SVT engine performance-Solutions!

[JohnK]

Brian,

What you are going through is exactly what a bunch of Superformance S1 owners experienced. These cars came with a Webcon system (Weber throttle bodies and Alpha (Magnetti Marelli) computer) They came from the UK with the calibration specific for these engines (the Mondeo in England / the US Focus). They had stock and an uprated calibrations for two different engine configurations. Both calibrations ran AWFUL. However, because most of them were stock level configurations, they weren't so touchy - as you seem to understand, the more highly tuned the engine, the touchier it will be and the more extreme any problems will be; i.e., terrible drivability.

I had the problem of tuning cars with this system dropped in my lap - figuring out the software, how to use the dyno, what to change in the calibration, how to develop a test protocol . . . I happened to have the right collection of skills and a great deal of luck and I got it to work before I was thrown out of the shop for making too much noise and taking too long. It was really challenging

 

Most relavant to your situation is that the calibration that I came up with was very very different than what the car was delivered with. I even spoke with the Webcon people (talk about an expensive phone call, at 3AM no less). They expressed considerable disbelief that their calibration was so far off. From what I learned, the tuning shops in England (unlike what I've seen in the US) are set up to do this kind of tuning and the tuners there really do understand the problem of howto get great drivability using an elextronic EMS- everything I learned convinced me that they know their business. So then, why doesn't their calibration work here? No one ever came up with an adequate explanation. But the fact was that the calibration they developed in England did not run here. The Webcon system was capable of delivering pretty remarkable performance from a stock engine when a good calibration had been developed - ask S1Steve about his experience at Pocono. The person I spoke of in my earlier post drove down from Philly to have me tune his car as a result of driving his hotted-up S1 on the track with S1Steve and seeing how S1Steve's car ran compared to his. He had the uprated engine and was amazed to see how far off his fuel was when we put it on the dyno - and he had built racing engines.

 

If you're running the Ford ECU, it expects all the pieces to be of a particular set: the injectors, the MAF, the idle air control, the ignition control unit, ... IDENTICAL for the ECU's hardware and firmware. The Ford system calculates, based on how many pounds of air are coming into the engine at the relavant instant, what the RPM is, what the trottle position is, what the throttle's angular velocity is, what the coolant temerature is, what the incoming air temperature is - and uses that information to calculate on a per-cylinder basis, how much fuel to deliver into the cylinder that's next to fire and at what degree BTDC to fire the spark. This happens for each cylinder all the while the engine is running. If any one of the components is not part of the correct set, the calculation will be wrong.

I said this system, unlike any other, is adaptable. Since it is a real time system, since it is using the pounds of air entering the engine instant by instant, and since it uses an alorithm, it doesn't care about whether there's an air filter in the induction or not, just how much air is available at the instant. I may not even care what the cam profile is if it's not too extreme. Within the limits imposed by the size of the injectors, you can get more power out of the system by improving things such as induction flow. But, this whole thing depends on all the parts of the Engine Managment System being a matched set.

 

A first guess is that getting the system to run well may be best done by making sure that the complete EMS has all correct and orignial spec components. This could be difficult but offers the greatest payoff - you should get really excelle

[JBH]

Before going too far and spending too much money, I recommend talking to someone that is a tuning specialist. I know Geoffrey Ring at Racetek Engineering in New York. He is a Motec Dealer, but familiar with many ECU systems and quite capable of tuning a Caterham. He also teaches at EFI University which trains both amateurs and professionals in theory, equipment and tuning.

 

Give him a call to see if he knows someone in your area. If he doesn't, it might still be cost effective to pay for his travel and time. Rather than spending days on a dyno when only hours are required may in the end save you money.

 

Just an idea.

[11Budlite]

"I know Geoffrey Ring at Racetek Engineering in New York."

 

Thanks for that info JBH. At some point, I want someone that really knows what they're doing fine tune my Emerald ECU and he looks to be about 2 hours from my house. I'm also a relatively new 911 owner, and it'd be fun to check out his place. I still need to get in touch with Tim at Marren Fuel Injection (about 25 mins from my house) to see what he can offer for fine-tuning my setup.

