More electrons please

Some time back, I had the brilliant idea of installing a new lightweight battery in the Porsche. The mental salve for such a purchase was that the stock battery was 4 years old and thus on the brink of complete failure. That it never exhibited any of the warning signs of a battery in stress (failure to hold a charge, low cranking power) was of no issue. Perhaps because the purchase of a new battery would allow me to reduce the weight of the battery by more than 50%, a heightened coefficient of cool (to an auto-aficionado), and would create the need for a bit of mechanical design work.
Surveying the market options, there were three brands to choose from. At the ultra-high end was a lithium racing battery. Despite weighing in at under 14 lbs, at $2k-$3k, even I could not conjure up a justification for one of those beauties. The standard choice used by many Porsche owners is an Odyssey battery with matching mounting bracket. As an aside, my desire to replace the stock battery has a tincture of validity. For whatever reason, those otherwise brilliant teutonic engineers chose to put a massive heavy battery in the car. And mounted it relatively high off the ground. More than 50 pounds high above the center of gravity. There must be some hidden reason for their choice, but it remains a mystery to every Porsche enthusiast. The Odyssey weighs in at under 20 lbs., with an extra pound or two with it's optional mounting bracket (more on that later). Altogether, a proven choice. Which leads me to the final option. A carbon-fiber battery made by a company with the odd name of Braille. Very high on the cool/gadget scale (CF for a battery?) and obviously imparts otherworldly capabilities in the car such that even a blind person could drive it. Also at about 20 lbs., with no bracket, it would knock about 60% off of the weight of the stock piece. So, though it was not commonly reviewed by the experts, and requiring some personal mechanical engineering to fashion a bracket, I took the mid-priced road less traveled.

Pause - Resume

Falling off of the blogging wagon does not necessarily mean the same for the blogging subject. Lots of hands on work in the past months, smiling (mostly) through it all. Makes me realize that, unlike a 'day job' in big tech in which projects, programs and strategies take months or years to play out, the focused act of rebuilding a car, drafting an investment perspective, or trading securities for that matter all have accelerated feedback and gratification. For an ex- (science geek and engineer) :) who was trained in cause and effect, it all feels rather homey.
I'll fill in some of the recent work shortly...

And lo, it was done. And all was good. Very, very good.

I now proclaim the suspension upgrade for my C2S (Cab) complete. I've taken a minimalist approach (for now) going only with Damptronics and sway (H&R) bars. By the way, I don't know whether it's because I chose H&R (versus GMG) or it's an 'S' and not a 'TT', but the bars I received had no adjustments in the front, and only three settings in the rear. In some respects, it simplifies life in that there are fewer setting combinatorics to play with, but I would have wanted more adjustment options. I've tried to follow Cannga's guidance (http://www.rennteam.com/forum/index.html?vs=19) to the letter, with the coilovers, stiffest sway bar setting in the rear (nothing to adjust in the front) and a GT3-like alignment. This was the last step so a few added comments re this phase. On my 'S', I was only able to squeeze about -1.0 degree of camber our of the hardware. I did not choose to upgrade the shock tower components to give me more adjustability so that's about it for now. And notwithstanding web comments to the contrary, it is quite incorrect that the alignment provides 75% of the value of these changes. Just the other way around. The new coilovers are da' bomb, the bars likely have some value (I could not do the before/after experiment here), and the alignment certainly helps in improving with turn-in (or dynamic instability if you're in to another line of machinery), but nothing breathtaking.

So, my findings are much like Cannga's. This basic mod is a must-have. For folks reading this blog who are not Porsche enthusiasts and question why anyone would hack such a valuable piece of machinery, I admit that perfection-seeking OCD behavior will always find a way. I do not know what further incremental value new drop links, et al would provide though I suspect it's diminishing returns. And this is not a track car. Just an um, aggressive drivers car. With matching driver. Further, although the ride is a bit firmer, I don't think that this aspect should sway (sorry, bad pun) anyone from going with it. The ride is just fine, immerses the driver in the experience, does create more road noise - especially tire noise - likely due to a smaller gap between chassis bottom and road enabling some interesting resonance, and the handling is night and day. Now that I've been driving it a bit, I do wish the Damptronics had some manual adjustments. Though I have no formal experience in this aspect of car mods, it feels like it could use yet more rebound damping to expunge the last of the P-car nose bob. It's mostly gone, but not thoroughly. So, for those who are still considering the change, and are comfortable shelling out nearly twice the entry fee, and want to get it 'just right', and are willing to defeat their PASM system, Moton's or alternative Bilstein's, Penske's, etc., might be the way to go. OTOH, I have no first-hand knowledge re the ride quality trade-off with these others so caveat emptor.

