IPv6 and Amazon EC2

Back in December, Amazon announced that IPv6 was finally available in the Ohio region (here). On January 25, they announced it had been rolled out worldwide (except China).

If you follow the link in the announcement, you can find all you need to know about enabling IPv6 in your VPC. But the detail that is missing is how to get your EC2 instances to start using IPv6.

That’s what this post is, and it’s insanely simple, but hard to figure out if you’re not already well versed in IPv6 stuff.

So here it is:
In your instance, in /etc/network/interfaces.d, there’s a file called eth0.cfg that looks something like this inside:

# The primary network interface
auto eth0
iface eth0 inet dhcp
  pre-up iptables-restore < /etc/iptables.rules

You need to make it look like this:

# The primary network interface
auto eth0
iface eth0 inet dhcp
  pre-up iptables-restore < /etc/iptables.rules
  
iface eth0 inet6 static
  address <IPv6 Address>
  mask 64

Where the IPv6 address is the address you’ve assigned to the instance in the AWS Console. After modifying this file, reboot the instance. It’ll put that static address on the interface and start picking up route advertisements.

Once that’s done, go into your security group and replicate your IPv4 firewall rules with identical rules using IPv6 source/destinations.

More fortifications!

A big theme of the last few years with this car has been fortifying it to survive motorsports better: Transmission kits, coolers, brake work, better seats, the new giant radiator – all of it an attempt to make it easier to drive and more reliable.

This post is about another item that won’t necessarily make the car go faster, but it will be much less liable to blow up. And that’s important. You don’t have a chance to win if you can’t even finish.

Over the past couple of years, I’ve coveted a setup Bray Lay showed me. He’d figured out how to use a machined aluminum spacer under the plenum with push-loc fittings to replace the vacuum lines. This setup imparts an enormous amount of confidence. One of the worse things that can happen to one of these cars is the fuel pressure regulator losing the reference signal from the intake manifold. Another big buzz-killer for my friends running speed-density ECUs is if the MAP sensor also loses signal. Both conditions will cause the engine to suddenly go lean.

In my case, the vacuum lines would often work their way loose from the vacuum block on the top of the engine.

Vacuum lines

The stock vacuum line setup, I had already pull the lines out, but this is exactly what they’d do.

I’ve also had the line come off the fuel pressure regulator.

Now, I’ve been lucky. Unlike a drag run, where you’re full-throttle and full boost for somewhere between 9 and 13 seconds, with autocross, you’re never at WOT for more than a second or two at a time. This aspect of my chosen hobby has prevented calamity. I’ve had these lines pop off – despite numerous zip ties and clamps – many times at events. However, I’m signed up for a couple of track days this year. Going WOT down a straight that’s nearly a mile long is in my future. A hose popping off under those conditions will be the end of the engine.

So, replacing these with push lock fittings is a no brainer. I had been looking at piecing a kit together myself, but recently Don Cruz at Cruze Perfomance has started offering a package with all the hard-to-get parts. And his kit is cheaper than putting it together yourself from Amazon and Fastenal. You may have to add a few things like some 1/8 NPT nipples, but you can get those at any hardware store.

The other item that needed addressing is what I believe to be my last oil leak: the turbocharger drain. The drain line on these car is notoriously difficult to reach when replacing it, so leaks are common once it has been disturbed. A click over to GN1 Performance out of California netted their turbo charger oil line kit, which uses a -4 supply line, a 60 micron filter, and a -8 drain. It’s expensive, but it is way cheaper than dealing with an engine fire.

So, first step in all of this? Tear down the top and front of the engine:

Engine after stuff removed

Engine after stuff removed

Essentially:

  1. Remove the ignition box and coils
  2. Remove the plenum and throttlebody assembly
  3. Remove the intake piping
  4. Move that upper radiator hose out of the way
  5. Move the lower radiator hose out of the way
  6. Remove the thermostat housing
  7. Remove the coolant tempurature sensor that’s to the passenger side of the thermostat housing
  8. Remove the turbocharger

With all of this out of the way, you can begin.

At the base of the block is a brass fitting that houses the oil pressure switch for the light in the dash and the feed line for the turbocharger. Clean it up, and remove the feed line from here:

Turbo oil feed line block connection

Turbo oil feed line block connection

The line kit comes with several shiny bits:

GN1 Line kit

GN1 Line kit

The oil line kit comes with two adapters:

Adapters. The one on the left goes into the fitting on the block, the right goes in the turbo.

Adapters. The one on the bottom goes into the fitting on the block, the top goes in the turbo.

The fitting on the bottom of the picture goes into the block fitting. The one on the top goes into the turbocharger. If you have one of those brass 90 degree fittings that adapts the turbo’s feed from NPT to a flared line, remove it. If, like me, you have a non-oem turbo, you may need a different adapter (I did, -4 AN to 1/8 NPT).

After getting those adapters in, flip the turbocharger over and clean all the gasket mess off the drain pad. Then attach the -8 AN male fitting with the included gasket and the old hardware.

On the engine side, remove the drain line fittings from the block. The black adapter in the kit screws directly into the block. Be sure to use some high-temp PTFE sealant for all of the pipe-thread connections, or they’ll leak. Don’t use the teflon tape, it won’t hold up under the heat.

Now comes what was the hard part for me. I installed the -8 line with the 90 degree swivel at the block side. The 45 degree is not a swivel end, so lining it up with the fitting on the turbo was a bear. GN1 recommends loosening the bolts on the turbo so you can clock the center section to line up with the drain lines. sadly for me, one of the bolts on the exhaust side was stuck, so I couldn’t pull that off.

I carefully screwed the fittings onto the turbocharger before I bolted it to the manifolds. That allowed me to wiggle it around and get the threads on the drain line started. Once they were on (but not tight!), I bolted the turbcharger back into place, then carefully tightened the fittings. Be aware, and make sure you don’t crimp the drain line.

