Secrets of the Mirage

[AScoda]

And I ended up with an acid-hole in one of my favorite pairs of shorts.

Chicks in New York pay top dollar for that kinda style.

Also, Acid-hole is a great name for a rock band.

[Loren]

Well, observers tell me that my car is still lifting a front wheel when pressed too hard. I can't feel it, but I know when it happened. I got a LOT of rear wheel lift and was hanging a fast sweeper pretty hard.

Gonna have to keep working at improving things. I'll start by lowering the car another half inch. And I guess I need to get started fabricating those ball joint spacers. Maybe order my front wheel spacers.

This bitch must be tamed!

[Loren]

Well, if you've been following the Custom Control Arms discussion, you'll know that I have solidified a plan. Here it is:

Ball Joint Extenders/Spacers are out. Can't work practically for me for a variety of reasons. Going with an off-the-shelf extended ball joint instead. That limits me to 1" of extended ball joint length (was previously targeting 1.5-1.75"), but with much more confidence that stuff won't break (the ball joints I'm using are regularly used on 4000 lb cars).

To fit said ball joint, I'm going to have to modify the lower control arm to receive it. Not a big deal, just a little cutting and welding. I was trying to avoid it, but it's not difficult or scary. Spare lower control arms are already on the way.

To acheive my goal of getting the roll center height above 3" with just 1" of ball joint height change, I'll also be extending the lower control arm by 1/2". In addition to being a factor in determing the roll center geometry, that will also have the effect of making the front track width wider, and without requiring a wheel spacer to do it. Win-win.

The last thing I'll be doing to help with the roll "moment" is actually raising the front of the car by half an inch. Yes, this raises the CG, but it raises the roll center MORE than the CG, thereby reducing that "roll moment", which is effectively the length of the lever through which the weight of the car is applied to body roll.

There is also some benefit to having the front slightly higher than the rear on a FWD car, which is a different discussion.

A lot of this is discussed in greater detail in the other thread. But, in a nutshell:
Stock CG (estimated) - 20"
Stock Roll Center (calculated) - 5.75"
That gives a Roll Moment of - 14.25"

Goal should be to be no MORE Roll Moment than that, and ideally less.
Less Roll Moment designed in the suspension geometry = less body roll and more stability... even without adding roll stiffness.

Simply lowering the car by 2.5", netted:
Lowered CG - 17.5"
Lowered Roll Center = -0.5" (that's BELOW ground level)
Lowered Roll Moment - 18"

The new plan, lowered 2" and with modified control arms and tall ball joints:
Modded CG - 18"
Modded Roll Center = 3"
Modded Roll Moment = 15"

Still not ideal, but should be a lot better.

[Loren]

The control arms showed up today. Took some quick measurements. And ordered a crapload of stuff!

• Weld-in Threaded Ball Joint Sleeves
• QA1 K772-style 1" Extended Ball Joints
• K772 Ball Joint Boots
• QA1 Ball Joint Tool Kit
• K772 Ball Joint Socket
• 2.5x1" 11 gauge rectangular tubing
• Adjustable Hole Cutter
• Cheap Generic K772 ball joint (to act as a heat sink and protect threads while welding the sleeves)

Sounds like a lot (and it is), but the parts only add up to around $180. Another $50 in tools and supplies. Oh, and $95 for the control arms.

So, the stock ball joint taper is 7 degrees. That means the replacement ball joint will fit the stock knuckle. The ball joints are actually pressed in and secured with circlips, btw. Not sure if the stock ones are that way or not. (nobody lists a replacement ball joint, not even Mitsubishi) Doesn't matter, anyway. While there are press-in style tall ball joints out there, they are a lot larger diameter. No way around modding the control arm.

The control arm is actually a lot heavier than I expected. It's stamped from 12 Gauge steel. The ball joint end is conveniently tapered down to almost exactly 2.5" wide with the folded sides tapering from about 3/4" down. So, it should be fairly easy to meld with a 2.5x1" rectangular tube and end up with a stronger semi-boxed control arm.

The ball joint sleeves are 1" tall, so they'll fit a hole the 1" tall tube perfectly.