Finally, I had a bit of a grin on my face when the fellow doing the alignment (at TCDesign - very nice folks (http://www.tcdesignfab.com/)) measured and found that, after lowering the car front and back 18 MM, I got the ride height corner variance to +/- 1.5 mm. He said that it was nearly impossible to do - and totally unnecessary for a street car! :) So I guess the techniques cited in a prior blog entry worked. Not a true corner balance by any means, but sufficient for street use. I chose not to drop the car the full 20 MM as discussed on another thread because of road object risks and my plan to substitute the GT3 bumper for the current stock one. Not having the comparison in lip heights, I am not sure yet what to expect. We shall see..

OCD update

Just a small update from the lowering/balancing exercise or the prior post. Though the technique used to measure and level the car (based on applying a caliper to the coilovers themselves) is accurate and repeatable, it does not take in to account any variance between those measuring points and the actual stance of the car on the ground. That is, once the car is lowered and allowed to settle, there was a side to side variance of 2mm in the front, and 1mm in the rear. Most would say that these numbers are within the practical needs of the exercise. However, in light of the title of this post, I sought to further refine them.
Clearly, I would need to use the cars stance as the actual benchmark to make all of this work, and I'd likely have to iterate a few times to get this right. So here's an alternate technique I used to provide further accuracy in the balancing process and one I'd likely use instead of the micrometer technique. Again, this is not intended to be instructional or advisory for anyone else. This is simply what I did to get this all right. Got it?
Measuring the actual ride height of the car is tricky. On the one hand, 'actual' is laudable, but literally interpreted, assumes perfectly aligned air pressure and a level surface. The Porsche factory technique is an example of this. I chose to factor out the tire/ground thing by measuring the distance from the lower wheel lip to the top of the fender. It's simple, repeatable and very much approximates 'actual'.
Since this was going to take a few attempts, the time to repeat the adjustment process and measure (as well as the wear, tear and risk of jacking/removing/adjusting/remounting/etc.) the wheels needs to be optimized. I chose the following - acknowledging that the goal of this exercise is to get side-to-side balance right. If front-to-back is required, modify these techniques to realize that priority. In this case, I did the front end, but the rear, with more difficulty and dexterity, can be done the same way.
1. Jack up one side of the car.
2. Remove the front wheel.
3. Loosen the coilover adjustment retainer (i.e. lower collar)
4. Turn the collar in whatever direction/distance estimate to level the car.
5. Remount the wheel and lower the car to the ground. Do not tighten the locking collar. Do not re-torque the wheel.
6. Measure. If it perfectly matches the opposing side, proceed to step 14.
7. If not, jack the car until the front wheel is off of the ground.
8. Power up the car (you needn't start it) to get the power steering working.
9. Turn the wheel to whichever lock gets added working space to reach the coilover collar (typically the direction of the side of the car you are working on).
10. Power off the car.
11. Reaching in to the coilover without removing the wheel again (through the space made by turning the wheel AND through the wheel spokes), turn the adjustment collar by hand in the direction and distance estimated to balance the car. Use the wrench if you can, but with the car off of the ground, you can do it bare handed.
12. Power up the car, straighten the wheel, power off the car and lower it.
13. Wait a few minutes here if you can. Bounce the car a few times to ensure it has settled. Then remeasure both sides. If it is balanced from side-to-side, proceed to step 14. Else, return to step 7.
14. Finally, raise the car, remove the wheel, re-lock the collars - being careful not to affect the adjustments.
15. Remount the wheel, lower the car, torque the lugs. Done.
This seems to be a better technique (for me) to affect small changes to the car - either initially or to experiment with different ride heights and different rake angles. A quick test drive of the car shows that the new suspension bits and adjustments make for a radical improvement in the driving performance of my 997S Cab. It IS a rougher more dynamic ride but not painfully so. If you prefer more of a GT-type car, I would stick with the OEM parts. If you want to drive a true sports car, I couldn't more strongly recommend his modification.