Installing the feed line was much simpler. Depending on your exhaust routing, you may want to run it either behind or in front of the turbo. Your choice. Just make sure the filter and lines aren’t touching any exhaust parts or rubbing any sharp corners.

Oil drain line

Oil Feed Line

Oil Drain Line

Oil Drain Line

With this part done, it was time to move upwards to the vacuum lines. Here’s kit:

Vacuum line kit

Vacuum line kit

It comes with an intake spacer (I got the 1″ thick option, drilled for an IAT sensor), 25 feet of nylon line, a block off plate kit to go where the stock vacuum block goes, and all the fittings.

Now, since I’m running a regular MAF based ECU, I don’t need to buy a fancy MAP sensor that can take the fittings, I’ll use a clamp on that guy, but the fuel pressure regulator needs to be modified to accept a push-lock fitting.

Fuel Pressure Regulator with new fitting.

Fuel Pressure Regulator with new fitting.

This was tricky. The fitting on the regulator was tiny. I carefully disassembled the regulator, then used a drill press to enlarge the hole in the top half, then tapped it out to 1/8 NPT. Then carefully reassembled it.

If you aren’t secure in doing this without ruining the regulator, Don Cruz will do it for you if you mail your regulator to him.

From this point, things were pretty straightforward, except for one or two little nags. You basically assemble it:

Plenum spacer installed

Plenum spacer installed

Now, looking closely, you can see my new IAT sensor. You can also see how close it is to the idle air controller housing. Yup, the IAT cannot be connected. If you are doing this, and you are running a stock throttlebody, ask Cruz performance if they’ll drill your IAT hole on the side or the back, not the front. Big clearance problem. I won’t be able to use this new sensor until I change to a different throttlebody.

PCV Lines hooked up

PCV Lines hooked up

Hooking up the PCV is simple enough.

The system comes with this cool vacuum distribution block. I had to make a bracket to mount it next to the plenum.

Distribution block on home-made bracket
Distribution block on home-made bracket

Distribution block mounted

Distribution block mounted

With this mounted, it’s a matter of connecting stuff. The two clamped rubber lines you see are fitted to 1/8 NPT nipples, with one going to the MAP and the other going to the vacuum lines that power the HVAC vents, evap canister, and cruise control.

The two push-lock fittings go to the fuel pressure regulator and the blow off valve, respectively. The open hole will be closed up using a pipe plug.

On the topic of the ignition. If you have a stock ignition system, it should clear the plenum fine after installing the spacer. I, however, have the Bob Bailey TR6 ignition, and the box is larger. I had to enlarge the holes on the ignition box bracket so I could slide the ignition box back enough to clear the plenum. A drill press is your friend here.

Finally, carefully install the orange O-ring in the channel on the block off plate (I used a little bit of Vasoline to get it to stay in place), and install it. The screws appear to be stainless, so I put some anti-seize on them to hopefully prevent them from galling into the aluminum throttle body housing.

Block off place

Block off plate

And really, that’s it. Hook up all the hoses, fill the radiator up, energize the fuel pump and set the pressure (since you screwed up the adjustment when you disassembled the FPR to drill it, remember?), and start it. Check it for leaks. To recap, you have disturbed the following:

  1. Upper radiator hose
  2. Lower radiator hose
  3. Thermosat neck
  4. Water temperature sensor
  5. Turbo feed line
  6. Turbo drain line
  7. All the vacuum lines
  8. All the intake piping

Double check it all, and start the car. Enjoy not having to worry about vacuum lines or oil leaks again!*, **

* Remember, -AN lines are not steel lines. They don’t last forever. You need to inspect them yearly, and replace the braided hose every five to ten years. Possibly more if you really put some miles on it.
** The Nylon line used in the push-lock fittings will get brittle due to heat. Inspect it and replace as necessary.

 

 

 

Installing a proper swaybar on a G-Body

“If you can raid the NASCAR parts bin, do it.”
- Dennis Grant

So, in my last entry, I barfed out my thoughts on sway bars for my car. Today’s entry is me actually acting on those thoughts. Today is a step-by-step on how I put a three piece swaybar on my 1987 Grand National.

Underside prior to removal of OE swaybar. Note the diagonal braces. They'll have to go.

Underside prior to removal of OE swaybar. Note the diagonal braces. They’ll have to go.

First, the underside as it was. Please pardon what appears to be an oily mess. The car has the factory undercoating on it, and while it currently doesn’t leak, all the previous leaks have turned the undercoating into what can only be described as black slime.

You can see in this picture the stock 32mm conventional swaybar, along with some add on chassis bracing. The diagonal jounce bars you see will be deleted to make room for the new bar.

Original rubber sway bar bushing

Original rubber sway bar bushing

In this picture, you can see the original sway bar chassis mount, and the 30 year old rubber bushing. The bar is supposed to be able to pivot in this mount as the suspension moves. It didn’t without considerable effort.

All the stuff I had to take off/

All the stuff I had to take off.

And here’s all the old stuff off the car. This is all you have to remove. The bar, mounts, endlinks, and that bracing if you have it.

What I’m installing is no less than a full-on NASCAR road race sway bar setup. Instead of a bent bar mounted in rubber mounts with a sandwiched bushing scheme at the ends, this bar is mounted in brass-raced pillow block bearings, with rod ends connecting the bar to the lower control arms. This setup will eliminate any bind from the crappy rubber bushings sticking and should sharpen steering response since the car will no longer have to wait for bushings to compress before the bar starts to work.

I used a 37.5″ torsion bar from Speedway Engineering. The bar is 1.25″ on the ends, but thickens to 1.5″. This meant the pillow blocks needed to be set a tad wider than the stock mount holes. I also needed to set the bar about an inch forward of the factory bar to clear the idler arm and pitman arms.

The original GM mounts were simply screwed into holes that had been drilled and threaded directly into the frame at the factory. This seems a bit sketch, but if the bar is set up correctly, the body mounts really should not see much load other than holding the assembly to the car. If it was good enough for GM, it was good enough for me. I mocked the bar and pillow blocks in place, and marked the rear mount hole locations.

