MR2 Owners Club Forum banner

SW20 MR2 for STS Take Three!

5K views 33 replies 6 participants last post by  gigglehurtz 
#1 ·
Figured I might as well start a thread to record the progress.

The car:
'93 MR2 slicktop, PS, no ABS. Known as "Gumby" in some circles. Picture below of it in E-Street trim at the 2019 Finger Lakes Tour.

68841


The plan:
Alex Wilhelm housings with custom valved inserts. More details to come, when they're available.
Megasquirt DIYPNP, with a custom break-out-board.
RacerX intake box with additional piping from the Vibrant catalog.

Currently I'm working on putting together the Megasquirt. Thankfully, the DIYPNP motherboard has everything necessary to run the 5S motors, but it took some research to be sure was going to work.
As far as I know, the 93+ ECUs for the 5S have a different PN, and have a different connector. This makes them incompatible with the 76 Denso kit that DIYAT sells. Fortunately, it seems that the connector for the 93+ connectors is still source-able, if you know where to look. Below, a picture of the progress so far.
68840
 
See less See more
2
#2 ·
Oh man, that's pretty exciting to see actually. I need to get on this Megasquirt plan and see if it works any better than the Greddy. Of course, I should maybe try again to get the Greddy working. Anyway, glad to see you're giving it a shot though and that Gumby still exists! (I recognize those wheels...)
 
#3 ·
Thanks! I'm hoping that the MS can squeeze a few more lbs-ft more torque out of the 5S. Fortunately, I know of a really good tuner local to me that has done a lot of Megasquirt stuff that should be able to get the most of the 5S. I've more or less accepted that the 5S isn't going to be making 95% of peak torque at the rev limit.
However, with the 25" diameter tires on the rear, it should put the 30-50 mph range in the 3800-5200 rpm range, if I recall correctly.

I've put in a rule request/clarification to see if it's possible to get allowances for Federal cars to use Cali cat locations. I've got a header plan for that too... I'll put a picture of that up when I have it in hand. It's pretty much the ideal design (by the Book) without asking Burns to draw something up from scratch. I'm hoping to improve peak torque, and if nothing else, match torque-to-weight ratio with the Miata/Civic (optimistic, but possibly achievable; I'm prepared to be disappointed).

I refreshed the front brakes last year with OE like pads, which are surprisingly really good (street/autox pads, not race pads), so I'm going to stick with that up front and put the Raybestos R-4S on the rear to hopefully maintain good trail-braking. I think the rear brakes are Hawk HPS now, which I never really liked on my Mazda3, since it felt like they needed to be heated or bedded each time before the first run. I recall a run at a local event where I forgot to take that into account, and there wasn't a braking point until a 180 sweeper which required slowing from 50 ish to 30 ish. I stood on the pedal and barely made the corner ?

Current preliminary suspension/wheels/tire setup (some money committed, some theoretical):
225/50/15 Yokos on Rota Slipstreams (5x114.3, +37). Hoping for no spacers up front. Option for 225/45/15 BFG scrubs from last year on the same wheels.
225/45/17 Yokos on heavy Enkeis. I bought some huge rear spacers to bring the offset to +20. Option for 225/45/17 new BFGs on the same wheels.

The suspension (plan) is kinda different from what everyone else has done.
  • 400/600 lbs/in F/R spring rates (nothing terribly unusual here)
  • Going with a TRD front bar and stock rear bar (I think the math checks out. The first runs will be very entertaining)
    • Basically, I'm going to try and keep the front at near stock ride height, and get the rear down in the 1.5" from stock on 23" diameter tires. So, it'll be around 0.5" lower with the 25" diameter tires.
  • Wilhelm housing and camber plates
  • I'm only replacing the inner LCA bushing with poly, and I impulse bought some replacement rubber bushings for the rear radius rod to LCA. I'm going with SuperPro bushings (I guess alllllllll those "influencers" have done their job). My plan to replace the front radius rods with new OE arms only partially succeeded; I only have the LH side. Apparently, Japan doesn't have any more.
  • Custom single-adjustable Penske inserts to fit the Wilhelm housings. I should have that in hand fairly soon, I think. Once the charges hit the card, I should know for certain!
  • Initial alignment settings are up in the air, and are likely to change.
 
#6 ·
So, you decided on a 3s header after all? Looks good! I was wondering whether that would provide some better options. Be interesting to see how the cat welding all turns out and what response you get on the Cali cat location.

