| MIATA BUILD |
|
|
 | |  | May 29, 2011 - More wiring!
I could do it faster, but I'm trying to build a car that will survive a week of stressful days without failing - and if it does, it should fail gracefully and be easy to debug. This means a lot of time looking at how I'm going to do something, seeing if there's a better way besides the initial obvious solution. Sometimes yes, sometimes no.
I wired up the fans yesterday, for example. I needed a good power supply for the second one, as I've learned in the past that these particular fans can pull over 20A each on startup. The usual way is to run a wire with an inline fuse. But there's a spot in the factory fusebox that isn't used on my car - it's for the ABS system. It also happens to be right beside the fuse for the main fan. So I removed the fusebox and pulled the appropriate connector off the bottom. Turns out the wire I needed simply wasn't there, but there was a spot in the connector plug for a spade connector. I pulled one out of an unusued plug and slotted it into place. Now I had a factory-designed 30A power source! I ran that down to a new relay which is triggered off the main fan. It worked out pretty slick. I even used the same wiring colors used by the main fan so that I know that yellow is a fan power wire, for example.
I also identified a fan override wire that will let me kick on the fans manually without waiting for the engine to get up to temp. I have an idea for that one, we'll see if it works out.
I also wired up the driver's side O2 sensor which needed extended wires and tightened up all the exhaust system connections. The car sounds pretty good. The PCM hasn't figured out how to idle yet so the car won't stay running without a bit of attention to the throttle pedal, but that's normal for the GM computer when it's lost power.
The new shocks ands the driveshaft should both arrive on Tuesday, and it's my goal to have the car ready for them when they arrive. It's looking promising.
entry 845 - tags: Wiring, conversion | | |  | May 31, 2011 - Visible progress!
I decided the wiring was in good enough shape to install the dash. The harness is still pretty accessible, but it's more awkward of course. It was quite a wrestling match trying to get the dash back in, it's the first time I've tried that while there was a windshield in the car.
The navigator's footrest is in, the defrost system is hooked up, the switches are wired - it's getting closer. There's no functioning speedometer in the car at the moment but with the rally computer that's not a big problem. I'll address that later.
I did realize that my current fan wiring does have a vulnerability - the right fan is completely dependent on the left fan for a trigger. If something goes wrong with the left one, I'll lose both. I know how to wire this so that both fans are independent, but I'll deal with that later.
entry 846 - tags: interior, conversion, wiring | | |  | May 31, 2011 - The engine bay is looking a bit better!
I was able to sort out an intake system that allows the use of both the hood and the driver's headlight - hopefully. It's not ideal, but it'll do for a start. I've also cleaned up and anchored the wiring. Other than a lack of a hood prop, this is pretty much ready to go.
entry 847 - tags: engine bay, conversion | | |  | May 31, 2011 - It's amazing how many small jobs eat up time.
I have a new set of Corbeau harnesses for the car, as the old ones were expired. So I need to set them up for length. It's not a big piece of work, but an hour or so of fiddling around was an hour I couldn't spend on something else.
entry 848 - tags: safety | | |  | May 31, 2011 - Aero!
The car will be spending more time at high speed than before, so I'm going to do a bit of work to stick it to the ground. I'll be doing a spoiler on the trunk lid, and this air dam will help balance it out. The relatively high ride height of the car will make the air dam less effective than it would be on a track car, but you work with what you have. The part is made of flexible polyurethane so it'll deal well with impacts.
Due to that higher ride height, it will probably be difficult to balance the downforce of a wing. Still, if I can get time, I'd like to test that.
entry 849 - tags: aerodynamics | | |  | June 1, 2011 - The installation of the V8 is going to require the reworking of a few things.
When I first built the car, I installed the rear subframe brace from a 2003. This not only stiffened up the rear of the car, it also provided protection. That wasn't just theoretical, inspection after the race showed how much abuse it had soaked up. Granted, I had some other problems that were causing this, but still.
