Ferrari Race Setups

Coventry | Murasama | BRM | Brabham | Ferrari | Eagle | Lotus

Chris Amon on the Ferrari 312:

"That '68 Ferrari was a gorgeous thing to drive, in that it was a car you could over-drive; you steered it with the throttle, really. What it lacked, though, was horsepower. Because it was a V12, it sounded as if it had all the grunt in the world, but we had a lot less than the Cosworth V8 or Honda's V12, and at Spa that was very apparent.

"In terms of driving pleasure, a fast lap at Spa was a fantastic sensation. Although we were down on power, and Spa was very quick - my pole lap there was well over 150 - the car was working beautifully. On that lap, as I came down the Masta Straight, after the kink, I caught up to Brian Redman, and I can remember going by him on the inside, at the entry to Stavelot, with my foot buried right in it, while he was having to brake in his Cooper! The thing felt tremendous."

Despite the Ferrari's horsepower disadvantage, Amon lapped four seconds faster than the field in practice.

"There's no way I shouldn't have won that race. I got a good start, and had more than 150 yards' lead at the end of the first lap, with Surtees behind me in the Honda. I thought I was going to open up the gap, and quietly disappear, but on the second lap at Burnenville I came up on Jo Bonnier, who was tooling along at about 10 mph, with a wheel hanging off. I had to back right off - which allowed Surtees to get a tow on the long climb towards the end of the lap. He came by me, and there was nothing I could do about it.

"I guess nowadays someone in that situation would just block the other guy, but we didn't do that sort of thing back then; it would never have occurred to any of us. Leaving aside any question of ethics, if you touched wheels with someone at most of the tracks we raced on, chances were you were going to hit trees or a house or something..."

On lap eight Amon's ill luck struck again, and a stone flicked up by the Honda penetrated the Ferrari's oil radiator.

"It happened at the Masta Kink, of all places, which was pretty well flat. Immediately I got oil on my rear tyre, and what happened next was one of the most horrific bloody experiences I ever had. I came out of the kink sideways - at close to 180 mph, I suppose - and I've absolutely no idea how I ever got it back. After that, all I could do was park the thing. I was furious that day! If it hadn't been for Bonnier, I really don't think anyone would have seen me..."

Surtees led until the Honda broke a suspension bracket on the 11th lap. Denny Hulme and Jackie Stewart traded the lead back and forth until Hulme broke a driveshaft. A brilliant drive by Stewart came to naught when his Matra-Ford ran out of fuel on the last lap, and Bruce McLaren won the only Grand Prix he would ever win in his own car.

It was also the first ever Grand Prix win for a McLaren car.

- Chris Amon quotes excerpted from Nigel Roebuck's column in the October 1998 issue of Motor Sport.

"Formula 1 cars are supposed to be difficult"

When I first drove a car in GPL, I was appalled at how difficult it was. This was back in late January, 1998, when Mike Lescault invited me to visit Papyrus. At the slightest provocation, the car would snap into violent oversteer. Matt Sentell acknowledged that it was difficult, but pointed out that you could catch the resulting slides if you knew what to do with the throttle. Mike Lescault remarked that he wasn't as good as Matt and Dave Kaemmer at doing this, but he was working on it.

Over the next few months, Mike improved, and started talking about how satisfying it was to have mastered the technique of driving in GPL. I had been practicing at home with pre-alpha versions since my initial visit to Papyrus, but I was still struggling. Mike was somewhat concerned about buyers having difficulty driving the cars, but the general feeling at Papyrus seemed to be that driving 1967 Formula 1 cars was supposed to be difficult.

Then the engineers implemented the Car Setup menu, and we could develop and store our own setups. We quickly discovered that there had been several fundamental things wrong with the default setup in earlier builds, including a rather large amount of positive camber on the rear wheels.

After a little development, the beta team began to share some setups with each other - and with Papyrus. With these setups, the car suddenly became much, much easier to drive, and a great deal more realistic. It was no longer necessary to catch the slides in every corner with just the right amount of throttle and opposite lock at the perfect moment to avoid several trips into the boonies every lap. Instead, the cars could be driven much like real cars, with a reasonable amount of power oversteer, and more progressive and predictable behavior. Soon we even developed setups that allowed trail braking.

Mike Lescault didn't like these setups at all. "Just when I'd mastered driving the cars the old way, now they've gotten so easy to drive that anyone can do it." He actually felt GPL was now too easy. He called the cars' new behavior "candy physics".

Although GPL got a little more difficult after tire temperatures were implemented, Mike was essentially talking about the ferociously demanding GPL we all know and love.

Mike Lescault would absolutely hate my new Ferrari setups.

Seeking balance

Last winter, I was discussing setups with the developer who was responsible for the default setups in GPL. I told him how much positive feedback I'd gotten from readers about my Coventry and Murasama setups, but he said he didn't like these setups because the front end washed out under power. He doesn't trail brake, and without trail braking and left-foot braking in midcorner, it's difficult to get the most out of my setups.

This comment stuck with me, and a few weeks ago, I decided to try to eliminate the midcorner understeer that has been characteristic of my setups. In the back of my mind, a theory had been forming. Was it possible that this understeer was caused by the front suspension bottoming on the outside bump stop, causing extreme weight transfer away from the inside front wheel and therefore terminal understeer?