Looking forward to seeing how the MS does when you get it into the car. (y)
 
#7 ·
Sloooooow rolling process. I sent my valve cover off to powder coat, and it's not quite done yet. The weather also hasn't quite broken yet. I need to finish the intake plumbing, and fab a bracket to hold the box in place. I also need to fab a bracket to hold the MS in place. So, instead, I'll post a picture of the wrapped header. ?
69551
 
#8 · (Edited)
The weather is predicting another cold front to drop in next week sometime 😕

However, I did confirm three important parameters for Megasquirt configuration on 93-95 5S-FE engines:
  1. The FSM is correct, the G signal pulses once per cam rotation (similar to the 3S-GTE distributor)
  2. The NE signal pulses 4 times for every cam rotation (very much not like the 3S-GTE dizzy)
  3. The dwell time for the coil is 10 ms Need to re-measure. EDIT: 3.0 ms looks correct.
I think those are the most important signals to get correct the first time! The fueling isn't quite as critical to get perfect right away. So long as the parameters are set correctly, the Megasquirt should be able to correctly fuel the car.

Also, I saw a letter in the May Fastrack about relocating the O2 sensor. Turns out the rules don't explicitly allow O2 sensor extension wires (though I know of many cars that are set up that way), so I'll have to write in about that so that way the cat relocation rule I wrote in about (that hasn't been responded to yet) can be successfully addressed.

EDIT:

I also "completed" the intake install. I have everything installed, but I may get some more clamps for the junctions that don't have one. I don't particularly like the tight radius bend off the air box, but it is what it is -- there isn't really a good way to add open radius bends and have it go into the throttle body with a 90 deg bend. I haven't really noticed any more or less power, but I also haven't really driven the car very much recently.

70151
 
#9 ·
I've been particularly negligent on working on the MR2 recently, considering the current situation.
I've driven the car with the intake on it more, and I think it gave a noticeable boost in power above 5K, where the power would fall off a cliff before.
I test fitted the 225/50/15 Yoko on the rear; there's no real clearance issue here, but I was curious to see how it would look. If the car was slammed, the tire would definitely hit the fender :D
I'll get around to trying the 17s on the rear, and installing the massive rear wheel spacers. I wanted to install shorter studs for the bolt-on wheel spacers, so that it shouldn't interfere with the back of the wheel.
I also installed the R4-S on the rear of the car. I think the resultant bias is good with the tire stagger that I'll be using, but I've only played with it on the street.
A Cusco rear strut bar also found its way on to the back of the car. Probably mostly for decoration, but I'll take every bit of chassis stiffness I can get with these old cars.

My shock builder is located just outside NYC, so he's been holed up and I haven't been able to reach him. Hoping to get the ball rolling on that again soon.

70532
 
#10 · (Edited)
Once again, negligence means I'm not working on the MR2 as much as I'd like. However, I'm enjoying the car with the 225 BFGs up front and the 225 Yokos on the back, even if the two tires have a completely different temperature profile. I made some of our local competitors sour, since they were in much more highly prepared cars and getting slaughtered by a car that is essentially an ES car with wider tires and an intake.

I also trialed the TRD front bar instead of the bigger 25 mm bar. The math suggests that the balance would be better, and I think it worked out. I'm hopeful that after I get the car on coilovers, I should be able to get the car balanced perfectly with the TRD front bar and the stock rear bar.

I picked up a cat back exhaust for a Japanese spec car, so that should facilitate putting the cat in the same place as the Cali car. It also makes welding up the cat fairly trivial, since it's just a pair of flanges, the cat itself, a length of straight pipe, and a reducer cone. Downside is, it's probably way heavier than something made from the Vibrant catalog. Upside is, I won't lose my hearing driving to and from events!

Also, I'm pretty confident I need to re-measure the dwell time on the coils. 10 ms is too long for the MS (they claim 3 to 4 ms is typical), so I'll double check and edit the post. I should also be able to guess at an appropriate value for injector duty cycle for starts, so I'll try to measure that and use that as a starting point in the base tune.

I need to figure out a good place to wire in the wideband controller, but I'd like to put it on switched power. I need to review the wiring diagrams so I can find a place to wire that in, since it's a pretty high-draw device.
 
#11 · (Edited)
I re-measured the dwell time for the stock ignitor. Looks like it's around 3.0 ms. The measurement doesn't vary with rpm, so I'm pretty sure the DMM triggered correctly.
71858

Idling at ~1500 rpm, the ECU is hitting the fuel injectors (each pair) every 2.0 ms.
I'm thinking the high idle is a vacuum leak somewhere, but I haven't had much success tracking it down. I replaced the IACV, and checked the connection between the ECU and the IACV, so I think it's something else going wrong. I can see with the DMM that the ECU is commanding the IACV closed, so I'm fairly confident that the ECU is working as intended.
EDIT: The ECU is working, but after running through the checklist in the BGB, I'm 99% certain that it's the IACV driver in the ECU that's not working as it should. Guess I should get the Megasquirt working...
 