Watch one, do one, teach one

As the title implies, the protocol that surgeons use for training can also be applied to wrenching cars. Thankfully though, this is not surgery, because I'm starting this whole process without the 'watch one' part. However unlike living beings, errors made in most things mechanical are reversible or correctable, so greater risks can be taken or less prior knowledge is required. Hugely valuable for someone with my limited skills.

Speaking of risks, I feel obliged to insert the personal version of a 'Safe Harbor' statement here. The tools, techniques, processes and opinions expressed herein are not intended to be formally instructional in nature, and I cannot accept any responsibility for issues or damages to yourself, your car, or your garage, created by performing the described actions. You're on your own, your mileage may vary, knock yourself out. Choose two.

Finally, pardon the length of this entry. It's a bit complicated and I tend to be, um, verbose.

The following describes the process I used to lower and level a 997S Cab after a set of Bilstein Damptronic coilovers and H&R sway bars were installed. Reviewing many Porsche forum entries and weighing them against the roads, driving hazards and habits that I'll be dealing with, I chose to lower the car 18mm from the stock ride height. The recommended range is 10mm - 20mm with the upper bound (where 'upper' is 'lower') being the preferred level. Bilstein recommends lowering a minimum of 10mm with the Damptronics. In general, with lower settings come improved handling, a rougher ride and the possibilities that your beloved sports car will become an urban farm plow - pushing it's front end through speed bumps, driveways and other metal wrinkling impediments.

The list of recommended tools are:
1. Floor jack (at least one)
2. Jack stands (at least two)
3. Torque wrench
4. Yard stick or equivalent - annotated in either millimeters or sixteenths of an inch
5. A micrometer - preferably digital (this is a personal preference but quite valuable)
6. The Billstein adjustment wrenches - co-packaged with the coilovers
7. Various breaker bars, metric sockets and wrenches and a soft headed hammer;
8. Laptop with a spreadsheet program - or a lab book and calculator. Or both :)
9. Dumbbell plates, equivalent to your driving weight. No fudging here.
10. Time. Unless you're expert, this takes a bit of time. But it's a bit like freshman chem class. A little bit of lab work and a lot of planning and documenting. You will not need to know how to interpolate a meniscus however.

Begin by measuring the baseline height of the car. There are numerous ways to do this, and they all assume that your car is currently set to the factory specs. Do this BEFORE the new coilovers are installed. Before performing these measurements, make sure the car is on level and smooth ground and try to have 1/2 tank of gas (minimizing variance as gas is consumed) in the tank. Do this twice. First without the dumbbell plates in the front seat, and then after. I used blankets to protect the upholstery before dropping the plates in.



Since this is a relatively exacting process, these measurements are important. After measuring, I would recommend that you construct a chart or spreadsheet, converting figures to millimeters, that lay out these initial measurements (pre- and post- weights), the chosen lowering amount, and the resultant height figure. You'll be using this to check your work when you're done. Create a column that is the amount of 'sag' that occurs when you add the weights. You'll be using this figure shortly.

I ended up measuring from the ground to the lip of the fender above the centerline of each wheel.

You'll be repeating this measurement a few times, so make sure you can do it accurately and repeatably. A couple of points here: I ended up using the measurements provided by a number of forum members to determine the proper stock ride height and using this as the baseline rather than measuring my own car. The figures I used are: front: 26.5 inches; rear: 27.5 inches. This is for a 997.1 S. The non-'S' has different tire specs so the height will be different. I originally tried measuring the gap from each of the four jack points to the ground. But they are less reliable because they are not perfectly flat.
Finally, I was sent this diagram, which is the Porsche factory method of measurement. It arrived too late to use, but you may want to try it.


After measuring, pour this data in to your spreadsheet or chart - converting figures to millimeters - to lay out the initial point, the chosen lowering amount, and the resultant height figure. You'll be using this to check your work when you're done.

Oh, did I mention that the adjustments can only be done with the wheels and tires off of the car? Thus you cannot determine if you've done all this correctly until you've reassembled the car and have had a chance to let it settle a bit. It adds to the excitement.

With all of this done, you need to get the wheels off. You can either do one corner at a time, one side at a time, one end at a time or all four at once. I chose the latter.

With the wheels removed, you'll be looking at an assembly like this (front):

The coilover assembly has two locking collars below the coil. These are the objects of your interest. The upper one is designed to pull the coil up and down against the strut and the other one is designed to compress against the upper and keep it all from moving - once set.