I felt that I had been careful to eliminate the problem of the rear suspension bottoming on the bump stops, but in most of my setups I was running the front suspension lower and softer than the rear. I wondered if this could be allowing the front to bottom under some conditions.

In addition, I had noticed that the understeer seemed to be much worse on steeply banked corners, such as Mexico's last turn, and the middle of Abbey at Silverstone. Could it be that the high G loading in banked corners was driving the front of the car down on the bump stops, even under acceleration?

So I started experimenting. Taking the Ferrari to Silverstone, I raised the front ride height to be equal to the rear, and stiffened the front springs a couple of clicks. Sure enough, the understeer was reduced, but it wasn't eliminated. So I went still stiffer at the front and higher all around, to 3 inches. Better, but still some understeer.

I eventually got up to 95 lb/in on the front springs, and 115 lb/in on the rear, and had the front anti-roll bar up to about 160 lb/in and 130 at the rear. At these settings, the midcorner power understeer was reduced, but something else far more important had also happened.

Something magical.

Sweet red candy

At these settings - settings which, had someone shown me a setup like this, I would have rejected as impossibly high and much too stiff - the Ferrari behaved completely differently than before.

There are several places at Silverstone where I have always had to be very, very careful to avoid having the car get away from me. These included the entry to Copse (the first turn) and the entry to Stowe (the turn at the end of the Hangar Straight). At both of these places, if I was trail braking a little too hard, or didn't hit the entry just right, the car would snap into oversteer and would be very difficult to recover. I didn't really understand why this was happening, I just knew I had to be careful there.

With this new, radical setup, the Ferrari showed me why. There are sharp dropoffs at the entries to these corners, and the car slams down hard on the suspension right in the middle of the transition between braking and cornering. In my old setups, I had thought I had eliminated touching down on the bump stops, but I was wrong! Even at "smooth and flat" Silverstone, the old setups were allowing the car to bottom on the suspension, and bottom hard - so hard that the car was snapping into violent oversteer if I pushed it just a little too hard.

With the new setup, the car just glided smoothly over these dropoffs, and kept right on going, never deviating from the line I set when turning in. Suddenly I could concentrate on getting my turn in point correct, and drawing a nice smooth line down to the apex, rather than expending a lot of attention being ready to catch a violent slide immediately after I turned in. The difference in the experience was amazing.

I also have always had to be careful at the exit of Abbey, where a big bump tended to kick the car sideways, and I had to be careful putting the power down out of several corners, particularly Copse and Club, because the track had a change in camber near the edge which caused the rear tires to break traction under power. With the new setup, the car's behavior in these situations was transformed. The bump at the exit of Abbey barely unsettled the car at all, and the transition across the crowns, though still noticeable, was very much muted, and could be handled with just a tiny lift of the throttle and/or a touch of opposite lock.

Another terrific side effect was that the car was much less upset by touching a curb or putting a wheel off the edge. Granted, I still had to lift if I got a wheel far enough off on the outside, and the car would still spin if I bashed a curb hard enough, but the effects were much less violent and more progressive than before.

But what about lap times?

I was still a little concerned about my overall lap times, because I believed that the higher center of gravity would cause more weight transfer to the outside, reducing overall grip, and stiff springing would further reduce grip. So I started pushing a little harder, going for a time. Within only a few laps, I had completely destroyed my personal best at Silverstone, going into the 1:28's for the first time ever with the Ferrari. What's more, I was far more consistent with the new setup than I had ever been with the old; it was almost laughably easy to run mid 1:29's, whereas previously a 29 was a big event.

My Best Lap Times in the Ferrari

Kyalami 1:22.07
Mexico City 1:49.96
Monaco 1:30.61
Monza 1:29.54
Mosport 1:23.01
Nurburgring 8:45.69
Rouen 1:59.83
Silverstone 1:28.80
Spa-Francorchamps 3:22.07
Watkins Glen 1:05.89
Zandvoort 1:26.84
In only a few cases, I haven't yet bettered my previous best, but those previous bests occurred after hours of focus on a single track.

The Fun Factor

Although my lap times improved, the biggest difference was in the fun factor. Whereas before I was always fighting the car, having to be ready to catch it when the tail snapped out of line, or struggle to get it turned in the middle of some corners, now the car just glided smooth as silk around the corners, going right where I pointed it. When it hit bumps, I was still aware of them, not because the car snapped sideways, but rather by the vertical movement of the car on the suspension.

Does it work at other circuits?

I took the new setup to Spa, Zandvoort, and the Glen. These places have more severe banking and more severe bumps than Silverstone, so I found I had to go even higher on the ride height to keep that silky smoothness.

The difference in the experience was incredible.

For example, there is a tricky area just after the hump at the end of the back straight at the Glen, where there is a sharp crown that has to be transitioned while braking and turning into the Loop. I've always had to be very, very careful here, because the tail was very prone to getting away from me during the transition over the crown, but with the new setup, this section of the corner became a non-event. The car just went over the crown and kept on truckin'.