#13 ·
Good read through your development ideas. Not sure what your custom MS board is doing. Expect that once you get O2 wired up you will run into MS for live VEAL tunning. It’s handy to have the digital gauges in cab through Shadowdash. Can use a cheapy Chinese droid HU for display and set custom warnings & alarms for parameters such as temps, AFR’s and oil press if you have sensor installed. HU can also datalog info if you want to analyse later. Might be handy to see your VE, knock and AFR’s to help dial in more timing. This would squeeze more hp from the 3S where you want it.

Jim
 
#14 ·
Glad you're enjoying it. The custom board is just to break out the pins on the 54 pin connector. I modeled it off the DIYAT break out board for the 64 pin connector. I'm not doing anything special on that board, but I was considering using it for distributor signal conditioning.

The wideband controller will be wired in through the 15 pin D-sub connector. The ECU is the only good place to grab switched power in the trunk. The class I'm aiming to build the car for doesn't allow for the addition or substitution of sensors for engine control (stock replacement only), so the wideband will just be used to get a base fueling map before switching back to the narrowband. I'll need to get some dyno time for really ragged edge timing advance and fueling.
 
#15 ·
You know what they say about all good things...

Was having some fun with cars today, and looks like the 5S decided that today was the day it was going to knock on heaven's door. I'm not in the position (logistically) to start a V6 swap, so I'll see if I can find a junkyard 5S and swap that in.

Anyone know any gotchas before I pick up any ol' 5S? I know the ideal situation would be to get an engine from an MR2 or a Celica, but I don't have the luxury of time. Looks like Camry longblocks can be swapped, but the balance shafts need to be deleted, and I think Toyota did something different with the distributor? Can the older dizzy be swapped?
 
#16 ·
Well, plenty to talk about here. I got a 5S-FE from an automatic 1993 Camry and started some documentation on the differences between the two. I haven't put anything into final draft form, but I hope it will be helpful for those who might want to follow a similar path next time. I find the changes that Toyota made between the two applications fascinating.

The MR2 engine was original to the car (shame) and spun the #3 bearing. All but one rod had notice-able play in it. My temptation is to save the block and replace the rotating assembly and head for something more fun. Right now the car is in pieces and I need to get the engine ready to drop in.
72762
 
#18 · (Edited)
Thanks, appreciate the offer. I yanked a fully dressed Camry engine from a pick and pull (a whole 178 including core charge) and almost have it ready to drop back into the car, after finding the numerous little differences between the Camry and the MR2/Celica implementation. It was very educational, but also annoying to modify the windage tray, oil pickup, and plug the oil feed for the balance shafts.
 
#25 ·
Time flies.

I didn't sell the car, nor did I die... but I have my MR2 in storage, with plans to get it out some time in May. I'm hoping to have it out sometime in the second half of the year.

Of course, even when I wasn't working on the car, there were things happening. I wrote to the SCCA Street Touring Advisory Committee to see what they thought about changing the catalytic converter rule so that the Cali downstream cat location can legally be used. This means that I can run the 3S header and have the cat just past the rear subframe, like the JDM 3S cars. The stock O2 sensor is moved to just after the last merge (this is the stock location for a 3S car), and I have a bung for the wideband just before the catalytic converter. I'll need to make an extension cable for the stock O2 sensor, but that's no problem.

Set up TL;DR: this setup (wheel+tire package, suspension, header+ECU) should make this car a ProSolo monster, compared to the Hondas and Miatas. This will be especially true if the event is wet.

From Alex W's graphs on toe and camber, I figured that my initial alignment will be:

-4 deg camber and zero toe up front;
-3.5 deg camber and 1/8" total toe in the rear.
400/600 F/R spring rates.
Stock ride height up front, rear 0.25" or 0.5" lower than stock.
TRD front bar, stock rear bar.
205/50/15 up front, 225/50/15 in the rear A052s on 15x7.5 wheels.

The A052 is a pretty beefy tire. The 225 on the 15x7.5 wheel actually is wider than a 225 on a 17x7.5 wheel. The Miata folks have a hard time fitting the 205/50/15 where the 225/45/15 BFG sorta fit. I'm not super worried about the stagger, but it is a concern that the car will be too "safe". Adding some front toe out should help to get rid of the toe induced understeer that will come mid-corner, if necessary. I will be street driving this car, so I don't want an aggressive toe setup for tire wear and stability reasons.

