Wicked Windsor In A Box: Pt 2

We left last month's story on the RDI 392 Street Pro with the engine ready for the dyno. At this point most project engines are loaded on the dyno, broken in, a few pulls made to verify the output, and that's it. There's nothing wrong with that, but let me remind you RDI's Preston Miller is an ex-Ford Motor Company development engineer with deep ties to Winston Cup. That's a business where remaining competitive means staying on the cutting edge.

A lot of being successful in Winston Cup is knowing how to build a lot of power reliably. With the power levels RDI engines were achieving creeping up and many motors going into road-race cars such as Cobras, Preston wanted to make sure that RDI was totally on top of crankshaft reliability. Consider what a busy drag racer may do in two years is, for a road racer, only equivalent to just a single practice session. This is where we pick up the story of our Street Pro 392 this month.

The dyno facility I elected to use for testing was Kenny Troutman's KT Engine Development in Concord, North Carolina. With more than 15 years of professional experience, Kenny is a dedicated circle track engine builder specializing in Ford-based Busch Grand National motors. His engines have taken championship honors in a number of the top classes and more wins and poles than he can remember. Not only does he build engines specifically for teams, but also runs an engine leasing service. I chose to use this facility not only because of Kenny's outstanding Ford know-how, but also because his dyno has tested a lot of crate motors over the last few years. This puts him in the position of knowing what is good and what is not. This knowledge alone was to prove a valuable indicator as to how effective our RDI unit was.

Break-In
The break-in procedure (on Valve-o-line mineral oil) was in part a build quality check of the RDI engine. On KT's dyno, oil is run through one of Holley's stainless steel screen micron filters. The advantage of going this route over paper filters is it allows you to see exactly what debris is being expelled from the engine and arrested by the filter. Half of what is arrested by a good paper filter is near invisible to the eye. The better the build quality, the less junk there will be in the filter.

After about an hour of low- to mid-speed cycling of the rpm at various loads and throttle openings, the engine was given a couple of pulls from as low as it would go to about 4,500 rpm. Even before setting up it was very apparent this was a stump puller with a good top end. At this point the Holley filter was opened up and inspected for debris. This proved a lot lower than I normally see. About 95 percent of what was apparent was lint from the parts cleaning paper. The rest looked to be the inevitable iron particles from the timing chain and gear break-in process. So far we were looking good.

Crankshaft Torsionals
Once the engine was broken in, ATI's JC Beattie went to work. JC's function in life, when not driving his oval-track car, is to service the needs of Winston Cup engine builders who require to have a damper specifically tuned to the needs of their race motors bottom-end combination. As RDI's Preston Miller points out, the ATI Super Damper is the overwhelming favorite among WC and BGN engine builders. Having a damper that really cracks down on torsional vibrations not only significantly extends crank life, but also adds power.

At first it may seem an unlikely piece of go faster equipment, but an effective crank damper cuts the amount of torsional vibration transmitted to the camshaft. In so doing it allows the cam to much more nearly deliver the intended lift/duration characteristics to the valvetrain. Forget light when it comes to dampers. Even for a drag racer, where acceleration is all that matters, a functional damper will out perform a light, non-functional one every time.

So how much power can a damper be worth? Forget about hearsay and some of the opinions of even well-known racers. Testing on my own dyno on a nominally 400hp motor showed a good damper was worth 5 hp over a typical factory damper (which was designed for the stock bottom end not a modified one) and over 11 hp over that delivered by a lightweight hub.

As far as measuring and refining the damper action was concerned, the plan was to have ATI build what, based on their huge data bank, they felt would be required for the longer than stock stroke (3.5 vs. 3.85) of the 392 Street Pro. Once the engine was on the dyno and broken in JC would set up ATI's torsional measuring gear, measure the crank's vibrations, then build the appropriate damper specifically for the Scat crank and rods, the Probe pistons, or rods and pistons of comparable weight.

Going into this test Beattie commented he did not expect to be far off the mark with the damper spec we started with. How right that statement was. We never got to see the selection process in action. Without a damper most cranks will suffer torsional vibrations up to a degree or more of double amplitude (+/- 1/2 a degree). Steel, as opposed to aluminum, can withstand a certain amount of deflection indefinitely. However, after a certain point the hours and minutes begin to count against it.

Within the industry it is generally accepted that a typical V-8 can, at what appears to be the most important third and fourth order harmonics, tolerate about 0.4 of a Degree Double Amplitude (+/- 0.2 of a degree) almost indefinitely. These third and fourth order harmonics also seem to be the ones that have the greatest adverse effect on power. The object of the exercise then is to damp crank vibrations to something below this 0.4 DDA threshold. As can be seen from the graph on page 159 the ATI Super Damper did just that. Below 4,750 rpm the torsionals exhibited by the Scat cast steel crank were very low. This was not unexpected as cast steel, like cast iron, has higher self-damping properties than are seen with forged or billet cranks.