At Zandvoort, all those places where the car used to snap sideways over bumps and in banked corners were not quite eradicated, but they became much more manageable. I could direct my attention to getting the line right rather than spending much of my mental capacity on being ready to catch the tail before it got away from me. Again, in just a few laps, I destroyed my personal best - a time which I had worked very long and hard to achieve - and I could run consistently within a half second of that time.

Then I went to Spa. Let me tell you, the experience of driving through the Masta kink with all four wheels suspended by their springs is a completely different experience than having the right rear go onto the bump stop halfway through the left-hander. It is awesome, delicious, marvelous - a glorious feeling.

Finally, I believe, I have stumbled on what has always been needed to experience GPL as it should be experienced. At last, the full and marvelous subtlety of GPL's incredible physics engine is fully revealed.

Ride Height in 1967

When I mentioned my radical new setup approach to Dave Kaemmer, he said that during his research for GPL, Ron Tauranac gave him very complete information about the 1967 Brabham.

Tauranac said that they used springs which gave wheel rates at the front of about 65 to 75 lbs./in., and a little stiffer at the rear. Except for the Nurburgring, they didn't change spring rates from track to track.

Instead, at each circuit, they would adjust the car's ride height so that it was as low as possible without bottoming anywhere. Dave recalled that this was typically around 4.5 inches, except at the Ring where it was around 5.5 inches.

I've examined a number of photographs of cars of the era in action and in repose - including a large poster of a 1968 Ferrari 312 which hangs in my home office. I could put my fist under the bottom of this Ferrari and have plenty of room to wiggle it up and down.

After examining many photos it seems pretty obvious to me that the static setting was typically close to the point where the halfshafts were parallel with the ground, which (assuming GPL's cars are physically correct, which I believe they are) would put them somewhere around 3.5 to 4.5 inches.

Under hard acceleration, the cars would squat so that the halfshafts angled down slightly toward the inside, to perhaps 2.5 to 3.5 inches ground clearance at the rear. (See Figure 1)

On reflection, it's obvious that the cars were designed to run this way. If the cars were run as low as we see them being run by many people in GPL, the universal joints would have quickly failed due to the stresses induced by running at such extreme angles.

Had Tauranac, Chapman, and the others wanted to run the cars very stiff and very low, they would have designed the transmission and driveshaft geometry to permit this. They didn't do so because making the suspension stiffer would have reduced the tires' grip.

Modern race car suspensions are extremely stiff only to cope with the effects of aerodynamic downforce. I believe that if wings and downforce-producing bodies were abolished, modern race car designers would return to softer springs and the ride heights to go with them.

How did we get so badly lost?

Right from the start, everyone - including me - has been operating on the assumption that lower is better. Lower means lower center of gravity, which in turn means less weight transfer to the outside wheels, and when the forces are more evenly distributed across the two tires on a given axle, there is more grip. Also, everyone knows that softer suspension gives more grip; every race car setup book states this flatly.

This is all true. But.

We were forgetting one thing: when the car's suspension bottoms on the outside suspension alone, there is an abrupt and extreme transfer of weight away from the inside wheel to the outside wheel. Also the tire on the outside wheel suddenly becomes the main springing medium, distorting it. All of these things reduce grip.

In addition, if the suspension at only one end of the car bottoms, the effect is the same as suddenly installing a very, very stiff anti-roll bar at that end. If it's the front, the front end washes out and won't stick again until the G loadings come down enough to allow the front to pop back up off the stop. If it's the back, the tail snaps out into violent oversteer that is very difficult to recover from unless it is caught almost instantly. The roll forces from the resulting slide nail that corner to the bump stop, and the tail loses so much grip that a spin is almost inevitable.

To look at it another way, the moment the suspension touches down on a bump stop at any corner, all of our carefully selected spring, damper, and anti-roll bar settings go out the window, their effects instantly overpowered by the much higher spring rate of the bump rubber.

Another thing to think about is that every time you have to wind in opposite lock to catch a slide, you're giving up lateral position on the track, reducing your ultimate speed through the corner.

Also consider what happens when you put a wheel up on a curb. If the suspension is already on or near the bump stop, and you lift up one corner, that corner and the kitty-corner opposite corner are going to go onto the bump stop - hard. For example, if you run over the curb at the apex with the right front on a right-hand corner, the left rear is going to go onto the bump stop, transferring weight away from the right rear and snapping the car hard right. The same thing will happen if you run up on the curb at the exit, for the same reason.

But why didn't anyone notice this before?

I believe that we've overlooked this because of one vital aspect lacking in any sim: seat of the pants feel. In a real car, we'd be feeling the vertical G's every time the car slams down onto those hard little bump rubbers. The shock would be transmitted right through the chassis to the seat bottom. We'd do something about it pronto.

But in GPL, we don't feel those vertical G's; instead we just get the car doing something twitchy and we think it's just that we weren't smooth enough or something, and try to do better next time.

We've all been driving around a serious setup flaw for months!

C'mon! You really expect me to believe that?

If you are skeptical about all this, watch a replay closely. Pick a replay that isn't your smoothest lap, because smoothness minimizes the effects of bottoming - but it's very difficult to drive consistently at the necessary level of smoothness.

