to seal that pan and was really satisfied with it, seems much more durable than RTV.
https://cimg8.ibsrv.net/gimg/www.rx7...da72177a67.jpg

https://cimg1.ibsrv.net/gimg/www.rx7...0f7a88d674.jpg

So, the car was tuned really well and had a fresh engine rebuild. A little bit of suspension work on the next update and we'll get into some track updates!

Be careful with the right stuff. Their newer formula isn't able to come in contact with gasoline, which gets in our oil pans sadly. I learned the hard way, it only lasted 1500 miles for me. Read your tube, it will say if it's the newer formula

I am a huge fan of Toyota FIPG black. That stuff is resistant to damn near everything.

I have put FIPG on a oil pan, stuck it on the crankcase, let it hold there by stickiness, then installed the bolts. It's that good.

Do you have any links to more information about NASA TT1-3? I could only find general information on what time trial is and then the rules for for how TT 1-3 are classified.

Also correct me if I'm wrong, but the TT1-3 classes are pretty much open in terms of aero/braking/engine mods, it just really comes down to tire size and power to weight right? I read the rules, but they don't even mention any of those things.

Suspension

Alright, I’ve been slacking on this thread lately, so it’s time to get this up to date. Reading Smokey the Talon’s thread inspired me to get back to it. This update takes place in January 2016.

https://cimg9.ibsrv.net/gimg/www.rx7...2832de5d36.jpg

I decided to ditch the CX Racing coilovers in favor of some quality suspension before my first track day. I’ve always been an Ohlins fan, and I live 30 minutes from their US headquarters, so I decided to go with a set of Ohlins DFV’s.

While setting the alignment with the new suspension, this is the best I could manage on the front (initially):
Front camber: -1.8 degrees
Front caster: 4.5 degrees

The caster was much lower than I wanted/expected. Upon further investigation, I found that the front upper control arm bushings had slipped, reducing the caster.

https://cimg4.ibsrv.net/gimg/www.rx7...24c363725a.jpg

As you can see above, there is a gap between the rubber “flange” on the bushing and the edge of the UCA. Braking forces had caused the bushings to slip and the UCA had moved further forward than it should have been; reducing caster. It’s really a crappy design – there’s no positive axial retention of the bushing, aside from the friction of the bushing within the bore.

So with that, I decided to make my own delrin offset upper control arm bushings. I wanted more caster obviously, a bit more camber, and a more reliable retention method for the UCA bushings. The lower compliance of a delrin bushing would be a welcome plus too.

I took most of the bushing design information from this thread: https://www.rx7club.com/suspension-w...shings-941277/

https://cimg2.ibsrv.net/gimg/www.rx7...8655fc215d.jpg

https://cimg8.ibsrv.net/gimg/www.rx7...96ce7bca86.jpg

https://cimg3.ibsrv.net/gimg/www.rx7...034106a62d.jpg

The bushing axes are offset to move the UCA inboard to allow for more camber. There is also a slight offset rearward to increase caster.

https://cimg9.ibsrv.net/gimg/www.rx7...49e130ff7a.jpg

I added my own twist on the design and turned some o-ring grooves into the aluminum sleeve. This allowed me to seal the edges of the sleeve to keep the grease in and keep water/dirt out.

So, the alignment I ended up at with the offset UCA bushings:
Front camber: -2.5 degrees
Front caster: 6.75 degrees
Front toe: 0
Rear camber: -2 degrees
Rear toe: 2mm in per side

https://cimg1.ibsrv.net/gimg/www.rx7...0e57702b8c.jpg

A friend and I found a set of Intercomp race scales and went in 50/50 on them. I made a set of ramps in order to drive up on them.

https://cimg5.ibsrv.net/gimg/www.rx7...180f320d1f.jpg

https://cimg8.ibsrv.net/gimg/www.rx7...7651ce4626.jpg

The car weighs 2815 lbs with me in it (I weigh 190 lbs) and ~5 gallons of gas. Weight distribution is 50.4% front.