Before loosening the assembly, you need to take one more set of baseline measurements. These are to establish how much you must alter the length of the coilover (actually the coil bottom to strut base) after adjusting it. This is where I got lazy. As you can see, the coilover axis is not exactly perpendicular to the ground. Thus, shortening it one millimeter will not achieve exactly one millimeter of lowering. You COULD calculate the angle from perpendicular that it makes, compute the cosine of the implied angle and then determine how much you must alter the hypotenuse to achieve the correct adjustment of the vertical leg of the triangle. Because it was 'close enough' (I'm rather OCD and it pains me to admit that I guesstimated it) I assumed the coilover was perpendicular. It turns out that it's within the range of accuracy of my measurements and so it's a reasonable corner-cutting lab tip.

Be careful to measure each of the four corners. Don't assume that the two fronts nor the two rears match. They don't, but they will when you are done.

In the front, I measured the distance from the top of the hub assembly to the bottom of the upper locking collar.

In the rear, it's a bit trickier. I measured from the bottom-most thread of the assembly, to the bottom of the upper locking collar. It is actually quite repeatable as long as you are careful, just not as simple.

Once you have these measurements, on your spreadsheets or chart, subtract the desired lowering amount from these four figures. With one caveat. To achieve the proper ride height when you're IN the car, the drivers side must be lowered less than the passenger side. The difference is the amount of 'sag' (from the weights in the drivers seat) that you measured earlier. So, drivers side lowering = lowering amount - sag. Passenger side lowering = lowering amount.

Proceed by using the provided wrenches to unlock the two collars or bolts. It takes some firm pressure to do so. This is where I had to tap one of the wrenches with a soft-headed hammer. Then lower the bottom collar a bit to make enough room to measure the changes you'll be making. Then rotate the upper collar in the direction to pull the coilover shaft 'up'. Continue to measure, using the exact same locations and technique you used initially, until you achieve the corner-specific reduction specified in your spreadsheet.

Then relock the collars, making sure that the upper collar does not move in the process. I marked the collar to visually ensure that it did not spin. Completing the first corner, simply rinse and repeat.

Once you're done, remount the tires, and torque them to 94 ft/lbs.

That's it.

By this time, the sun is setting, the NFL playoff positions have been established and it's time to take your newly suspended car for a ride. Be careful when first driving to clear driveways and speed bumps. You'll need to get the feel of how much extra care you'll need in avoiding objects. Don't ask me how I know this...

Oh, the finished product:

Home surgery for fun and profit

When I last left our story readers, I was awaiting the return of my Porsche - with upgraded coilovers and roll bars. There was a small diversion when the shop determined that the car also needed front rotors and pads, requiring an extra day or so for the parts to arrive, but now the fun begins.
The fun, such as it is, is to tune the settings that were left in rough form at the garage. For reasons of time and cost, I asked the PVG folks to return the car to me before they could complete the lowering/adjusting process. So I thought, 'well, how hard could that be?'. You're about to find out.
I want to restate that the PVG folks are great to deal with. And the day I was there, there was an incredibly eclectic collection of cars sprinkled around the shop and lot: a Shelby GT350, a 1980's Rolls Royce, a 1980's Maserati, an original Mini and many others. I assume that they upload the skillset to work on these things a la 'The Matrix'. No other way they could know all this stuff.
A few Porsche blog forum readers have asked me to lay out the process I used to lower/balance the car. I'm happy to do so, but please remember that I'm a complete rookie and amateur at all this, that I have no previous experience doing this and thus likely didn't use the Porsche standard techniques to complete this process.
OTOH, it was a great opportunity to do the things that appeal to the 'XY' chromosome set: play with tools, cars and watch NFL playoff football - all while in my garage on an unusually warm northern California day.
So, taking a wrench to a big dollar sports car? What could go wrong?? :)

Before I go on, here's a photo or two:

Stock ride height:



The parts (there are three other coilovers in the box - promise):

Upgrade: Part deux

I spent the better part of December poking around the net in search of an ideal solution to the Porsche suspension upgrade quest. Lots of good information. The best thread and thoughtful advice came from this thread:

http://www.rennteam.com/forum/index.html?vs=19#20073008 where contributor Cannga discussed his research, analysis and results.