Remember that the minimum length for bump stops is one inch. Dave Kaemmer recently confirmed to me that when the car is about one inch off the ground, the suspension contacts the top of a one inch bump stop. Keep in mind that at a length of one inch, the spring rate of the bump stops is very, very high - many times higher than that of the springs at their stiffest setting.

Now watch how many times the bottom of the car goes down to one inch or less above the ground. And remember that every time it does, the spring rate has just gone up dramatically at whichever corner of the car that's down at one inch or less. Notice how often the rear of the car goes down by itself, while the front stays above one inch - and what happens to the car's direction of travel when this occurs.

In my humble opinion, the tradeoffs resulting from low ride height simply aren't worth it. By getting the car up into the range in which the suspension was designed to operate, and keeping the car off the bump stops (except in extreme situations such as those encountered at the Ring and the chicane at Monaco) I found the car's behavior is transformed.

Now it drives like a car, not a like a wild animal.

Ok, maybe it's still a wild animal, but now it's a sane wild animal instead of a crazy one.

Ferrari 312 at Zandvoort

Figure 1. A 1968 Ferrari 312 under hard acceleration uphill out of Hugenholzbocht, a slow corner at Zandvoort.

This condition will tend to maximize squat. Note the ground clearance and the driveshaft angles.

Click on the photo to see an enlargement.
GPL is not difficult to drive!

What would have happened to GPL's sales if it had shipped with default setups like these? Would it have gained such a reputation for being difficult to drive?

I don't think so.

As John Wallace recently commented to the GPL beta team, these cars certainly are not easy to learn to drive, but once you've mastered the art and discipline of throttle control, and have adjusted to the level of cornering and braking grip available, they are simply marvelous.

Set up properly and handled with respect, the Ferrari is very, very easy to drive.

And oh, so sweet.

GPL's Ferrari

Aside from the small wing which Amon elected to use on the Ferrari at Spa in 1968, the Ferrari 312 modeled in GPL is essentially the car which he described above. Folks who have driven GPL's Ferrari will recognize the uncanny way in which the simulation matches the description of the real thing: its superb handling, the delightful way it can be steered with the throttle - and its lack of grunt compared with the Honda and Cosworth engines.

With the shortest wheelbase of all the cars in GPL, the Ferrari is characterized by its delicious responsiveness to longitudinal weight transfer. It has a marvelous ability to be pivoted around the front wheels with a touch of brake. Blessed with light weight and superb suspension characteristics, it telegraphs to the driver exactly what it is doing in response to control inputs. It is very sensitive to the throttle, and its responses are smooth and progressive. It is quite "talkative" when slithering over bumps and scrabbling for grip under hard acceleration.

Despite its brutal power, the 312 is a finesse car - a car for the artist at the wheel.

Being a Ferrari, the 312's defining characteristic is of course its engine. This is actually a wonderful engine, with a smooth, broad power band and plenty of punch. At its 10,500 rpm redline it's only 5 hp down on the Eagle, and has 10 more prancing big ones than the Cosworth. It also may lack a little of the midrange grunt found in the Weslake and Cosworth.

The Ferrari is a couple of miles per hour down on the Eagle and Lotus in a straight line, which suggests that perhaps that terrific chassis is just a wee bit greedy in terms of drag. This puts the red car at a bit of a disadvantage to the Eagle and Lotus at Monza and possibly Spa, but anywhere else, the Ferrari is a weapon to be reckoned with.

That mid-corner understeer

So did I ever cure that mid-corner understeer?

Not entirely. In fact, all else being equal, raising the ride height and stiffening the springs and dampers seems to make the car understeer a bit more. I suspect this was probably because I was going somewhat stiffer proportionally on the front springs.

While developing this collection of setups, I initially kept my original anti-roll bar relationships; I went higher overall, but if the front was three clicks stiffer in my old setup, I kept it three clicks stiffer in the new setup. I found myself actually hurling the car into some corners, using the rally-driver technique of rotating the car slightly to the left before turning to the right for a right-hander, to get some polar momentum going and help get the tail out.

Eventually I realized that this was really too much work, and I started changing the balance toward less understeer. As I did so, I found I didn't need to work as hard at the wheel, and I seemed to be able to go just as fast, maybe a little faster.

I also realized something else. Previously, I'd favored strongly understeering setups because my reflexes simply aren't fast enough to catch all the little slides that the bumps threw the car into with lower-riding setups - slides that quickly became big, unrecoverable ones when I didn't react quickly enough. And working so hard to catch these slides caused me to quickly fatigue.

Divergent Stability

An airplane is said to be "stable" or "unstable" depending on its behavior when deviated from flying straight and level.

For example, when the pilot deflects the stick to the left and rolls the airplane into a 30 degree bank, and then releases the stick, will the airplane return to level, stay in the bank, or increase its roll angle into a steeper bank?

If the airplane returns to straight and level flight, it is considered to have "positive stability". If it remains in a 30 degree bank, it is considered to have "neutral stability".

If, however, the angle of bank increases, the airplane is considered to be "unstable" or to have "negative stability".

If the angle of bank increases at an increasing rate, the airplane is considered to be "divergent"; that is, not only is it unstable, but the more it rolls, the faster it wants to roll.

At the very best, divergent instability causes increased pilot workload, and at worst it can be very dangerous, making the aircraft unflyable. Divergent stability is considered very undesirable in aircraft and designers are careful to avoid this characteristic.