The spring and bar package is actually a bit soft as well. On high grip surfaces, this might be a limitation. I may need to change up my bar package to get the roll under control. On the plus side, the car should work great on low grip surfaces and in the wet.

Ride heights are to maintain good geometric roll stiffness and keep the car from getting into wild toe land. No roll center correction in this class, so have to make due with what's available. The nice thing about this platform is that the CG is already pretty low, so making the car lower without fixing the other issues that come up is chasing small gains. Lowering the rear ride height is based on evidence that suggests that the NSX uses a backward sloping roll axis. This should contribute to good turn-in and mid-corner stability.
 
#26 ·
Ride heights are to maintain good geometric roll stiffness and keep the car from getting into wild toe land. No roll center correction in this class, so have to make due with what's available. The nice thing about this platform is that the CG is already pretty low, so making the car lower without fixing the other issues that come up is chasing small gains. Lowering the rear ride height is based on evidence that suggests that the NSX uses a backward sloping roll axis. This should contribute to good turn-in and mid-corner stability.
This is also something you get with my geometry correction kit (front RCA's thicker than rears), and it seems to work well there. Sounds like a solid starting point for you.
 
#28 ·
Alrighty, first event in the books. Wet event, coilovers installed, no alignment, and completely new Nankang CR-1s on the front. The car suffered from understeer (shocking, I know) and I apparently clapped cones with the rear tires on all my runs. There was a STS CRX running in the same group as me, codriven on Yoks, and they were about 2 sec ahead. The car definitely has potential in these situations, because it can put down power. I was able to launch the car way harder than I thought I could. The engine mods seem to give it an edge in acceleration out of digs (honestly the car feels like it makes some serious torque around 4.5k, it is surprising). I definitely didn't drive the car well either, so being 2s off is a promising enough result for me.

I replaced almost all the ball joints on the front, except for the inner tie rod ends, which also are in need. Coilover install went about as expected. The hole front and rear needed to be opened up for clearance to the adjusters.
Blue Gas Electric blue Circle Machine


With the Rota +37 wheels, the clearance to the coilover in the front is... Tight. I actually needed to use a cammed strut bolt and a Toyota crash bolt to get clearance.
Automotive tire Tread Tire Fender Material property


The only problem with this is that I wasn't able to hit the camber numbers I wanted in the front. I only can get -3.5 deg, which is not bad, but not quite. I need a dry event to know if this limits front grip enough that I actually need to do something about it. I'm thinking next event will be in June sometime, maybe in Tucson or Las Vegas.

Fun fact, with the front camber plates, I still get positive scrub radius 😅 . I think I also lost some steering angle, doesn't seem the wheel wants to go to max lock. I'll need to add a ~5mm wheel spacer anyway if I want to hit my camber number, so hopefully that improves clearance without smashing into other things.

Tire Wheel Automotive tire Plant Tread
 

Attachments

#29 ·
Sounds like a great starting point. Wet events were some of our best because of the extra traction in the MR2. On Contis in the wet the car was unbeatable.

Our fronts were +25 for context, so spacing out with 15's should still be fine. If you find yourself going up in tire diameter, that's when you might start hitting liners or pinch welds, but I think you'll be fine for now in that range.

Penskes look nice, sir. Jealous!
 
#30 ·
Finally, one dry event in the books. I had to throw money at the car to fix a cam seal leak and a dead wheel bearing. Since they were in there, I had them install some extended wheel studs and I got 10 mm spacers. I did a string alignment and got around 4 deg negative camber up front, with plenty of space to the strut if I wanted more camber. I did a small tweak -- I raised the front ride height by adding 0.5" of preload. Not ideal, but it keeps the front off the bumpstops, and I think I'm still safe on spring travel. I might want to go with a 8" long spring in the future for additional travel, but that's a small tweak.

The car sits at 1" lower than the BGB says the stock ride height should be (but I swear the car sits about as low as it did on Konis and stock springs)

The event with Arizona Border Region SCCA at Musselman circuit (basically a really nice kart track) proved to me the setup philosophy was sound: I was able to abuse lift off oversteer, but still put power down off the corner with no drama (and just a hint of understeer), and the car slalomed like nothing I've ever driven. I'm sure I left time out there in some of the faster sweepers, because frankly I just wasn't prepared for the grip level. I found 0.5s to 1s every run, and I only got four runs. I'm sure if I had more runs, I would've found another 0.5s or 1s out there. It basically just does what it's told to do, and it does it so much faster than any front engined car I've ever driven. I'm sure if a more confident/capable driver would've driven it, it could've been top PAX easily.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top