 

Bruce

[athens7]

The holidays, and year end at work got me behind, but here are the results of my last runs. The VCT is restored, the original factory PCM program is restored, the cams have all been timed using factory timing tools. I left the adjustable exhaust cam gear on for future tuning, but it is set to zero. In effect, the car is as it was when I bought it. The first run is in this condition, the second with the performance tune that came from Diablosport when I bought the Predator, not a custom tune. The power output is as good as the custom tune from NC, without the driveability problems. The dip is still there, and the air/fuel ratio is still a little off at the torque peak, but my tuner feels both items can be improved with a little custom tuning. According to Ben at RMSC, limited slip diff equipped Sevens can expect about 19% drive train loss on a Dynojet dyno, so I'm making about 185 hp and 158 lb-ft of torque. This compares favorably with the factory SVT's output of 170 hp and 145 lb-ft torque (and my car's torque peak is achieved at 500 lower rpm). While this is still not 200 hp, as advertised by Caterham, it's pretty good. My choice is now to either install stock SVT cams and re-test, or leave this cam set in and tune for best results. The general consensus has been that the claimed 20 hp increase due to the Kent cams is optimistic, at best, as may be the claimed 10 hp increase from the K&N filter and the Caterham header. Most of the people with whom I've spoken believe 20 hp in total from all 3 mods is probably more realistic. If so, replaceing the stock cams will likely just be money spent to reduce power. Everyone who has looked at the motor agrees it is sound and defect free, and a body of opinion is forming that the dip is the result of the stock cam phasing not being correct for the new cams, a problem which may be addressable through PCM mapping. My guys also believe the smoother and stronger curves (as compared to the original run in NC) are due to the incorrect timing of the Kent cams when they were originally installed on my car, prior to sale.

One delightful side effect of the re-installation and timing of the cams is that the engine's cutting off on clutch engagement from higher rpms appears to have stopped! As this problem was the reason this whole quest began, it was indeed an exciting moment when I tested it and the car didn't die. My mechanic indicated that camshaft timing issues could have been the cause.

So, tune this set up, or go back to dead stock and start from scratch? As always opinions are welcome. I did pick up a new complete SVT cylinder head, and I have a set of stock cams. I'm playing with the idea (not right away) of porting this head, installing the stock cams, and having a complete second option to bolt on. The guy I bought the head from is an avid SVT buff, as well as a mechanic, and claims 210 hp from his similar setup. Upgraditis is a chronic disease!http://www.usa7s.com/aspnetforum/upload/36573913_Dec 19 2007 chart.jpg

[Mandurath]

While its still not the real driving season, I would get what you have running right. Sounds like you found some folks that can do it correctly. Upgradeitis is a serious disease and can only be cured by new partsicillin.

[Kitcat]

Why upgrade? You have a car now that can be reliably driven and enjoyed. It also has an astonishing performance envelop. Finally, it looks fantastic. If after a year you are tired of driving the fastest, coolest vehicle in your area, then upgrade.(IMHO).

[athens7]

Considering all the Sevens I see advertised with 200+ hp Zetecs (I have yet to see a dyno chart to support one of these), and the representations made by the sellers as to power, my primary goal was to ascertain whether or not my engine had a problem or problems, and to solve the cut off issue. I can certainly live with the power my car produces, especially if we can reduce the size of the torque dip. It's much more fun to drive the car than to throw money at it (I'm not a do it yourselfer like Mazda). Btw, 68 degrees and sunny yesterday and today. It sure was nice to get the top off and go for a couple of blats!:7fume:

Nice smiley, too!

[Al N.]

It's funny that you mention that Brad, because I too have never seen a 200 hp chart from a naturally aspirated Zetec in a Se7en. I've heard rumors of 220 and 240 hp cars, but never proof.

 

(Yes, yes, Mazda, we know what kind of power the LS-1 makes....)

 

Maybe 2008 will be the year that one of us gets to that point. Quite honestly, I'd like it to be me.

[Keith]

I'm a little late to this thread - but I'm not surprised. I've spent time on the track in the company of a "230 hp" SV, and in a straight line it has trouble keeping up with my 148 rwhp car.

 

Drivetrain loss isn't as simple as a percentage. A lot of people like to view it that way because it exaggerates higher hp numbers nicely. The truth is that it's partially a percentage and partially fixed. I prefer to use a fixed amount for a given drivetrain configuration instead of a percentage, mostly because I'd rather under- than over-report.

 

When testing a 1990 Miata, you see a 26 hp loss at the wheels when compared to the factory rating. 116 hp becomes 90. Grab yourself a 2005 Mazdaspeed turbo Miata, and 178 hp becomes 152. Hmm.

[Birkin42]

I would think that losses are a combination of those created by things that are not torque/hp dependant and those that are. For example, losses due to seal friction and circulating fluid would be primarily dependant on temperature and speed, but not on torque or hp. Then there is the losses due to things like the gear mesh which will differ with torque transmitted.

How much is due to each type I'm not really sure. I would think though that if you add 62 hp or 53% increase to an engine that the drivetrain losses will increase somewhat. I don't think they'd go up by 53% as well, but I wouldn't be surprised to it increase by say 20% (speculation on my part).

[sdca7]

I just got my car back from the dyno after a new engine rebuild today. I posted the results here:

 

http://www.californiacaterhamclub.com/chat/showthread.php?p=6699#post6699

 

 

I'm satisfied!