The same concepts and terminology can be applied to cars.

GPL's cars, when riding low enough and soft enough to contact the bump stops, exhibit divergent behavior in terms of slip angles. Once the car's rear tires go beyond a certain slip angle (and the outside rear suspension contacts the bump stops) the car has an increasing tendency to increase its slip angle until it - rather quickly - spins out.

As with airplanes, divergent instability in race cars is very undesirable.

With the new ride height and stiffer suspension, the tail now is much more "catchable"; its breakaway is far more progressive, and doesn't become divergent after reaching a certain slip angle. I can now ride much closer to the peak of the slip angle curve and get away with it.

So, despite the rather leisurely reflexes with which I've been endowed, I can handle setups which are biased more toward neutral. I can balance the tail on the throttle, rather than letting steering wheel angle tell me when I'm on the limit.

The car still understeers in the middle of many corners, but in most circumstances the understeer much reduced because I'm running a setup that's more balanced overall. And the front end doesn't wash out entirely in banked corners under power as it did before.

Other parameters

If you closely inspect these setups, aside from the ride heights and stiffness, you'll find little remarkable about them beyond careful tuning - and an almost complete departure from anything asymmetrical. I believe the only remaining asymmetry is in the rear camber at Monza, and that's because this helps reduce the understeer through Ascari while perhaps helping grip a little through the right-handers.

Tire pressures. There seems to be a passion for "zeroing the tire temps", tweaking tire pressures and cambers to get the temps even all across the surface of each tire.

I tend to disagree with this. It requires running the front and rear tires at or near the same pressure, and I've found that this seems to make the car more unpredictable. I simply can't get the car to be as consistent with equal or nearly equal tire pressures front and rear.

After a lot of experimenting, I've settled on 19 psi front, 22 psi rear, and I'm using these pressures for all these setups.

Camber. I've found that using asymmetrical camber destabilizes the car under braking. It doesn't seem to really add enough lateral grip to offset this, except perhaps at Monza.

I've stuck with .75 degrees negative camber in the front, and .25 negative in the rear. This seems to give good turn-in and excellent stability under braking. Although the outside edge of the left rear tire will sometimes run a degree or two higher than the inside edge, keeping the tire flatter to the road seems to result in improved traction under acceleration, as well as improved braking stability. These seem to offset any slight reduction in lateral grip.

Brake Balance. I've had to move more toward the front on brake balance, typically using 58%. This is because the higher ride height causes more weight transfer to the front. Again, the tradeoff is more than worth it, in my opinion.

Dampers. In general, I've found that bump settings of 2 or 3 clicks, and rebound of 3 clicks works best. Softer and the car wallows around and gets too unsettled when it hits a curb; stiffer and it tends to dart and weave around over the bumps and also gets too unsettled over curbs. I feel that the gradations between clicks is too coarse to do much fine tuning, such as asymmetrical damper settings as used currently in Champcars.

One interesting thing I noticed was that once I started keeping the car off the bump stops so the springs were doing the suspending, I needed softer dampers (2 clicks on bump) for Spa. However, this setting didn't work for Kyalami; I needed the dampers to be one click stiffer (3 clicks on bump).

I believe that this is due to the difference in speeds; at Spa, where cornering speeds are high, the bumps accelerate the wheels at a higher rate, so softer damping is needed to avoid making the car bounce around. At Kyalami, where cornering speeds are low and the track is relatively smooth, stiffer dampers help make the car more responsive.

On the other hand, at Zandvoort, which is one of the roughest tracks, I had to use 2 clicks on bump even though the cornering speeds are relatively low; otherwise the car is too nervous.

Gearing. Some time ago Dave Kaemmer suggested that using taller low gears and putting the top gears closer together would be beneficial for lap times. After all, any power in excess of that which overcomes the available traction (i.e. incites wheelspin) is wasted. On the other hand, using high gears that are relatively close together is useful for high speed acceleration because it allows you to stay in 3rd and 4th, near the power peak and taking advantage of greater torque multiplication, for a longer time on long straights.

Recently when I was testing a Eagle setup for Spa which was developed by Wolf Woeger, I noticed that Wolf was using this type of gearing. I put Wolf's gearing on my own Eagle setup and went two seconds faster immediately. I've also tried this type of gearing on the Brabham at Spa and found a quite noticeable improvement.

I've used the same philosophy on many of these Ferrari setups. In addition to helping acceleration at high speeds, this makes the car still easier to drive, because with less of an excess of torque in the lower gears, it's easier to manage the throttle to keep the tires just on the limit of adhesion under acceleration.

Oversteer, understeer, or neutral?

Naturally, these setups are tuned for my driving style. I trail brake, and I use left-foot brake in the middle of the corners to steer and tuck the nose. I've always liked to feel the limit of grip in mid-corner with the steering wheel, not by having the tail come out. This has dictated an understeering setup.

I also felt that an understeering setup allows me to get the power down earlier and better during the exit phase of the corners, giving me higher speed out of the corner and down the straight.

However, not everyone uses this style, and many of the most skilled drivers use a more neutral setup and balance the car on the limit of rear tire traction throughout most corners. As I mentioned above, I've begun to move in that direction as well.