Drilling new holes for the pillow blocks.

Drilling new holes for the pillow blocks. Don’t use this as a guide on yours, MEASURE!

I then drilled them and tapped them to 3/8-24, and mounted the pillow blocks.

Pillow Block on new rear mount.

You can see in the picture that the frame slopes upward away from the mount. A spacer had to be made. In my instance, a lug nut ground to the proper angle to sit flush on the frame while also providing a level mount for the pillow block worked perfectly.

The next step is the bar arms. I got my from Coleman Racing. They’re 49 spline and 17″ long, with a 30 degree bend in the vertical plane. However, to clear the frame and meet up with the end links, they needed to be bent 45 degrees outward. Since these bar arms are 0.75″ thick steel, that was a concern. I considered trying to heat them and bend them with a hand sledge, but that didn’t look like it was going to pan out well. Thus came the only really specialty service required for this install: a 50-ton press.

I got lucky, really. The good folks at RLC Fabrication had a few down minutes and were able to bend these arms for me. Much thanks to them.

New bar arms after bending.

New bar arms after bending.

Now, this being a non-standard install, it’s important to mock stuff up every single step of the way. Here is one of the bar arms on the end of the swaybar for the first time. It looks great, right?

First mock up

First mock up

Wrong. There were issues. First off, the bar arms would hit the steering box bolt heads on the driver side. Shifting the bar over to the driver side to clear the bolt heads rendered the passenger side nearly immobile as it hit the frame on the way up. When you get down to the finished pictures you will see that I flipped the bars side to side. It provided way more frame clearance that way.

Now, after getting the arms in place, I used a piece of baling wire jammed into the rod ends as a sort of variable length end link. I used this to set the end link length, and mark the arm bars for drilling. You want the end links as close to vertical as you can get them at ride height. Once I had an overall length, I cut some threaded rod to fit and assembled the end links. Don’t forget jam nuts!

Once you’ve marked the spot, drill the holes. With metal this thick and tough a drill press is mandatory. Use a lot of cutting oil or you’ll dull the bit. We drilled a small pilot hole, then stepped up to the 3/8″ we needed.

End link mocked into the arm. The spacers provide room for the rod end to articulate.

End link mocked into the arm. The spacers provide room for the rod end to articulate.

Here’s the end link on the bar arm. Note the spacers. Without them, you lose a lot of articulation range on the rod end.

Now, right at this point is the important part: Make sure you have clearance. Mock it up, and articulate the suspension to make sure nothing hits. If something makes contact while you’re driving around, you will at the very least break something. At the worse, you’ll crash the car. Do not screw around with this step. Don’t save time. Don’t shave corners. Make sure the suspension can cycle through its entire range of motion with the wheels at full steering lock in both directions. If there is any interference anywhere, fix it now.

Powdercoated the arms chrome. Blang Blang.

Powdercoated the arms chrome. Blang Blang.

This is the purty part. We powdercoated the bar arms chrome. I think they came out quite nice. Mirror finish. With a Craftsman powdercoating gun from Sears. How about them apples?

Finished product. Arms had to be flipped to provide extra clearance close to the frame.

Finished product. Arms had to be flipped to provide extra clearance close to the frame.

Here’s the driver side snugged in. In the picture, you can see the pillow blocks, the shaft collar used to keep the bar from sliding side to side, the bar arm, and the end link. This photo was taken with the suspension at full droop. Even at full droop with the wheel cranked over to full lock, there is plenty of clearance between the bar arm and the tie rod.

All Done!

All done!

So there you have it. Total install time was about six hours, not including travel time to the store and to get the bar arms bent. I think with all the parts and properly bent bar arms ready to go, I could do this install again in about two hours.

Driving impressions? I’ve only had it around the block, and there is really no way to to test it properly on the street. But what I did get is that the setup is very responsive. It’s also more comfortable. With the new setup, the sway bar is not involved until the car rolls. So hitting expansion joints on the highway? Way better than before. The old bar was definitely binding up. However, street impressions are for naught. The real proof will come a the first event of 2017. During 2016, I was routinely pulling 1 to 1.1 lateral G in the car. It would understeer and wash out quite a bit as the inside front tire came off the ground.

If this bar succeeds in keeping the tires planted, I expect a front end grip improvement that should be measurable. Keep an eye on this space for an update in the spring.

Here’s a photo album with additional photos:

https://goo.gl/photos/enNV6Aki1ppkbELW7

And now, the parts list. I got the actual sway bar, the bar arms, and the pillow blocks from Coleman Racing. Other parts as noted.

1x Speedway 1.25/1.5 Hollow torsion bar, 608-49-150
2x Coleman Racing Products Sway bar arm, steel, 30 degree, 21915
2x Sway bar pillow block bearing, 12328
2x aluminum shaft collar, 1.25″ inner diameter, Amazon.com
1x 12″ fully threaded 3/8-24 rod, McMaster-Carr
2x (for me, you might need 4) Rod end, 3/8-24 female shank, 3/8 ID, McMaster-Carr
4x Steel unthreaded spacer, 3/8 ID, 3/8 length, McMaster-Carr
1x Pack of 4 Grade 9 3/8-16×3 hex head bolt, McMaster-Carr
1x Pack of 10 3/8-16 nylock nuts, McMaster-Carr

There you have it. A big shout out to my Dad for the lift, heated garage, tools, time, and you know, raising me. Also RLC Fabrication for getting me out of a bind with the bar arms; and Coleman Racing Products.

 

On the topic of anti-sway bars

Anti-sway bars are the stuff of myth and legend. Especially for the General Motors Metric mid-size platform (the G-Body). You can search and read forums and old magazine articles until your eyes bleed, and you will come away with the distinct impression that nobody really knows how to deal with them. How do they work? how big do you need them to be? Is the bar from supplier X going to be enough? Will it be too much?