The plan is to drill a hole in the box tubing to fit the ball joint sleeve, and then cut the tube in half right in the middle of the hole. The ball joint sleeves will then be welded to the semi-circle cut ends of the tubing.

Once some materials start showing up, I can start cutting and fitting stuff! I'm not planning to do the welding myself. I may tack it together as-needed. But, I'll probably have someone with better equipment do the final welding.

[CaptainSquirts]

nobody lists a replacement ball joint, not even Mitsubishi

Had this exact problem. Only way to get oem ball joints were to purchase the whole control arms and they came pressed with them. I have an extra tall ball joint from whiteline if you wanted to measure to see if it had better fitment. Also can you not just press the ball joints out of your current control arms and fab up the new ones to go in?

[Loren]

Whitline (or anybody else) doesn't list a ball joint that is compatible with the 2012+ Mirage. Nor do they list specs anywhere. They don't even tell you how much taller the ball joints are. Just "Roll Center Correction Kit". Not enough information.

I'll be pressing out the stock ball joints to get them out of the way. But, the hole they fit in is much smaller than the replacement.

[Loren]

I don't know if I've mentioned this before, but the Mirage has the slight tendency to cut power in certain situations for no apparent reason. I haven't been able to reproduce it. It just sometimes does it. I've been wondering if maybe it was the stability control no being disabled with the traction control. Today, I learned that is not the case. Cold weather/tires left me with much less grip than usual. The front tires would heat up a bit, but not the rears... which led to an oversteer situation. For the first time, I was able to get the rear to slide out a little, and almost got the car into its first spin. All without any lost of power. So, good news: Stability Control is turned off with Traction Control. (now, if I could remember to turn off traction control before half-way through my first run...)

So, that's not happening. My new theory is that it's doing what a lot of new cars do, though maybe with a little more tolerance. I left-foot brake. A lot. I will quite often tap the brake with my left foot to slow or stabilize the car a bit without ever lifting off of the throttle. This makes some cars lose their minds! Newer Mazdas and Fords, for sure. What happens is that the ECU sees repeated brake and throttle inputs at the same time... and assumes there is a fault in the throttle position sensor. It then does the safe thing and cuts power.

I'll have to play with it some more to see if i can verify that's what's happening. If it is, I've read that one possible work-around is to interrupt the brake pedal signal that goes to the ECU. If the ECU never sees the brake input, it can't get confused when there's brake and throttle input at the same time. With traction control off, it shouldn't create a stability control problem. But, maybe could create an ABS fault of some sort, unless the ABS computer is completely separate from the ECU.

As always, so much to think about, and nothing is ever as easy as it seems. I should buy another old British car. Their electronics don't work most of the time, but they are drop dead simple.

[Carracer]

I don't left foot brake but I'm sure I felt it pull power on me. I'm curious if its more about erratic handling of the throttle.

[Loren]

I don't left foot brake but I'm sure I felt it pull power on me. I'm curious if its more about erratic handling of the throttle.

Well, that's easy. Don't lift.

[Loren]

Now, with parts in hand, this is a beautiful thing! The ball joints are, as expected, MUCH beefier than that "cute" little stock ball joint. And, as I half-expected, even the "standard" length K772 style ball joint, because it is beefier, is already taller than the stock one by probably a half inch. That means that even with just a "1 inch extended" ball joint, I'm still getting a 1.5" change. Cool!

And, because the replacement ball joint is a significantly larger diameter, I can "move" the ball joint out and effectively make the control arm longer WITHOUT really adding any material length to it. Outer circumference of the new ball joint will be in the same place as the old one, but because it's larger, its center will be further out than stock. That's what I wanted to do, it just turns out that it will be a lot easier to do. The only steel I'll have to add is for reinforcement to box it all in as-needed.

Really happy with how this is looking. The only part I'm missing is the piece of box channel steel, which should be here Friday. (why did I order steel from Seattle???)

And someone to do the welding.

Joe, are you out there?

[jbrannon7]

Of course. I will probably need to get a fresh bottle of gas.