 

Justin

 

:7fume:

[slngsht]

great results.... opening the throttle all the way is definitely the cheapest mod

[JohnK]

I'm a little late to this thread - but I'm not surprised. I've spent time on the track in the company of a "230 hp" SV, and in a straight line it has trouble keeping up with my 148 rwhp car.

 

Drivetrain loss isn't as simple as a percentage. A lot of people like to view it that way because it exaggerates higher hp numbers nicely. The truth is that it's partially a percentage and partially fixed. I prefer to use a fixed amount for a given drivetrain configuration instead of a percentage, mostly because I'd rather under- than over-report.

 

When testing a 1990 Miata, you see a 26 hp loss at the wheels when compared to the factory rating. 116 hp becomes 90. Grab yourself a 2005 Mazdaspeed turbo Miata, and 178 hp becomes 152. Hmm.

 

I'll offer that this isn't about frictional losses, it's about how well the car has its engine's "drivability" developed. You may strap a car down on a dyno and fiddle with things in order to record high peak HP at wide open throttle ( WOT) and wherever your torque peak lies, but this is pretty much a laboratory exercise and doesn't represent real-world issues. The last several cars I developed calibrations for were assessed by the owners in terms of how they felt on the road - and I can't remember the last time one of my clients asked for a printout of a pull on the dyno. The Dynojet, which damn near everyone seems to rely on as a measuring tool, is suitable for cars with carbs and distributors and only gives you a snapshot of a condition when everything is in one particular state. It is not a useful tool for measuring and developing output when running an electronic engine management system.

Given current engine management technology, you can easily set things up so that you can FRY your engine driving down the road at 50 MPH. What this means is that, if you're just doing WOT pulls on a Dynojet, all you're tuning for is WOT. If you don't have a "load-bearing" dyno, and a test protocol for doing mid-range tuning, your car which measured 230 RWHP will most likely be an unmanageable slug that just happens to make power at WOT. Conversely, if you want performance and you're running an electronic management system, your tuner has to be able to set fuel and spark at optimums at each 500 RPM point on your power curve. If you can find someone who can do this, you will be dumfounded at the way your car behaves and you will most likely never even think about what it records on the dyno at the rear wheels.

Cheers,

[Keith]

There's a Rototest dyno sitting about 50' from my desk Interestingly, every single one of our customers asks for a dyno chart after tuning.

 

Driveability is harder than WOT peak power, but it's possible to have both. Of course, I deal almost exclusively with cars with modern adjustable engine management. So we expect full driveability on every car that goes on the dyno. Once you have that, you still concentrate on output - because given the same part-throttle behaviour, the car with more horsepower is going to be faster.

 

My example of a 148 rwhp car pulling away from a "230 hp" Zetec is under WOT conditions, of course. Neither car appeared to have any evil part-throttle behavior.

 

But driveability doesn't have anything to do with drivetrain loss really, which is what I was discussing. Here's an interesting chart for you:

http://www.flyinmiata.com/tech/dyno_runs/NC_sweep_times.pdf

 

That's the same car, tested in the same gear on two consecutive dyno runs with no changes in between. What's the difference? On one run, we set the sweep time to 15 seconds. The next, we let the car take 25 seconds to accelerate from 2000-7000 rpm. The difference you see is the inertial losses in my opinion, although we've had some good conversations around here about that.

[horizenjob]

A few comments and suggestions. If you find a dyno shop that seems to report

%10 less then others in the area, that's probably the honest shop. Taking the

measurement slowly is probably better. My mechanic used to emphasize reading

the area under the HP curve for the range your interested in. It helps even

out the little peaks and valleys and the area is what you use on the track.

 

I'm a little skeptical of the difference between crank and wheel measurements.

Something to wonder about, there are 750 watts in a horsepower. So if you

lose 26 HP in the drive train - that's about 20 kilowatts going into the diff and

transmission. That seems high, but maybe. An electric space heater is typically

1.5 kilowatts.

 

Also wondering if some of these transatlantic issues are caused by different

gasoline formulations? Do they use %10 alcohol in Europe?

 

 

[athens7]

Several trips to the tuners and $$$$ later, here I am again, a VERY happy camper!

 

http://www.usa7s.com/aspnetforum/upload/2092993627_Feb 11 2008 simple chart.jpg

 

This is the final run in February, after 7 hours of adjustments to the ECU programming. Our goal was to optimize the air/fuel ratio, and maximize timing wwithout detonation. We were also hoping to eliminate the dip in the power curves between 2500-3500 rpm. We discovered certain things:

-Adaptive fuel programming sucks. In my car's case, short term fuel trims were all too lean, leading to a gradual leaning out of the air/fuel mixture. Eliminated by resetting the ECU, the problem can reappear in as little as 200 miles. Eliminating the adaptive fuel programming eliminated the problem.