When it comes to swaybars on the GM G-body, the conventional wisdom (and the product offerings) seem to center around going slightly bigger. If you’re fortunate enough to have a G-Body that came with the F41 suspension package, you have a 32mm solid front bar, and a smaller rear bar attached to the lower control arms. The aftermarket supplies 34 and 36mm solid and hollow conventional (1 piece bent) bars, and that’s about it, except for Ridetech, which can sell you a NASCAR style torsion bar that’s 1.5″ (38mm) in diameter.

So what do you really need? My opinions – and these are my opinions, but I’ll explain them – follow.

The answer to which swaybar you need? “It depends.”  But I’ll go ahead and spit out one answer early: If you are not racing the car, the F41 front and rear bars are all you need.

The more I dig into the actual engineering on this car,  the more I’m impressed by GM. Some of the perceived deficiencies in the platform are really the result of cost cutting or compromises made in the name of comfort, not bad engineering. The frame is a good example. It’s a c-channel structure. People knock it for being floppy. It turns out, GM engineered the frame to work with the body as a system. If you replace the squishy rubber body bushings with a better bushing material, all that frame flex goes away and the car feels as solid as a new unibody model. There’s no need to add weight or cost by boxing the factory frame, or replacing the frame entirely with a costly aftermarket frame. Once you identify and address the compromise (soft body bushings) things work as designed, and the design isn’t bad.

The suspension on these cars is no different. The geometry was parts-bin engineering, a metric-converted version of the A-body from the late 1960s. The design goals were cost and comfort. The front suspension was built without a lot of caster. Why? Caster stresses the power steering system. They’d have had to add a power steering cooler to all these cars if they’d run the kind of caster modern cars run. And all these modern cars have power steering coolers on them now.

With the F41 package, GM definitely subscribed to the soft spring, stiff bar mentality. And it works beautifully. So I’ll say it:

Unless you’re racing the car on race compound tires, the F41 swaybars are exactly what the car needs. Any more front bar without changes to the front suspension will make it push, and any more rear bar will make the car’s snap oversteer problem even more snappy.

Now, what if you’re racing? Things get more difficult. To keep the car planted and all four tires on the ground in a turn, you need to understand the geometry in the front, your shocks, and your tires. Fix the geometry and the tires, and the car will start heaving further, and will eventually start picking the inside front tire up. When you get your car to this point, it’s time to step up in front bar size.

Buick Turning hard

Buick in a Turn, check the inside front tire

It is obvious from this picture that the car is going to require more anti-sway of some kind to keep the inside tire planted, and transfer load from the outside tire, which is getting overworked.

Doing this with springs won’t help. Controlling roll with the springs limits the suspension travel without providing any load transfer to the inside wheels. Doing this with springs also requires some serious shocks.

So, do it with the anti-sway bar.

Why? First, you keep your softer springs, which keeps your ride tolerable and keeps your costs down by allowing you to run with a less expensive over-the-counter shock package. Second, Newton’s Third Law of motion means that in roll, the compression on the outside of the car will twist the bar and push the inside tire down onto the pavement, increasing the grip on the inside tire. More grip is what we want, not necessarily less roll.

That gets us to “which bar?”

Here, too, I will provide my answer: “Not the one in the catalog you’re looking at.”

To do it right, you need to give up on a conventional single-piece bent bar like you’d get from Hotchkis. Their 34mm bar isn’t big enough. Also, conventional-style endlinks that use poly bushings won’t work. The additional rate destroys the bushing material and creates slop in the links, which makes your bar not work at all. Additionally, the conventional frame mount bushings have a lot of friction and also complicate making your bar work.

To do your swaybar correctly, you need a three-piece unit like they use on actual race cars. Ridetech came to this conclusion when they designed their Musclebar(tm), but their design was built to be an easy bolt-in, so they make some compromises. Namely they welded stuff where it could have been bolted.  I can only assume they did this for durability on a street application where people wouldn’t be checking the clamp fasteners often enough. They also only have two choices for the center bar, and my friends have found it to not be enough for autocross.

So hit up actual race car part suppliers. Speedway Engineering and Schroeder Steering offer splined torsion bars in a large number of sizes. Once you fab up the mounting and get your sway bar arms bent right and lined up, you can easily swap out the torsion bar. Instead of having to buy a whole swaybar from Ridetech for $600 each time, you just swap out the torsion bar for $150-$300.  The torsion bars also don’t take up much space, so you can just throw them in the trailer. If need to make an adjustment at the track that can’t be accomplished by moving the link mounts, you can just pull out a whole different bar and swap it in a few minutes.

My suggestion is to go bigg-ish, with a bar rate somewhere close to 500-600lb-in, then increase the size until the car starts to push in corners. Then back off a step.

But what about the rear? I’ve not seen an instance where a stiffer rear bar will help this platform. In fact, you need to allow the rear to articulate as much as possible. Stiffer rear anti-sway bars will actually cause the car to pick the inside rear tire up off the ground (solid axle!). You’ll lose traction coming out of the corner and it’ll tear up your differential having that inside wheel freewheeling.

 

CAM East and other musings

Last year, I attended the CAM Challenge East in Peru, Indiana. I wrote up that experience, which was overwhelmingly positive. You can take a waltz down memory lane here.

This past weekend, I went again. Plus more. To say the past four days have been a blur would be a vast understatement.  On Friday, I made my way up to Grissom. I had to leave a lot later this year than last due to a well check for my daughter. We got onsite just past six. Even then, I was able to tech, register, and walk the course. But there wasn’t much chumming about.

My daughter came along with me this time, and seemed to have a good time looking at the cars and playing with her stuff and reading. Oh my. The reading. Read three books front to back this weekend. I won the lottery with this kid.