To adapt to these new, more neutral setups, I had to be smoother and less aggressive when turning in, and be ready to catch the car when the tail starts to come around. Once I developed these habits, I found the more neutral balance of these setups to be very comfortable and a great deal of fun.

Tweaking for your style

I feel that these setups are readily adjusted to suit any style. First, be sure you have the tires good and warm. It takes one to three fairly brisk laps at most circuits to bring the tires up to temperature and have the car's handling stabilize to a mild understeer. When the tires are cold, the car will feel much "looser"; it will tend to feel twitchy and the rear end will feel "slidey".

Spooled vs. Open Diffs

Dave Kaemmer commented:

"One thing I noticed in the article (now this is picky, and it's the only thing I read that made me stop and think about disagreeing) is that you said "to make the car looser...take out a clutch". Taking out a clutch should make the diff lock up less, which, all else being equal, should give less limit oversteer, although I must concede that it would give more oversteer up to the limit.

"Think of it this way: the limited slip diff is designed to allow you to select behavior which lies between the extremes of the spool and the open diff. A spool gives equal wheel speeds, and unequal forces at the contact patches (roughly proportional to the load on each wheel). The open diff gives unequal wheel speeds, but equal forces at each wheel (no net yawing torque). For a limited slip diff somewhere in the middle, anything that tends to unlock the diff will move its behavior towards that of an open diff, whereas anything which gives more lockup will give more spool-like behavior. All else being equal, a spool will exhibit more oversteer (near the limit) than an open diff."

Note that in a Salisbury-type differential such as those used by GPL's cars, a higher ramp angle (e.g. 85 degrees) gives less locking, while a lower ramp angle (e.g. 30 degrees) gives more locking. On the other hand, higher preload - more clutches - gives more locking.

With the tires warm, if the car understeers too much for you, take two clicks out of the front bar and go up two clicks on the rear bar (don't go softer on overall roll stiffness unless you also go higher on ride height). You can also cut rear toe, and take out a clutch or two. [Dave Kaemmer disagrees with this; see sidebar.] If you don't trail brake, go a click or two to the rear on the brake bias.

If the car still understeers too much, go to 85/45 on the diff. If you're an oversteer junkie, 60/60, one clutch, and anti-roll bars with about 40 to 50 lb/in higher stiffness rear vs. front will give you the wild ride you crave.

I've included examples of "pushier" and "looser" setups for Silverstone, along with a setup for the Advanced Trainer at Silverstone. Even though it doesn't suit my style, I went almost as quickly with my "loose" Silverstone setup as I did with my preferred setup.

I've also included "pushy" setups for several of the other tracks, for those who simply aren't comfortable with my more neutral setups. If you want more understeer at other tracks, go softer on the front bar and stiffer on the rear, and add some rear toe.

Keep in mind that since styles vary, the setup I call "Loose" may still be a little "pushy" for some people, but going a little farther with the bars, toe, and diff will bring the car to you.

I have to admit that the looser setups, although they take more delicacy than I'm used to during turn in, have a delicious feel once they start to slide. You can really hang that tail out and keep it there. There's a delightful slithery, flowing feel to the way the car transitions from understeer to oversteer and back, especially under power.

Going lower and softer

Once you start tweaking, you will almost certainly be tempted to lower the ride height and soften the springs and/or bars.

Don't do it!

At least, give yourself twenty or thirty laps at a given track with the setup at the height and stiffness I am using. Get used to how the car feels when it's flowing over the bumps rather than bouncing off them. Then, if you like, try going a little lower and/or softer.

I am pretty sure you won't like what you find. Personally, I can't believe I've been running such unstable setups for so long, believing they were good, believing the only deficiency was in my own inability to react instantly to bump- and banking- induced divergent oversteer.

So what do other people think about these setups?

Your Ferrari article is great! I had never seen Chris Amon's quote before--I couldn't believe how good a description it was of the GPL Ferrari! I am glad to see that all that data we tracked down was for the most part right on. [Your] setups do indeed handle better than low riders.

- Dave Kaemmer

Wolfgang Woeger recently posted a 3:13 replay of a hotlap at Spa, along with his his setup. After watching his hotlap, and adopting his gearing, I knocked about two seconds off my previous best at Spa. Then I sent Wolfgang a pair of my Eagle setups (built with the same philosophy as these Ferrari setups) and my original high-flying Ferrari setup for Silverstone.

Here's what Wolfgang had to say:

I did about 15-20 laps [at Spa with your Eagle setup] with the slight change of tire pressure rear to 21 and the brake to 55. That way I could drive easy laps in the 3:15; best was a 3:14:8x. It was a very good lap, so I would say that I'm 1 second slower than in "my" setup.

BUT: The feeling was VERY different. I could run comfortable through corners that otherwise gave me a heart attack when feeling a bump. With this setup I could accelerate much better out of the slow corners; the slides were way easier to control.

Next track I tried was Zandvoort [with your] Eagle. And now I couldn't believe what I felt and saw. IT WAS LIKE ZANDY GOT A NEW ASPHALT AND THE BUMPS WERE GONE!!!! [Wolf's emphasis]

Wonderful, Alison! Soooo cool to drive....I did about 30 laps here and the best was a 1:24:2x. I did with another setup already a 1:23:98; Greger Huttu did a 1:23:8x. But needless to say it can't compare the ride over the bumps of a high 1:23 to low 1:24 laps with the comfortable driving - VERY impressive, it really lead me to rethink my setups.