-The MAFS map in the ECU in its stock form doesn't fit the demands of the Seven. Changes to the air/fuel ratio settings didn't translate to real changes in the air/fuel ratio, until the MAFS map itself was changed.

-Heat soak in the engine prevented replicable runs. The second run was always down on peak power. The engine coolant temperature would get as high as 212 degrees.

The a/f ratio graph was smoother and more ideal, but persisted in being too lean at the torque peak. After the session, I found the car to have much better performance at both part and full throttle.

 

A couple of SVT technicians had advised me to install a 180 degree thermostat in the car, as opposed to the stock 195 degree unit. Thinking it would help with the heat soak issue, as well as durabilty, I did so, and returned to the dyno on March 18th. We also wanted to adjust the exhaust cam timing, and change the switchover point on the intake runners, as I had lost some midrange torque in the last session when we dropped the switchover point to 4100 rpm.

 

http://www.usa7s.com/aspnetforum/upload/1847002569_Mar 18 2008 before & after +2deg exh cam timing.jpg

 

This chart shows the first run of the visit, with the only change from February being the addition of the new thermostat. The later run is with the exhaust cam timing advanced 2 degrees and the intake runner switchover point moved to 5500 rpm. There is a HUGE gain in midrange power, and the elimination of the over-lean condition at the torque peak. We were actually able to increase the air slightly from 4500 rpm up. The dip in the power curves is also narrowed by about 400 rpm.

 

http://www.usa7s.com/aspnetforum/upload/226029952_Mar 18 2008 3rd & 4th gear.jpg

 

This last chart shows runs in 3rd and 4th gear. I show this for the "which gear yields the most efficient power" discussions.

 

All my runs were completely replicable in sets of three. I have gotten much feedback on the drivetrain loss issue for determining true engine output on a chassis dyno, most of it conflicting. Regardless of what loss exists, I know this: my motor puts out 3 more lbs-ft torque at the wheels than a stock SVT makes AT THE CRANK, and does it at 500 fewer rpm :hurray: ! Intake air temperature is also important; there was a noticeable difference in power at higher intake air temperatures. This is a problem with the SVT intake location. A cold air intake solution may be the next item on the to do list.I have now seen some other dyno runs for SVT motors with hot cams, and they all have a dip at some point in the power curve. The consensus among my sources is that the Cosworth/SVT head flows all the air it can, and volumetric efficency problems can crop up with more aggressive cams. The car goes like a scalded dog, and I believe it will be more durable with cooler operating temperatures and a more optimal a/f ratio. My motor sure has come a long way since October of last year! These mods are easily duplicated, and my tuner has the data to install the tune using a Diablosport Predator. Also necessary are the 180 degree thermostat and an adjustable exhaust cam gear. Your results may vary, but I would encourage SVT owners to watch the a/f ratios on your cars, if you are using the stock ECU. Hard running and lean operation can be an expen

[Kitcat]

Thanks for all the data. It is very informative. For a naturally aspirated 2.0 liter street engine, that seems like a lot of power. I don't want to think abt what all the additional development cost. Now you can get around to driving/enjoying it:).

[BusaNostra]

I never been a dyno queen person. What I do know, the owners who gets a lower horsepower will cry foul and blame it all to the dyno unit. All the conspiracy theory on particular dyno unit vs the other brand.

Those who gets a higher horsepower reading are convinced the unit is the most accurate brand. You feed the brain with a positive numbers and he'll be satisfied.

I've seen them all in the dyno shop.

 

I concentrated more in suspension setup. The seven is very critical in geometry for track or simply for main street. All those horsepower without a proper damper, spring rate, tires, geometry setup is USELESS to the 7. Like a rocket without guidance.

A highly suspension setup 7 will beat any 400hp rice & iron.

Edited June 20, 2009 by BusaNostra

[MHKflyer52]

I never been a dyno queen person. What I do know, the owners who gets a lower horsepower will cry foul and blame it all to the dyno unit. All the conspiracy theory on particular dyno unit vs the other brand.

Those who gets a higher horsepower reading are convinced the unit is the most accurate brand. You feed the brain with a positive numbers and he'll be satisfied.

I've seen them all in the dyno shop.

 

I concentrated more in suspension setup. The seven is very critical in geometry for track or simply for main street. All those horsepower without a proper damper, spring rate, tires, geometry setup is USELESS to the 7. Like a rocket without guidance.

A highly suspension setup 7 will beat any 400hp rice & iron.

 

So true and even more with a driver that is smooth and knows how to use the avaiable power to the best advantage.