In order to make room for the V8 exhaust, I had to go to the less beefy version of the brace used from 1999-02. The two forward cross braces are gone because the exhaust pipes hang too low - and that means they're going to be susceptible to damage. I've got some ideas, but it's one of the things I'll have to address.
entry 850 - tags: exhaust, conversion | | |  | June 1, 2011 - I've been working on the car quite a bit over the past couple of weeks, and it's become more intense for the past couple of days.
That's no surprise given the upcoming track test. So I stopped taking pictures as I installed suspension, bled the brakes, installed the seats, bolted on the roof, etc - but check it out! It's together. That means it's time for the first test drive.
It wasn't an epic drive, just a half mile down the road and back. The goal was just to determine that all of the various systems were working. This car has undergone quite a bit of change since it last turned a wheel under its own power. So how was it? Good! Not perfect, as there were a few little problems: a seeping oil fitting, a dragging rear brake, a bit of interference between the steering column and a heat shield, and low power steering fluid. But those are to be expected and were quickly sorted out. There were no real concerns. The car runs well, shifted well and the chassis still feels like the Targa Miata. Woohoo! Now that I know nothing terrible is going on, I'll exercise it a bit more next time. I'm still breaking the engine in, though.
By sheer coincidence, the first drive of the car took place almost exactly four years ago. I then spent an extra year honing the car for the 2008 Targa. I've got a shorter schedule this time, but the car's much more complete.
entry 851 - tags: first drive, v8 | | |  | June 1, 2011 - The test drive was done without a hood.
The current intake system won't fit under it without some trimming! I'm going to try a slightly different tack first. I've got a few ideas...
entry 852 - tags: intake | | |  | June 6, 2011 - A few days of very hard work and no updates.
So let's catch up, shall we?
First, the new intake. This has been a bit of a challenge because the throttle body is 4" and the MAF is 3.5". On the rest of the cars built at FM, we've used a 4" MAF. So I've had to play with a couple of different tubing sizes. The previous iteration dropped down to 3.5" then back up again, but there's no reason for that and it put a big 4" tube in the tightest spot. So I stopped by a muffler shop for a tight 90 degree mandrel bend of 3.5" tubing and assembled this setup.
Yes, the filter is right above the left header. I'll work on that later. Right now, the focus is getting the car to the track on Saturday.
entry 853 - tags: intake, conversion | | |  | June 6, 2011 - Unfortunately, even my "low profile" intake wouldn't fit under the hood.
Then I discovered that, even with no intake at all, the hood wouldn't close. I spotted the power steering reservoir as a problem. But removing that didn't solve it either - the throttle body hit the hood!
After checking with Tyler, the fabricator who builds the V8 cars at FM, I found out this was normal. Some of the hood structure would have to be removed. So I proceeded to make copious amounts of noise. And finally, the hood closed.
entry 854 - tags: hood, intake, conversion | | |  | June 6, 2011 - Lots of final details in the last day before going to the track.
Including some more test driving, and a problem is appearing. The car's running oddly, with a lot of hesitation at times. After some poking around with the wiring, I finally got the OBD-II port to work and discovered that neither of the O2 sensors are providing a real signal. They're flatlined at 0.455v. Turns out, that's the same reading you get if the sensors are disconnected.
Some more poking around at wiring and all the connections appear to be good. Bill and I are doing a bunch of research to see what normal O2 wiring readings should be, we're seeing some things that don't quite make sense. For example, it seems the GM computer feeds around 5v down the "high signal" line and 0.455v less down the "low signal" line. Is that right? I'm still trying to determine yes or no.
entry 855 - tags: wiring, conversion | | |  | June 6, 2011 - Before dropping the car to the ground for the last time, I decided to check the weight.
I'd been hoping for about 2300 lbs wet. Well, not quite. It's closer to 2400 by the time a full load of fuel is taken into account - I was only at around 1/4 tank. Assuming there's room for another 8 gallons of fuel in the car, that's about 50 lbs and it'll all land on the rear wheels. That brings the fore/aft weight distribution to 54:46. By the time I've loaded in a spare and the tools in the trunk, I think the final balance should be pretty good. It's interesting to note that the first time I weighed the car in the initial form, it had a 53.8% front weight bias. But by the time it went into the trailer for the race with a spare tire and tools and the like, that had changed to 50.2% front. So I'm not that far off.