The 3rd I tried was then Silverstone [with the Ferrari]. I did the brake to 56 and had such a good drive there again - my best lap out of about 35 laps was a 1:27:03. I was never before in the Ferrari here, but I know that another Elite driver has a 1:26:7x here, and I should be able to be 0.3 - 0.5 sec. faster than him. I had little problems here in the braking zones to keep the car straight; maybe it's just a Joystick handicap, but normally I don't have it. In accelerating out of ANY corner-type it was wonderful again.

Thanks, Alison, for giving me some really new perspectives on setting up the cars! :))

Later, Wolfgang added:

I did a 3:13:98 with a setup that used a static ride height higher than 2.5 at Spa in the Eagle. Also Ian Lake did some testing and is in the mid 1:23 in the Lotus at Zandvoort [with 2.5+ ride height], and we both think that after more testings we are 0.3 to 0.6 seconds away from the fastest laps now.

- Wolfgang Woeger

Steve Smith, on my original high-flying Ferrari setup for Silverstone:

Holy moly! That's the finest setup I've ever driven in GPL, my own included. It's as easy to drive as an F2. A keeper. Congratulations.

- Steve Smith

Fantastic job; you have a way to describe all parameters in a beautiful fashion. [Your Silverstone setup] is very forgiving in almost all aspects.

- Eric Cote

I tried some of Alison's new Ferrari setups tonight for the first time. Alison has come up with some revolutionary thinking on setting up your GPL racer. I took her loose Ferrari setup at Silverstone where I have a lot of track time in the Ferrari. Slapped my drive train on it and did a 1:27.21 in just 10 laps. Only 2 hundredths off my PB.

What is remarkable about these setups is the grip out of the corners (the secret to doing fast laps IMO). This must be mostly due to the symmetrical suspension setup. The car seemed very stable all around. I felt like I was cheating. <G>

All this with the ride height of 3.5 inches! So that's what the top of my tires look like. My only draw back was my driving style. When I pushed it hard tire temps became unacceptable. But with cleaner driving it would be much easier to be consistent and faster with her setups for the average driver.

Check out her site and read what she has to say about setting up your car; it will open your eyes to some new ways to set them up.

- Chris Moses

Great setups, Alison.

I just tried [your setup] on the Mexico track which is one of the hardest tracks for me, and after 10 laps I had a 1:49.86 lap. That's over a second better than my previous best lap in Lotus. It's all Ferrari for me from now on.

- Jan Otto Ruud

Alison has done a great job on this and has added a lot to the GPL community. These setups of Alison's are easy to drive and do not exhibit the normal understeer of Alison's early setups.

That was the one thing I never liked about Alison's early setups, but these are very well balanced and are a must have addition to any F1 lover's garage.

Well done Alison!!

- Ron Ayton

I have waited for these (Ferrari) set-ups for a long time. I first 'discovered' your work on the BRM set-ups and was (almost) instantly a success in a car I had until then been very poor in at best. Thanks for all your hard work! Now, at long last the Ferrari is eminently drivable...

- J M Tinney

Thanks for all the work you put into the setups you do! The Ferrari is so much easier to drive and it's still VERY fast. All my best laps are in the Honda/Murasama with your setups. I think I'll drive a Ferrari for a while ;)

Thanks again!

- David Cook

Thanks for the Ferrari setups. My previous best at Kyalami in the Sweet Red Machine was 1.21.80. This was after hundreds of laps. I ran seven, yes seven, laps with your setup and put up a 1.21.52. As an added bonus, the car was about as difficult to drive as my Ford Ranger. I don't know whether to laugh or cry! :) Thanks again.

- Berry Kaiser

Okay, I have a couple of replays in my GPL file [named] something like "luvualison" (blush) from back in October or November when I tried your Cooper setups and GPL became INSTANTLY more accessible.

I just now pulled out of the pits at Spa in an "Alison-tuned" Ferrari. My pb for that marque had been 3:32.78. On lap 2, I turned a 3:29.65, and on lap 3, a 3:28.69! Going through Masta is almost like pulling into the driveway!

I predict we'll see a lot of red on VROC this weekend!

- Kurt Steinbock

Wow.... I don't know what to say.... the Ferrari's easier to drive than my XR4Ti.....

I'd never thought of the bump-stops giving so much trouble!

- Mark Stahl

Well done again Alison...more interesting reading and great new setups to boot! Keep up the good work :-)

- Ian Robertson

Alison, I tried your new Ferrari setups today. I can describe my inital thoughts in just one word: "WOW". The car handled beautifully yet didn't lose any speed at all almost with the higher ride height. I was able to do a fast lap (for me anyway) at Silverstone without spinning in just my 5th at speed lap around the track. I was able to run my quickest lap ever at Zandvoort within about 4 laps and then smashed my old record later. I can't wait to try the other tracks now! Thanks so much Alison! Great going!