As for the event, it was a repeat of last year. Fast course, lots of awesome cars, everybody having fun. Most of the points I made in my post last year carry forward. The high dollar equipment showed up once again. Once again, nobody cared. We were all there to have fun. And Fun Was had.

I had a co-driver. James Bishir, who I commended last year as an exemplary n00b, had some highly publicized car trouble at Putnam a few weeks ago. Circumstances conspired against him, and his car just wasn’t ready. So in exchange for lodging and breakfast, he co-drove mine. Wouldn’t you know it, the stinker ended up faster than me by the end of the day! I beat him overall at the event due to a quicker morning session, but he’s a quick study.  Once he gets his car back together, he’ll be able to hurt some feelings.

But CAM East wasn’t the only thing on my plate. After dinner Saturday, my daughter and I packed up and headed for home. We got back about 9pm. We both got cleaned up and passed out… only to get back up again at 6am Sunday. There was another event at NCM that I needed run in order to earn my regional year end points. My daughter was also slated to make her autocross debut in a friends’s kart.

So, we made it to NCM with plenty of time. The SCCA Targa event was finishing up there. It was interesting. Randy Pobst was there, and his introduction to the Kentucky Region of the SCCA included a demonstration on how to properly perform intercourse with an Exocet by one of our esteemed STR drivers.

Sadly, my daughter was met with some heartbreak. It turned out she was too small to safely operate my friend’s kart. She was crushed, but she’s a trooper. She bounced back quickly, helped me get my tires changed, and the rest of the day went smoothly. Until a good buddy accidentally locked my keys in my car while doing me a solid and rolling my windows up during a cloudburst. Ooops. It was nothing an old Corvette antenna couldn’t fix, though.

And here’s where we get to the car. This was the first event weekend since installing a recirculating blow off valve. At the Wilmington Champ tour, I was plagued by lag. Every time I had to lift off the throttle and then get back on it, I could count to two or three in my head before the turbocharger came back. On the Wilmington course, it had to have cost me at least two seconds.

So I put a Tial 50mm ventilator on the car.

TIAL 50mm Recirculating BOV

TIAL 50mm Recirculating BOV

For those not familiar, this valve allows pressurized air that gets blocked by a suddenly closed throttle plate to be bypassed around the turbocharger and fed back in the inlet side. This prevents air from reverting backwards through the compressor and stalling it. The result is the wheel keeps spinning while the throttle is closed, and when I mash the gas back down, I have to wait less time for the turbocharger to come back up to speed.

Here’s a datalog chart from the Wilmington Champ tour that clearly shows the problem. After a throttle closed event, the boost takes a long time to come back:

Arrows point to extreme lag events

Arrows point to extreme lag events

Here’s a chart from Sunday’s KYSCCA event:

Post BOV. No more lag.

Post BOV. No more lag.

As you can see, there’s no more lag. The car is responsive enough to actually drift around corners without spinning. That’s not easy to do with one of these.

The car placed higher than expected this weekend. At CAM East, I was in the top half after the morning sessions. I drove less well in the afternoon and fell to 21st out of 34. I would have been 41st of 67 if I’d been in CAM-C like I was last year, which compares very favorably to my finish from last year.

At the KYSCCA event, It was the same story. I won CAM-T easily, and would have been just a fraction of a second out of a trophy in CAM-C if I’d run there.

All in all, it was a fantastic weekend. Exhausting, but fantastic. The car performed flawlessly. It was a moral victory of the highest order.

And now, the videos. Here’s my best run from Peru:

And my best from the KYSCCA event:

A BMW Reflection

A post on Jalopnik today jogged my memory about an experience I had owning a BMW. Below, you’ll find an updated version of a tale I spun many moons ago over on 502streetscene.com. Enjoy.

Once upon a time, I bought a BMW. 1999 540i with the manual six speed transmission. I thought I’d made it. I had  a BMW. A desirable one. The Benchmark for a four door sports sedan. I thought I’d stolen it. Paid nine grand for it, and after replacing the clutch and the tires, I was cruising.

To this day, I still remember how well it drove. How you could drive it all day and not be tired because the seats were just that good. The stereo was shit until I replaced the speakers, but everything else about the car was amazing.

Then it all went downhill.

Rist off, it started idling like crap. The intake re-seal had to be done. It’s a typical item item, not a big deal. That was expected when I bought it. Knocked the job out in a weekend. No biggie, but the sheer number of fasteners and the low quality of the gaskets that crumbled to dust after just 100,000 miles was disappointing. Ford used much higher quality gaskets on my then-wife’s Sable. They were still nice and bendy at 110K when I replaced them.

Then I started getting Digital Throttle Control codes, and eventually it went to failsafe and wouldn’t move. Both TPS sensors in the throttlebody were fried. Root cause? The electrical connector on the computer-controlled thermostat leaked, and coolant wicked up the wires all the way back to the DME and shorted out a bunch of shit. Cost to repair that was a $300 throttle body, a $180 thermostat, a few connectors spliced into the wiring harness to stop water if it leaked again, and oh, I had to tear the top of the motor off and do the intake manifold re-seal all over again. Oh, and corrosion from the coolant shorted a pin that ran the secondary air injection pump to the #1 TPS… inside the DME. That meant unplugging the secondary air injection system, which is an emissions component, which means the car could no longer be registered anywhere that has emissions testing without a DME replacement.

Then the shitty plastic snap-on connectors they use on the radiator hoses failed catastrophically and without warning, dumping all my coolant out on the road in J-Town. Normally a cooling system failure is preceded by a leak. Not on a BMW. That shit just explodes.

THEN the real fun began. One day I start the car and it’s making this high pitch squeal. It’s coming from the driver side valve cover. Pull the valve cover, and there are chunks of what turned out to be timing chain guide all over the inside of the engine. BMW uses a very brittle and cheap plastic on the timing chain guides. If the tensioner isn’t replaced at the proper interval, the chain goes slack, beats the guides, and they crumble.