[Loren]

I'm putting you on notice. I should be ready for your services sometime next week.

I'm sure I'll be putting some photos up here once I've started cutting and fitting things. Might start on that tomorrow.

[Loren]

Or, maybe I'll start playing with it tonight. I removed the snap ring (What? Snap ring pliers? No, angle grinder!) and pressed out the stock ball joint.


This photo shows the control arm sans ball joint, the threaded ring that will be welded to it, the stock ball joint, a standard K772 ball joint, and the extended QA1 K772 ball joint.


And this is sort of the position the welded ring will go in. (flush with the top)


And a comparison of the three ball joints.

It's hard to get an exact measurement, but with the ball joint out and the boot removed, I can better estimate where the center of the ball is. And I'm pretty comfortable with calling the difference in length 1.5".

Tomorrow, I'll get started cutting the control arm to fit the threaded sleeve.

[Loren]

Good news, bad news.

The good news is that I got the control arms cut. Couldn't use the circle-cutter for that because it requires a center hole, and there was no center. So, I marked the circumference of the threaded sleeve and freehand cut it. Came out nice enough, and within 1/16" difference between the two.


Ready for the bottom reinforcement boxing to be cut, and then for it all to be welded together.

The bad news is that I decided to double-check some clearances before I got too crazy with things. I knew the clearance to the wheel would be close. Closer than I thought. In fact, it's about 1/8" too tall! Easy enough to fix. (and this is why I liked the idea of the rebuildable ball joints with different stud length options) I'll just return the 1" extended studs and get the 1/2" extended studs.

That brings the 1.5" total ball joint movement that I was excited about back to... 1". But, it's still a major improvement.

The other measurement I verified was the outboard clearance of the end of the control arm to... whatever. The whatever happens to be the brake rotor, or the brake shield plate before it. The new assembly, because the new ball joint is a larger diameter, will extend further. (new and old ball joints share the same center, defined by the knuckle, so the LENGTH of the control arm is not a factor here) Clearance to the brake shield is minimal, but the new ball joint would clear it... barely. Fortunately, the ball joint stud is longer, which places the body of the ball joint BELOW the brake rotor, anyway. Makes it completely a non-issue unless I fit larger brakes. (which I won't do for autocross, don't need bigger brakes for autocross)

So, still in good shape. Need those shorter ball joint studs.

Oh, and because I trimmed back the control arm from the edge of the existing ball joint hole, the total length of the control arm (to the center of the stud) only changed by 1/8-3/16". I'll verify, but I probably don't need to fit longer tie rod ends. (I should still consider bump steer correction to fix the angle of the tie rod, but there should be enough threads on them to not fall apart)

[nathanwilliams617]

The freehand cuts look good, nice and tight. Should allow for some strong welds.

[Loren]

Yeah, I was pleased with how it came out. If I get nice, tight fits on my boxing-in pieces on the underside, it should be plenty strong. Stronger than the OE part.

[Loren]

My little steel shipment arrived yesterday. Got to work on it today. I started with like an 8" piece of rectangular 2.5x1" box channel. It's almost comical how LITTLE of it I actually used. It ended up whittled down to just one side (a flat reinforcement for the bottom of the end of the stock control arm) for about 2.5", and then just about a half-inch tall triangle of the side going up to the ball joint. Could have made it from flat steel just as well, but it wouldn't have been quite as neat.

What I have is far from perfect, but it will weld up nice and strong. Gonna get over to Joe's whenever he has a minute.

[Loren]

The control arms went together brilliantly! Huge thanks to Joe Brannon for handling the melty bit.


Here's an action shot of Joe making sparks.


And this is the semi-finished product.

I still need to do a bit of grinding to clean things up, and then add paint. But, the hard part is done.

[Loren]

BTW, that unwelded spot in the middle on the bottom is a drain hole. Didn't want them collecting water and rusting out from the inside!

[nathanwilliams617]

Do the threads still work smoothly after the heat stress? Also unless it's for clearance issues, I would not clean up much with a grinder, b4 painting. More material stronger the weld, + it's officially a race car now. Good job guys.