- Dan Belcher

Woohoo! Just went into the 26's with the Honda at Monaco; this is a feat I know... BUT...... I did it with a 2.5 setup!! I'm 0.3 FASTER at Monza..... the ONLY place I am losing time so far is Spa (oh I miss the sparks through Eau Rouge), where I am still 1.4 off my personal best.

HAHAHA THANKYOU ALISON! Read what this gal says, she knows her stuff!!!

- Tim Wheatley

What's next?

Well, of course there are setups for the Eagle to do, and the Lotus, and bringing my earlier setups up to date with these new discoveries. Preliminary testing suggests that the Eagle responds extremely well to settings virtually identical to these Ferrari setups, and the Brabham also works well on settings that are quite similar. The BRM gets a very noticeable improvement from higher ride heights and stiffer roll bars. The Coventry and Murasama don't seem to get as much improvement, but I will need to do more experimenting with them to be sure. The Lotus improves but I haven't yet found the settings to give it the right balance; with settings like I'm using for the Ferrari it seems to have terminal understeer.

And of course, the Trainers also respond to this new approach. Preliminary testing suggests that the Advanced Trainers become very docile and forgiving, but require somewhat different balance than the GP cars.

Dave Kaemmer suggests some further avenues for experimentation, which I fully intend to explore:

I wonder if you couldn't go further. Another half inch of ride height would allow you to soften the springs, which should help grip in high speed corners. It's an interesting avenue...

I still think the 85/30 diff is an abomination... Now with your new higher ride heights and the taming of bump stop induced panic, perhaps it is finally time to explore 45/85's and other more normal diff settings. For sure the 85 coast side gives a feeling much more like the Barber Dodge cars, where you have to be very careful with throttle and brake release at corner entry. [See Dave's explanation of differential behavior in the sidebar.]

After trying these setups, Eric Cote proposed some interesting ideas and sent me a Silverstone setup developed from my own setup that has really made me reconsider some of my other assumptions, such as my use of 22 psi rear tire pressures and my use of 4 clutches in the diff.

Eric has also challenged my neglect of dampers as a tuning tool. Eric says, "Since the Ferrari doesn't run on the bump stops anymore, the fine tuning of dampers in corner transition (refer to [my] damper table) actually works beautifully; in previous setups (low rider type), when running on bump stops, changing dampers had little effect [or the effect was] non-existent."

Eric has updated his damper table, which he originally developed for GP2, so that it is now even more applicable to GPL. Check out Eric's comments in detail here.

I feel that, in a way, the whole ball game has started over. Now that we know how to keep the cars off the bump stops - and we have good evidence that this works and is a desireable thing to do - there are a great many new things to try, new directions to explore.

Amazing, isn't it? Almost nine months since GPL's release and we're still discovering new things, breaking new ground.

That Ferrari V12 wail

The Ferrari sound that comes with GPL is great, but unfortunately it isn't really a Ferrari and it lacks that distinctive metallic tone that seems to characterize all twelve-cylinder Ferraris. Steve Smith pointed out to me, "For some reason, all Ferraris have that sound, and nothing else sounds quite like it." After hearing a pair of 3-liter twelve cylinder Ferrari F1 cars at the Glen last summer, I have to agree.

For the purist, there now available a more authentic sample. A very clever fellow named Patrick has adapted a recording of Nick Mason's 1970 Ferrari 512S V12 to work with GPL.

You can download Patrick's excellent sounds from his site. I highly recommend all of them. Patrick's Ferrari sample is quite good. The pitch of the 5-liter sports car engine is a bit deeper than a 3-liter, but it's got that essential Ferrari V12 sound.

Using the same techniques pioneered by Patrick, Thomas Brown has developed a superb new sound for the Ferrari, taken from a 3-liter V12 312P Ferrari sports car of the same era as GPL's Ferrari 312, and it has that fabulous Ferrari sound. It's the most authentic-sounding and best sound for the Ferrari yet, IMHO.

You can download Thomas Brown's scintillating sounds from the Fast Lane.

With a good sound card connected to a good home stereo system, and the volume on "realistic", this Ferrari sound is superb. (See the sidebar on my BRM page for tips on setting the volume.)

The setups

Alison's Ferrari Race Setups (06/10/99 - 5 kb) - Race setups for the Ferrari.

These setups go into the setups folder under your GPL players folder. For example, my setups folder is:


This setup collection was zipped with the setup path included. You can simply unzip the entire archive into your player's folder.

There's also a GPL export HTML file with my best Ferrari times.

Note that these are setups for the GP car. It is possible to use them for the Trainers, although you'll need to rework the gearing.

To use a GP setup for a trainer, you simply need to change its extension. GP setups use an extension ending in 1, while the Advanced Trainer and Basic Trainer's setups end in 2 and three, respectively.

Thus, to use my Ferrari Monza GP setup for the Advanced Trainer, create a copy of Ali_Fer_Monza.fe1 and rename it to Ali_Fer_Monza.fe2. Then revise the gearing to be appropriate for the lower redline of the Advanced Trainer engine.

I've also found it useful to change the balance a little more toward neutral for the Trainers. For the Advanced Trainer, try taking out one clutch from the diff, going down a click on the front ARB and up a click on the rear, and down a click or two on rear toe. You may need to go even further for the Basic Trainer.