The kicker? There’s no replacement interval for the tensioner in ANY of the BMW service literature or the owner’s manual, which means most of these cars are running around with slack tensioners. From reading other peoples’ experiences on bimmerforums, the tensioner should be changed about every 50K miles or so.

But alas, it wasn’t on mine. Replacing the chain guides is a 23 hour job according to the book. It requires over a thousand dollars in special tools to block the cams and the crankshaft at TDC so nothing moves while you have the chain off, you have to tear the engine down to the bare longblock, and the car must be re-timed and the adaptations in the computer cleared or it’ll run like shit when you put it back together. The crank bolt must be torqued to 100ft-lb, then turned another 150 degrees in three more steps. I borrowed a torque wrench that did torque angle. It quit when I hit 500ft-lb on the bolt, and I was only halfway through the second tightening.

I did the guide replacement myself and then had it towed to Stein for them to re-time it. That lopped 15 hours off the bill, and it was still three grand.

Oh, and I spent an hour with a set of needle nose pliers pulling chunks of chain guide out of the oil pickup. Had that stray piece not gotten jammed in the right spot and made the noise and alerted me to the problem, I’d have never known, and the pickup would have eventually been completely blocked and the engine would have been oil starved and completely ruined.

Oh, and behind the chain guides is an oil separator. It’s made of brittle plastic and will break as soon as you touch it. Once it breaks, the car smokes like a freight train. Replacing it requires tearing the entire engine down again, because it’s behind the damn timing chain.

So, I got all that fixed. Car was running great… for a week. Then the steering interlock broke, immobilizing the car. Towed back to Stein, they had it two weeks waiting for the interlock, new keys, and a new ECS module. $600 more.

I put it up for sale right after that. In a single year, the car had cost me $7500 in parts and labor, $2000 in depreciation, used up all of my tows on my AAA membership, and was actually in-service for just 10 of the thirteen months I owned it.

I added it up after I sold it. I literally would have been cheaper for me to walk down to the BMW dealership and lease a BRAND NEW 550i than it was for me to own that E39 for a year. Literally. Lump together purchase price, parts, labor, and depreciation and divide by 12 and I could have driven a brand new car instead. Maddening.

As for doing the work yourself, a good friend once told me that a BMW owner needs but two tools: a cell phone and a checkbook. I used PTO to take many days off work to fix that damn car.

The only good thing, maintenance wise, about that car is changing the oil. With the canister filter and easy-to-reach drain, I didn’t even have to jack the thing up. Fifteen minute job… of course, by the time you buy the $30 filter kit and eight quarts of the $7.99 Mobil1 or Castrol Euro formula BMW LL certified oil, you have an $90 DIY oil change on your hands.

Like I said, I LOVED that car when it ran, I really did, but it made me pay for the pleasure.

The Road to (and from) the Optima Search for the Ultimate Street Car

Fair warning, this will be longer than a typical post.

Way back in February, I did something crazy. I entered the Optima Search for the Ultimate Street Car event schedule for the NCM Motorsports Park the weekend of June 10th. At the same time, I ordered my big Weld RT-S71B forged 18×9.5″ wheels. I was feeling giddy. A big name event, actual track time, big sponsors, lots of photo ops, and high-dollar competition that was surely going to crush me, but would be awesome to be able to compete against.

The best part? No work assignments! The second best part? I had five months to prepare! Easy?

Turns out, not so much.

My entire mission the past year or so has been to make changes to the car to reinforce it. After the engine rebuild, it’s been about longevity and reliability. I got new front suspension arms not because what I had didn’t work (it did), but because the new stuff was stronger and had 15 years’ more engineering know-how put into it. I built an engine that can make 500-600 horsepower on race gas and 30psi of boost, and have elected to (attempt) to run it at 17psi on pump gas, because I don’t want it to blow up. Instead of taking the leap to a Megasquirt, I jumped on an opportunity to simply add a blue-tooth enabled connection to my Powerlogger so I can monitor the engine without a lengthy development effort or cutting up the dash to fit more gauges.

So in the run up to the USCA event, it’s all been about evaluating the car and fixing stuff.

The first thing to fix was my new wheels. One of them ended up not being round.

Now, I have no idea how that happened, but Weld took the wheel back no questions asked and fixed it, so good on them. However, they couldn’t get it fixed and back to me in time for the USCA event, so I ran it on my old 245mm Dunlops. Omen #1.

Omen #2 was a trip down to Lexington for an autocross in May. It was hot. 90 degrees. On the way home, I had my handy new Bailey Engineering Scanmaster-G set on the coolant temp, and noticed that I was running 195 degrees on level ground. When cruising uphill at 70mph, the temperature climbed above 200. Not good, especially when I was anticipating having to run boost down a 1 mile long straight at NCM a month later.

So, after I got home and let the car cool down, I popped the radiator cap and looked inside.

the crusty insides of a 29 year old radiator

the crusty insides of a 29 year old radiator

Yup, thirty years had taken its toll. The tubes were crusty and full of deposits, the oil coolers were covered in slime. Once I had it out of the car, I found several pinhole leaks that had sealed themselves with corrosion. All in all, this radiator had lived its useful life. So I ordered a new one from GNS Performance. The radiator they sent me was a work of art. All aluminum, dual 1300+ CFM Spal fans. Lovely.

new radiator next to the old one, note how much thicker the core is

new radiator next to the old one, note how much thicker the core is

If this didn’t fix my cooling issues, nothing would. Thankfully, it fixed them. I did have a battle with the relays, though. The mounting brackets are crap. If you buy this radiator, zip-tie the relays to the brackets before you attempt to install it. Otherwise, they’ll come off the brackets and drag on the ground.

Omen 2 dispatched.

Omen 3, and the one I should have taken to heart and withdrawn from the competition and gotten my money back, was when I noticed my passenger side axle was leaking. Figuring the bearings where shot, I got new bearings and seals. When I popped the rear cover to get the C-clips out, I found this:

Missing ring gear tooth

Located missing ring gear tooth

Yes, there was a tooth missing from the ring gear. Conveniently, it had found its way to the magnet on the back cover and not done any further damage.