More Suspension

I like to develop a good understanding of my cars, especially when it comes to the suspension system. There’s nothing wrong with trial and error, but track days are expensive, so I would like to limit the “errors” in the suspension tuning process to maximize my enjoyment of the car at each track day.

https://cimg1.ibsrv.net/gimg/www.rx7...9cf8b8cd01.jpg

I decided to reverse engineer the suspension kinematics to build a suspension model and quantify some key parameters. I’ve gone through this same process on my RX-8 a couple of years ago, and it’s been helpful to understand the camber curves, roll center heights, and spring/damper motion ratios when choosing alignment settings and spring/ARB rates. To measure the suspension points, I used a plumb bob to mark the X and Y position of each suspension point on my garage floor. I pinched the string of the plumb bob and held it up to a ruler to find the Z position of each point.

https://cimg2.ibsrv.net/gimg/www.rx7...8ccbb619ed.png

This shows the right-side suspension of the FD in Solidworks, along with the ARBs. My Solidworks license is up, so I had to use eDrawings for this screenshot, that’s why the aliasing is so bad.

https://cimg4.ibsrv.net/gimg/www.rx7...d04ae4be42.jpg

Above is my RX-8 suspension model. I went through a little more effort to make this one look pretty.

I also bought a program called Dynatune XL. It’s an Excel-based suspension kinematics solver, it’s much cheaper than similar programs and serves my needs nicely.

https://cimg9.ibsrv.net/gimg/www.rx7...b58ef981ed.png

Some specs:
https://cimg0.ibsrv.net/gimg/www.rx7...243c889595.jpg

So, what are the key takeaways?

-The front roll center height is really high compared to the rear, most cars are the opposite. This means that some roll stiffness is baked into the suspension kinematics and the front can be run a little bit softer. It also means the front of the car is really quick to respond (I think we’ve all felt this) since the high roll center generates immediate jacking force upon turn in.

-The front camber curve is ~twice as aggressive as the rear. So, be careful about running less static camber in the rear, because the dynamic camber in the rear will be a lot less. It’s probably beneficial to run close to equal camber front and rear (and it could be better to run more rear camber than front).

-The rear motion ratio is higher than the front. So, equal front and rear spring rates mean that the rear wheel rates are ~28% stiffer than the front.

Completely awesome you're able to get and share that suspension info with the community. Thank you for that. Keep up the good work. So did you get on track in 2016 at all?

Final Setup Tweaks

The Ohlins came with 11kg/mm springs on both the front and rear. As I mentioned earlier, the rear motion ratio on the FD is higher than the front; so the standard Ohlins spring rates would result in a higher rear stiffness. My go-to calculation for a car setup (at least for spring & ARB rates) has always been the Tire Lateral Load Transfer Distribution (TLLTD). Formulas for this calculation can be found in Race Car Vehicle Dynamics by Milliken – I highly recommend this book for anyone interested in car setup or vehicle dynamics.

Overview of TLLTD:

- The axle which has more load transfer will have less grip (tires reduce their friction coefficient, or “saturate,” when heavily loaded)

- As an example, a car that’s really stiff in the front and has a large front ARB compared to the rear will have a TLLTD that is biased toward the front. Most RWD production cars have a TLLTD of ~60% front, so the car is biased toward understeer.

- If the TLLTD is 50% front, the car will be “neutral” on the skidpad

- For rear-wheel drive cars, we need some extra grip “left-over” in the rear to accelerate the car out of the corner. I usually shoot for a TLLTD of 55-60% front to make the car feel stable.

I’ve been developing this car setup spreadsheet since my FSAE days:

https://cimg4.ibsrv.net/gimg/www.rx7...1585e1e2f9.jpg

So with the standard 11kg/mm F&R springs on the Ohlins and the stock ARBs, the TLLTD comes to 51.8% front and a roll stiffness of 1.88 degrees per g. That’s low of a TLLTD for my tastes, that setup would have trouble accelerating out of corners without breaking rear traction.

I had some 13 kg/mm springs laying around that I swapped onto the front coilovers. That brought the TLLTD to 54.4% front, which was much closer to my 55-60% target. Roll stiffness ended up at 1.77 degrees per g.

So, confident in my setup and ready to hit the track, the car got its final dose of preparation; a bath.

https://cimg4.ibsrv.net/gimg/www.rx7...6584ea1912.jpg

I put some new tires on the RPF1's: Hankook R-S3 - 245/40-17 front, 255/40-17 rear. I also installed new brake pads, Carbotech XP-10 fronts and XP-8 rears.

March 2016 – VIR Trackday with Chin Motorsports

My first trackday this year was at a track that I had never driven before (VIR North & Full), with a group I had never run with before (Chin), in a car I had never tracked before (FD). Sounds like a great plan! The first day of the weekend was on VIR North, and the second day was on VIR full. I ran in the intermediate group – I just had to get a quick check ride in the first session and I was all set.