The reason I was hoping it would be 2300 lbs is that I was assuming a 200 lb weight gain over the 2100 lb naturally aspirated form. But when I think back, I think it was closer to 2200 in race weight. There will be a few more changes, but probably nothing drastic.
entry 856 - tags: weight, conversion | | |  | June 6, 2011 - Track test!
I took the car out to High Plains Raceway, which was coincidentally the last track it ran on before coming apart for the big personality change. At that time, it was very nicely sorted with just a bit of turn-in understeer from the new diff and a shock that was low on fluid. The engine was in fine form.
Well, the car looks the same. But the noises coming out are quite a bit different. After my first session, I had to fix a leaking brake caliper. The Wilwood front calipers have a fitting that screws into the caliper, and the bleed screw threads into that. The leak was where the fitting goes into the caliper, and it was a gusher. It took me a while to figure out what happened - I think when Adam was dismantling the car, he loosened the bleed screw to drain the brakes. But the fitting backed off instead. When I re-bled the system, I used a pressure bleeder that only puts 10 psi through the system so I didn't discover the loose fitting. No worries, it was easy enough to tighten up once I found the problem and I hadn't really lost that much fluid.
Unfortunately, the hesitation problem was still pretty bad and the engine was down on power. I'd estimate it at around 230-250 hp at the wheels. Not shabby, but about 150 hp low. I played around with a few things but wasn't able to identify anything with my meager resources at the track. So I worked on sorting the chassis.
In the second session, I decided the car was understeering too much. There was also a clunk from the front end under hard cornering. I softened the front sway bar and found an alignment bolt that was not loose but not tight enough. I also discovered that my throttle pedal was only allowing about 3/4 throttle, something I was sure I'd checked! Adjustment ensued.
The third session didn't show a real difference in power, disappointingly. Once in a while, the car would clear its throat and just leap forward - even at 100 mph, it was enough to bounce my head off the headrest when that happened. But not consistently. But the understeer was almost all gone and the car was mobile again. The clunk was also gone. So that's a success.
I decided to call it a day after that. I still had another three sessions or so to go, but with the engine misbehaving there was no point. The handling was in good shape and I'd confirmed the rest of the car was working as intended. There are still a number of small things to do, such as move the shift lever forward about 1" so it's easier to reach in 2nd and 4th gear. But as long as I get the engine happy, the car's in decent shape.
entry 857 - tags: testing | | |  | June 8, 2011 - Progress.
The O2 sensors have now reported for duty. It's quite possible that one of them was incorrectly diagnosed as wrong, thanks to some confusion on my part between Bank 2 Sensor 1 and Bank 1 Sensor 2 in my logging program. Possible, but not certain! Regardless, I can read them both happily now.
Unfortunately, that hasn't solved the problem. The sensors are displaying what I can feel - the car is going lean for 15-20 seconds, then rich for 15-20 seconds. One interesting thing is that the car still isn't going into closed loop as far as I can tell. I haven't spotted the problem in the datalogs yet, but I'll get there.
entry 858 - tags: conversion | | |  | June 9, 2011 - A-ha!
Bill and I spent some time on the car last night, and determined that the injector duty cycle is changing when the car goes rich. That means it's not a mechanical problem such as a fuel pump, but bad instructions from the engine computer. We also discovered that unplugging the MAF makes the car go rich, which points towards a real problem with the programming.
After discussing this with Steve from V8Roadsters, it seems that our computer was reflashed for the wrong type of MAF. Unplugging it forces the computer to go back to the static tables, but they're set up for sea level and not 5000' of altitude so they're too rich. To add to the fun, the computer was given a custom operating system so our copy of HP Tuners will no longer communicate with it.
I've been put in touch with the programmer who set this thing up, and hopefully we'll have an answer soon. I've got another track day scheduled for Saturday, and I'd love to have the car on the track in full health. I'm not confident it'll be sorted in that time, but at least we know where the problem is!
entry 859 - tags: conversion, programming | | |  | June 13, 2011 - Track test again!