There’s a lesson here: DO NOT LEAVE OUT THE MAGNETS when you overhaul stuff. They were put there for a reason. There is no way to tell when that tooth broke. It had been at least three years since I popped that cover, maybe more. If that tooth hadn’t stuck to the magnet, instead finding its way back in-between the ring and pinion, the rear would have locked up, then shattered, and the car would have spun off and likely hit something unpleasant.

When my Dad and I pulled the wheel bearings, we found the passenger side bearing had, in fact, spun. Oops. It’s likely that whenever that bearing stuck, then spun, that’s when the tooth came off the ring gear. Or not. Hard to tell.

Anyway, we found this Tuesday night the week of the event. After a judicious application of the plastic wrench (thanks for the metaphor, Rich), I had a new ring and pinion on the way from Summit. It got here by 10am the next morning. Along with a new installation kit.

The next day, as we were installing the new ring and pinion, we discovered the installation kit had come with the wrong side bearings. GAH! Thankfully,  a local truck parts house had the proper bearings. My good friend, Tom Bell of Bell Motor Service helped me get stuff pressed off/on, and we got the diff back together Wednesday evening. Summit racing even took back the incorrect bearings and refunded me $58.

Then came the really hard part. Ring and pinion sets must be properly broken in, or they will fail. I needed to put 500 miles on the car by the time I got to Bowling Green – less than 36 hours from the time we buttoned the diff up.

So, Thursday was a driving day. My daughter packed up some books and videos and her MP3 player, and we climbed into the Buick early Thursday morning. We took the back roads to Newport and ate lunch at the Haufbrau Haus, then took I-71 back home. That got us 350 miles.

Friday morning, I packed up and headed for Bowling Green, again taking the back roads to extend the mileage and vary the speed. I arrived at the track just past noon, having put just over 500 miles on the ring and pinion. I paid $100 for a garage and parked the car so it could cool off. If you’re ever doing a track day event, pay for a garage. Being able to get out of the sun is worth every cent.

While it was cooling off, I took some time to walk around the event.

This twin turbo Camaro could have its power level dialed in anywhere from 500 to 1300 horsepower!

The equipment present was fantastic. It was also HOT. I don’t think it go below 90 degrees at night.

Anyway, I changed the differential fluid at the track, and thought all was well.

The next morning was the Speed Stop challenge. We started on a section of the road course, accelerated down a hill, up another hill, and had to stop the car in a box just over the crest. Much tougher than it sounds. I got one good run, then the car started stalling and sputtering. It wouldn’t rev past 3000 rpm, which, coincidentally is when the fuel injection system switches from sequential to batch fire. That’s important. Remember that.

After making a few more attempts at runs, I finally limped it back to the garage. I pulled the logs from the runs out of the Scanmaster and found that, curiously, when the engine burped, every single sensor spiked. This was good. It meant this was a problem internal to the computer, not a problem with the engine. So I pulled the computer out.

It was so hot to the touch, I nearly dropped it. After setting it on the concrete floor to sink some heat out of it, I opened it up. I wish I took a picture, but what I found was amazing. A ground had completely burned up inside the ECM. Now, since these ECMs sink a lot of current, they have a bunch of ground pins on their connectors, since one pin with a single 16ga wire isn’t enough to handle the multiple amps that ground through the computer. One of those pins had overloaded and melted, leaving the ECM with insufficient ground capacity. The epoxy that’s used to weatherproof the unit had melted in places. It had gotten hot, and likely had one or more internal short circuits.

This is where the 3000 rpm thing comes in. The car was stalling at 3000 rpm. When the fuel injection switched from sequential (1 injector grounded at a time) to batch (6 injectors grounded) the current overloaded the ECM and caused it to reset. At best, the car stumbled. At worst, it stalled completely.

Now this is where a small miracle occurred. Where do you get an ECM for a 1987 Buick Grand National on Saturday? In Bowling Green? Not at a store, that’s for sure. I called them all.

But wait, each year Bowling Green hosts the Buick GS Nationals! There had to be somebody nearby  that raced Buicks that had an ECM on the shelf. I called my friends at Boost Crew Motorsports, and within an hour, a kind soul brought me a loaner ECM.

So, while my Speed Stop runs were poop, I had a new ECM and an afternoon of autocrossing to get done. I also had the Design and Engineering portion, which I crushed. Top Ten finish in that section. Go me.

The autocross was HOT. The Dunlops didn’t like the heat and washed out halfway through my first run. I got two more runs done before the brand new cooling fans quit.

That’s right. 98 degrees and I had no cooling fans. It was at this point I threw in the towel. It was too hot and I was too tired. Racing further risked damaging the car worse or me losing my temper. I wasn’t having any fun. It was time to go home.

My good buddy Dave happened to be at the event working, and he’d brought his truck. A quick call to U-haul for a car trailer and we were loaded up and headed home.

Northbound and down…

Ironically, this would be the second time Uncle Dave had bailed me out of a racing-induced failure two hours from home. He was the one that got me home in 2014 after I blew my head gasket at IRP. I really need to get my own truck and trailer.

Once I got home and had time to properly troubleshoot, I found the root of the problem:

Melted!

Let me take a paragraph to explain what you’re looking at. It’s a ground, melted into a loom. The ground had gotten hot, probably from working loose and arcing after I installed the new fans (which grounded through this ring). As it melted the plastic, the plastic eventually encased the ring and separated it from the bolt head that was grounding it to intake manifold. This severed the ground and disabled the fans. I think it also re-directed a bunch of current through the ECM, which is what burned it up.

Needless to say, this particular ground has been fixed, and fixed right. A reman ECM was sourced and a spare I had in the garage has gone into the trunk “just in case.” The entire weekend, including the radiator, ring and pinion, entry fee, and hotel cost me over two grand.