The car made the 3hr trip to Virginia without any issues:

https://cimg9.ibsrv.net/gimg/www.rx7...4e88f1eeb8.jpg

https://cimg2.ibsrv.net/gimg/www.rx7...4c3fa0dc3c.jpg

For day 1, I ran on my lowest boost setting (8-9psi) so I could get used to the car and make sure everything stayed in once piece. Within the first session I could already tell how much more fun this car was going to be compared to my RX-8. It’s much more lively, easy to toss around, actually has some torque to play with, and it turns in much quicker. After the first session, I realized that I had filled my small catch can halfway (welcome to trackdays with an FD). I emptied it and added a second catch can (Dasani) for some more capacity.

https://cimg3.ibsrv.net/gimg/www.rx7...da582736b5.jpg

The car held up well for the entire first day, I just had to empty the catch can(s) periodically. Also, I learned that I couldn’t run the car below half a tank of fuel or it would fuel starve. VIR has a long left-turning section (NASCAR & left hook) that sloshes fuel away from the pickup.

That evening, I had a chance to look over my data logs to make sure things were behaving as I expected. Everything looked great; temperatures were cool, AFR was safe and consistent, fuel pressure never dipped, and nothing looked unusual.

I neglected to turn on “spot meter” for this video, so excuse the horrible exposure:


Stats from VIR North
Lap time: 1:44.7
Boost map: 8-9 psi
Weather: High 56F
Max coolant temp: 188F
Max oil temp: 205F
Max IAT: 102F
Top speed: 132 mph

Day 2 was on VIR full, and I decided to turn the boost up to 12 psi based on the success of the previous day.

Anybody who hasn’t driven VIR and lives close enough to do so… do it. VIR is such an awesome track to drive, the snake and the uphill esses are some of the best sections of racetrack in the country.

After during the 2nd session of the day, my brakes started to develop a huge vibration that I could feel through the pedal and steering wheel. The full course configuration added a much faster braking zone to the track, so I think I was just overheating the XP-10 brakes on the front and either warping the rotors or transferring excess brake pad material to the rotors. The vibration got so bad that I could hardly hold on to the steering wheel in the braking zones, and the pads would knock back on the straights to the point that I would have to double pump the brakes at the beginning of braking zones. I spent the rest of the day taking it easier than I would have wanted in the braking zones.

Otherwise the car ran fairly well. The setup was pretty neutral, some understeer, some oversteer, but it was easy to manage. I was getting some bottoming on the left-front through hog pen, and it was starting to wear through the inner fender liner and had chunked the tire in a few places. Also, I think my steering rack bushings were allowing the rack to move around a bit, because the steering wheel would be off center on some straights on the track.

I was much more agressive on day 2, except for in the braking zones. Also I fixed the settings on my Gopro, so this video is much better to watch:


Stats from VIR Full
Lap time: 2:12.0
Boost map: 12 psi
Weather: High 57F
Max coolant temp: 189F
Max oil temp: 205F
Max IAT: 117F
Top speed: 143 mph

https://cimg7.ibsrv.net/gimg/www.rx7...a8c09297ca.png

https://cimg2.ibsrv.net/gimg/www.rx7...6642c0eb81.png

https://cimg7.ibsrv.net/gimg/www.rx7...b84a644de5.jpg

Overall it was a successful first trackday for the car. Tons of fun and only a few issues that will be easy enough to fix in the future. I'm sure the car has a 2:10 in it at VIR with working brakes and more seat time.

To-fix list:
- Brakes: The overheating and vibration were the biggest issues I needed to deal with. Also, I went through an entire set of front pads in the weekend... brand new on Saturday morning, 1mm of pad thickness Sunday evening.
- Bottoming: Rubbing the inner fender liner was getting annoying
- Catch can: Yeah... I needed a bigger catch can

Awesome update!

that 7670 response sounds as good as the primary turbo on the stock twins... crazy.

with that wheel setup did you have to roll or pull and fenders ? it looks good. not too aggressive and mild !! +45 front and 38rear ?

Man... Kind of sad to hear the Carbotech's wearing down that quickly. They're pretty pricey pads and I was thinking of going to those over the HP+ on my rig too.

nice work!