This time, I was at Grand Junction Motor Speedway. It's a very familiar track and I've got a good library of lap times there, so it's a good test on if the car's any faster or not.
I went back and forth with the computer programmer, sending in a number of logs for various drive cycles. The 5000' elevation doesn't actually make a difference, of course, as the computer is running off a MAP sensor with the MAF unplugged. It's just tuned wrong. Since I didn't have the ability to make any changes, there really wasn't anything I could do. The computer is going back to him to be unlocked (again) and hopefully a new tune. I almost nixed the test session on the track because, even with the MAF disconnected and the car running in closed loop, it wasn't completely happy. On the morning of the track day, however, I decided to run it anyhow.
The track is tight, with slow corners connected by short straights. Definitely not a place to sort out high-speed handling although there is one section that's 70 mph in a big sweeper. The biggest problem is that there are very few high-speed corner entries. But it's a well-known venue.
The car was particularly unhappy under braking. Last year, before the engine swap, I'd found the rear brakes simply weren't contributing enough. And they're still not. On my first session, I managed to hit 75 mph in the fastest section and then had to get fairly creative while trying to gather the car up for an off-camber braking zone. Usually, we can't run more than about 70 mph through there! I held it together, but also locked the front wheels in the process. The tires were already slightly flatspotted but after that episode they felt completely square. So they had a tendency to lock very easily. The car was also understeering on the tightest corners, likely due to the clutch-type differential. It was similar to what I'd seen with the OS Giken diff with the four-cylinder.
Still, even with some handling quirks and a driver that really wasn't doing a great job behind the wheel, I managed to hustle the car around faster than ever. Not a lot faster due to the braking problems, but it counts. My fastest time was a 1:03.725 before I flat-spotted the tires, and it got worse from there. My previous best was a 1:03.796. Like I said, not a lot faster! It's interesting to note that in 2007, the first time this car took to the track, I turned a 1:03.733 in the Seven.
As always, the GM diff hooked up extremely well. I could light up the rear coming off a corner, but it took more effort than you'd expect. That might partially be due to the rich air/fuel mixture cutting power, but it's in line with the sort of traction this diff provides.
There's more time in it with round tires and a driver who's settled down a bit. By the end of the day, I knew the tires were toast so I just started playing around with leaving black stripes around the track like the child I am. You have to, right?
I did have one exciting moment. In my third session, I was just coming off the last corner on to the straight when the engine died. A half-second later, I got a big whiff of fuel. I immediately cut the ignition (the fuel pump control in the engine computer was disabled by the tuner, so it's run off the ignition) because I had a feeling I knew what had happened. And I was right. The push-on fitting for the fuel feed had come apart popped off. It happened a couple of times during the build and I suspect it was damaged or defective. It had been good for a while, though. When that happens, the fuel feed sprays 60 psi fuel all over the engine. I was still moving fairly well and right by the pit entrance, so I brought it into the pits and bailed out to open the hood right away. Luckily, no fire. Whew. A couple of hours later, I had a replacement and everything seems to be working as intended. We've used these fittings on a lot of cars and never had a problem.
So, not a bad day. I'm going to change out the spring rates to the expected Targa setup and set up the ride height for the rally. Then I'll start to really fine-tune the handling. I'm also going to pull out the engine computer and send it back to be reflashed. That will let us tune it properly for this engine.
entry 860 - tags: testing, fuel, V8 | | |  | June 13, 2011 - The new wheels are here!
The selection was pretty tough, actually. I'm allowed to use a total of 8 tires during the race, so I need 8 wheels. Given that I'm going to run a 225/45-15 (more on the tire selection later), I wanted something wider than the 15x7 wheels I have now. That's a shame, because I have a full set of 8 of those. Oh well.
So I wanted either a 15x8 or 15x9. At Flyin' Miata, we have a collection of both. I decided to go with the narrower of the two because it'll give me a bit more sidewall flexibility for following the rough surface. That narrowed it down to three basic choices: the TR Motorsport C1, 949Racing 6UL and the new Konig Wideopen. The C1 is the least expensive of the lot, but the heaviest by several pounds. The Konig is a brand new wheel and looks to be a pretty good value, but it's unproven. More importantly, we didn't have any I could borrow so I'd need a full set of 8 of either.