It seems like the options are relatively constrained - assuming one wants to preserve the features of the PASM system. It boiled down to the following categories:

- Replacement springs for the existing struts. Good - maintaining PASM functions, low cost; Minus - still left with the questionable shock units, limited adjustments due to the OEM shock design, PASM system may protest when the car is lowered to the desired level.

- Race-quality coilovers. Good - really high quality technology, extremely broad range of adjustment across many parameters; Minus - defeats the PASM system, generally provides a very harsh track-centric ride, very costly.

- PASM-compatible coilovers. Good - PASM compatible (duh), improved ride/handling, only modestly degraded ride comfort, replaceable coils for further performance modifications, adjustable; Minus - some ride degradation, fewer adjustments than race-quality versions, pricey (though less than most race-quality versions).

In addition, of equal value (at least) is replacement of the anti-sway (i.e. sway) bars. There are a few brands to choose from, most seem about the same, and few individuals on the web cited either a preference or much understanding about the strengths/weaknesses of each. One value is the choice of settings via multiple mounting holes.

Given all these factors, I chose to go with the Billstein Damptonic PASM-compatible coilovers. And a set of sway bars from H&M. I bought the bunch from AWE tuning (http://awe-tuning.com/) because they were highly regarded by the Porsche forum cognescetti, nice folks to deal with, and great end-of-year prices.

The car is over at Portola Valley Garage (http://www.pvgarage.com/) for the install. Then a realignment. Then the drivers report. Pretty excited about it actually...

Lemme Upgrade upgrade...

Try as I might to keep the Porsche and Ford stock, I find myself succumbing to the need to personalize and improve each of them. In the case of the Ford, there are few things to do, or so I claim, without altering the historic character, personality and collector value of the machine. The Porsche (a 997S Cab) has, much to my wife's lament, no such limitations.

The value of pursuing an evolutionary (versus revolutionary) development program is that Porsche 911's have had 40 years to build an ecosystem - of enhancement parts, owners clubs, online forums, racing programs and more. As I learn more about it all, the scale and variance of Porsche's world is just enormous. There's a lot of value in understanding how Porsche has managed the dynamics of this world, how it's helped lift Porsche to be the the premier sports car maker in the world, and how it could be applied to many other businesses - particularly SV businesses - to achieve similar success. But of the many values this ecosystem provides, is the broad level of expertise and tuner parts to turn any Porsche in to any kind of car you'd like.

Right out of the wrapper, as great a drivers car as the 911 is, there are some clear deficiencies that beg to be addressed. Specifically, and exacerbated by the rear-engine design of the car, anything that can be done to upgrade the suspension is, to quote Martha Stuart, 'a good thing'. Here's why.

Like all engineering implementations, the P-car suspension is a study in trade-offs. The smooth-riding nature of the car sacrifices road-holding, remaining flat in hard cornering and directional control when on the edge. No doubt there have been cost trade-offs (even at the price of a 911) when choosing the suspension bits of the stock car. Specifically the coilover shocks, roll bars, and links. Neither are adjustable beyond the basic setting of the dash-mounted PASM (Porsche Active Suspension Management) controls.

And finally, there are the physics challenges of a rear-engined design. White-coated lab techs in Stuttgart have done everything short of defying gravity to make a horribly flawed design run with the big dogs. But, on the edge, and challenged with the compromises of ride comfort and cost tradeoffs, the car gets wiggy.

So, I set out to upgrade the suspension to address all of this. Accepting that I'd be trading off some comfort to achieve my goals, the intent was to:

1. Reduce the famous Porsche nose bob. For every 911, Boxster/Cayman and 914 I've ever driven, they have this odd tendency to allow the front end to bob up and down, like the suspension is made out of rubber bands;
2. Maintain consistent steering accuracy when on the edge. At 8/10's the Porsche is a model citizen. But any further, and the twisting of the suspension - front to back - coupled with body lean and extreme angles of the rear suspension arms create a real challenge for the driver. Though a real hoot to deal with, it's unnecessarily dangerous and hinders the ultimate potential of the car. Further, pushing the PASM button simply hardens the ride, causing the car to skid across bumps while turning, and radically increasing the likelihood that you'll be needing dental work after your ride. Honestly, in current form, the 'hard' suspension setting is just useless.
3. Reduce body lean. Again, at more than 8/10's, the body roll exacerbates the suspension geometry changes, and induces more yawing, direction changes, and reduced tire footprint than seem necessary.

Stand by for a nick to the checkbook...