That said, I didn’t do too badly. Thanks to my top ten Design and Engineering score along with Autocross and Speed Stop times that were above the bottom third, I managed to not be DFL despite scoring a big fat 0 on the road course portion. That’s a big deal.

I plan on trying again next year. I think with my big wheels on the car and all these other gremlins sorted, the Buick should turn some heads next year.

This thing actually works pretty good!

So, there have been more than a few posts on here regarding  my foolish endeavors prepping my Grand National for SCCA autocross competition. Some regard it as silliness, most others think it’s pretty badass. I’m having fun with it, though, and that’s all that really matters.

That, and results. Is what I’m doing working? How do you tell?

You tell with data. You collect data, and you analyze it, and the data will tell you if what you are doing is working or not. Without data, I’m just talking out of my ass.

So I got some data, then overlaid it on this video:

This is telemetry collected using superimposed on a video feed from my Go-Pro. If you watch the little g-meter in the bottom left, you’ll see the car hit 1.1g lateral acceleration, not in a crazy offset, but a sustained turn.

1.1g. Sideways. In a 1987 Buick Grand National.
Granted, this was at the Wilmington Air Park in Ohio, which is concrete. This car would not be able to do that on asphalt. Or would it?

That one was on asphalt, on a really cold day. It hit 1.0g. I’m happy with it.

What put the car over the top? What made this possible? Tires.

Big, beefy, sticky BFGoodrich Rival S tires. In 275/35R18. The biggest I could fit under the car without cutting it up.

Rear view of new tires

Rear view of new tires

Front tire at full left lock

Front tire at full left lock

What made getting this much rubber underneath a car that came with 215mm wide tires originally? Careful measurement and custom offset wheels. I aquired a set of Weld RT-S71B forged wheels for these meats. Getting them on the car and balanced required on-car balancing, since these rims are lug-centric. Weld can also only manufacture to a half-inch on the offsets. These wheels needed 1/4″ spacers on all four corners to truly get them to not hit stuff.

New Weld Wheels

The results are remarkable. I’m at the point now where I finally feel I’m in need of a bigger front swaybar. You see, the car is rolling a bit too much now and putting too much load on the outside front tire. It likely always has, but now I can prove it:

Trying to drive out of the tires

Trying to drive out of the tires

This photo was captured by the people at autoxpix.com, and shows quite clearly the wheel attempting to escape the bead of the tire during the 1.1g turn in the first video. I’ll attempt to compensate for this with more front tire pressure moving forward, but a better front anti-sway bar is going to be the real fix. Anti-sway bars don’t just make the car roll less in turns, the extra roll resistance actually transfers load from the outside wheel to the inside wheel, allowing the inside tire to handle more of work.

I don’t know when I’ll be able to get that done, I need to treat my property for termites soon, and that costs about the same as the swaybar I need. Oh, the woes of being a grownup.

Cadillac CTS-V and the steering thing

My Cadillac CTS-V sprang a really bad power steering leak the other day. Big leak. Giant puddle on the floor. Turns out it was the pinion seal on the rack and pinion assembly. Changing in the car would be a bit insane, so I had to pull the rack.

Now, there’s no procedure for removing the rack and pinion in the factory service manual, and I haven’t been able to find one on the internetz. So I’m writing one. Now. Step by step. No pictures, because you really don’t need them.

  1. Raise the front of the car as far as you can, set it down on jackstands positioned under the lift points on the subframe  (there are little arrows on the skirt showing you where).
  2. Remove the front wheels
  3. Remove the brake calipers and hang them off the upper control arms using wire or zip ties.
  4. Remove the brake rotors. Don’t beat them off with a hammer, they’re held on with a little torx head screw.
  5. Separate the tie rod ends from the spindle. I did it by running the nut almost all the way off, then whacking them with a ball peen hammer.
  6. Now is where it gets fun. Set a jack under the engine cradle, then remove the two driver side engine cradle bolts (21mm heads). Gently lower the driver side of the engine cradle onto a jack stand.
  7. Loosen the motor mounts from the engine cradle. The rear motor mount nuts are 21mm. The fronts are an 18mm nut welded to the frame and you’ll need to get a wrench on top of the bolt, that’s 13mm.  Loosen them as far as you can without taking the nuts off.
  8. Raise the engine with a jack, being careful not to crush anything up top.
  9. Remove the bolt holding the steering shaft to the pinion.
  10. Trace the wires coming out of the black cylinder that’s screwed into the rack gearbox portion to a connector right in the front of the car next to the ABS module. Disconnect and pull the wires out.
  11. Unscrew the black cylinder, being careful not to twist the wires. They’ll break if you twist them, and if they break, you’ll have to buy a whole new whatever that thing is.
  12. Disconnect the two power steering lines from the rack gearbox.
  13. Remove the anti sway bar.
  14. Unbolt the two bolts on either side that hold the rack down.
  15. Wiggle the rack out by moving it forward into the space vacated by the swaybar and pulling it out through the driver side wheelwell.

It’s not difficult, but it is time consuming. Take your time. Reinstall is reverse of removal.

It’s Straight!

Owing to an uneven fender overhang that’s existed since I bought my Grand National, I’ve always been suspicious as to whether or not my car was straight or had been in an accident before I got my hands on it.

Even after I replaced all the body bushings and yanked the car straight (as far as my measurements could tell), the uneven overhang persisted. It finally bugged me enough I called up a body shop here in town known for frame work and building race cars, and had them put it on their frame rack to measure it all.

Turns out the car is straight as an arrow and square as Dr. Sheldon Cooper. The fenders themselves are misshappen. Good ‘ol 1980s GM quality.

So the overhangs will remain, but the chassis is 100% healthy, even after all the racing I’ve been doing with the car. I have just a few things left to get installed before the season starts: a brace to keep the engine from rocking over too far under load, and new oil cooler lines. Both are in my possession and likely will go on this weekend.