Meanwhile, the 6UL is both the most expensive and the lightest at 12.0 lbs on our scale. It also seems to be pretty strong. But we have a set of 4 of them at FM, so I only needed 4 new ones to get my full complement. That makes them the least expensive option really.
The black is a satin sheen. With polished lips, they'd look pretty killer. Better than other black/polished 6ULs I've seen that had a glossy center. They're not going to stay black, though. The car needs white wheels, and Emilio at 949 says the black is the best base color for paint. I just have to decide if I'm going to use proper car paint or use Duplicolor wheel paint. The advantage of the latter is that it's easy to touch up should the need arise.
I can hear howls of protest now about my painting plans. I'm okay with that.
entry 861 - tags: wheels | | |  | June 15, 2011 - Time for some paint on the wheels.
I decided to paint them with rattlecan paint instead of proper car paint - even though I had all my painting gear out for working on another car. The main reason was that it's easy to patch if it gets chipped. I'm using Duplicolor wheel paint, and I've had good luck with it in the past on three other sets of wheels. It's also just plain easier to deal with.
First, I scuffed them up a bit with a Scotchbrite pad, then wiped them down with wax and grease remover. Emilio at 949Racing had told me that they didn't need to be scuffed, but it can't hurt to give the paint a bit more to adhere to.
Unfortunately, the local auto parts store only had one can of white. It would have been more than enough to do just the faces, but when you're trying to paint the barrels of 8" wide wheels that just doesn't quite do it. One can did this - they're covered, but a little patchy. I'm not too worried about that, the inside of the wheel isn't that visible. I shall obtain another can later and do the front faces. Those will be blinding white.
entry 862 - tags: wheels | | |  | June 17, 2011 - The tires are here!
Long-time readers may remember the decision process for the tires in 2008. I finally went with a 205/50-15 Toyo RA1, and I'm happy with that.
But the rules have changed. Specifically, all tires must have a minimum treadwear rating of 140. That means no more nice sticky R compound tires. This leaves me with what I call the "Azenis class" of tire, the gummy street tires designed for autocross. We are allowed to run 8 tires this time instead of 6. Given the ability of this car to spin the wheels, that may not be a bad thing.
The tire size was fairly easy this time around. With my new power/weight ratio, I want as much tire as I can get. And in the stock tire diameter, that means a 225/45-15. In that size, I have the choice of two tires: The Hankook R-S3 and the Toyo R1R. I'd love to be able to run the Dunlop Star Spec, but it's simply not available in the size.
Asking around, it appears the Toyo is a better tire in the wet and possibly overall. So, 225/45-15 Toyo R1Rs it is! I like to call them "pirate tires". Arrrr!
There actually were a couple of other options out there. The Toyo T1R comes in a 245/35-16, but that's not the right tire for the job. The R1R comes in a 245/35-17 that could work, but it needs a 17x8 wheel at minimum and preferably something wider - which isn't really available. So we're going with the light and tested setup we have.
entry 863 - tags: tires | | |  | June 28, 2011 - Remember the fuel line fitting that popped off?
I did some digging after the last failure on the track and discovered I wasn't the only one who had experienced this. In fact, one of our customers had even had the same experience. There was definitely a problem.
Luckily, I also came across a new design from the same manufacturer. It wasn't even production yet, but a fellow on the LS1Tech forum had a pre-production piece. I spent some time on the phone with Russell fittings and managed to get my hands on the part number for the new design. We returned all the stock of the old parts that were on the shelf at Flyin' Miata and the new ones have just arrived.
That's the new one on the right. As you can see, it screws together instead of relying on a piece of plastic. Much more reliable. Even though the factory uses a plastic clip design, there's something about this particular implementation that just isn't right. Russell makes a big deal about proper installation when you call their tech line, but even then they fail - and the factory ones don't need the same installation procedure they call for. The new one is a vast improvement. This should be the end of surprise fuel baths for the engine.
entry 864 - tags: fuel | | |
|
THE DIARY