The bottom line here is, so long as the engine spec does not allow excessive rpm and/or out-power the crank, a cast steel piece is actually better than a forged or billet piece and a heck of a lot cheaper. The numbers at the top end right up to our self imposed 5,800-rpm limit show the ATI Super Damper was more than up to the job of keeping torsionals under control with a maximum of 0.14 and 0.30 DDA for the third and fourth order harmonics respectively. Now we were secure in the knowledge our crank was likely to stay in one piece for a long time we moved on to our next phase of dyno test--namely carburetion.

Mega-Buck Winston Cup vs. Demon
Now for the carb test promised last month. As mentioned the plan was something a little different to the normal like-for-like shootout between a Holley and a Barry Grant Demon carb. The plan here was to pit the Demon against Kenny' Troutman's dyno shop Winston Cup 830 Holley. This carb was initially built by one of the country's leading WC carb builders and has since been progressively refined by Kenny during who knows how many circle track motor dyno pulls.

The first task was to ring out all the power from the WC Holley. This meant not only calibrating the carb spot on, but also the ignition timing. The speed of combustion changes as the mixture is changed. The fastest rate of burn usually occurs around about 14 to 14.5:1 air/fuel ratios. This means to see the exact nature of a jet change the timing also needs to be re-optimized. Timing then, should be considered a secondary part of any jetting exercise.

This being the case a few words on our Performance Distributors GM HEI distributor adaptation for SB Fords. First the custom mechanical advance curve built into it proved to be right on the money. Second the micro advance adjuster allowed for very small adjustments to tune for the absolute optimum timing. The procedure here was we jetted, timed and tested and slowly the power came up.

Being familiar with the idiosyncrasies of the Ford Racing/Victor Jr. intakes, Kenny was to use his experience to stagger jet the combination. Kenny's stagger jetting utilized three sizes of jets for the four corners of the carb. There is little point in my giving you those jet sizes, as they will be a little different for each motor. This aspect of the motor tune is just one reason why setting up on the dyno is a good idea. Granted at $750 a full eight-hour day (plus fuel) it's not cheap, but when you have an expert like Kenny on the job the dyno set up finds every ounce of power the motor is capable of.

But back to the plot: After dialing in the WC Holley's fuel curve to the best possible and following this along with exploratory timing changes in 1-degree increments we had a motor that would hit 499 point something hp on every pull--but never the magic 500. At 32 and 33 degrees of total advance just about the same results were delivered from this highly repeatable engine.

This factor in itself is worthy of mention. Last month I said a motor that won't repeat each power run closely makes a poor dyno test mule. This RDI unit would deliver curves that laid one over the other time after time. Out of faint hope more than anything I suggested to Kenny we try 32 1/2 degrees total timing. Normally a 1/2 a degree is of no consequence especially as it is difficult to turn a distributor by just half a degree. However with the minimal scatter the Performance Distributors unit was giving and the micro adjustment changing by 1/2 a degree was no problem.

And, yes, I liked the results--501 hp. The magic 500 barrier was now broken. By utilizing the 1/2-degree timing capability the RDI motor returned 500-plus hp pull after pull. At this point we could safely assume all the power possible had been wrung out of the motor via jetting and timing. The WC Holley had delivered extremely well. Even though it is a race carb it idled smoothly and was very responsive. Also the brake specific fuel consumption of this motor with the Holley was exceptionally good.

To make the point a part throttle test to simulate a loaded half-ton truck (5,500 pounds gross) going down the freeway at 70 mph was done. The more the throttle is closed the harder it is to get good BSFC figures. For this test any brake specific fuel consumption figure under 0.55 lbs/hp/hour can be considered very good. Our WC Holley-equipped motor delivered 0.53. At full throttle the Brake Specific Fuel Consumption was consistently in the 0.39 to 0.44 range. In plain English this motor was proving very fuel-efficient. Now it was the turn of the Demon to show its paces.

First the Demon required a different fuel fitting setup to the Holley, so be prepared for that eventuality. Once on the motor the fuel level was checked at the very effective site plug. At this point the idle was set. With the idle screws barely over half a turn out the motor idled at a smooth 650 rpm while pulling a healthy 14-15-inch vacuum. From here on the same diligent calibration procedure was carried out on the Demon carb.

Although we were supplied with air correctors for both the idle and main jet circuits they proved unnecessary. The fuel curve, like on the Holley before, was about as good as it gets. This left us only with the main jets to deal with. Also since we had now amassed a considerable amount of time jetting with the Holley carb we knew pretty well where things should be. In a relatively short time the Demon was delivering its best.

And that best proved very well as you can see from the output comparisons on the graph below. From 2,800 to 4,400 the Demon was up on the WC Holley with differences in torque amounting to as much as 18 lb-ft at 2,900 rpm. From 4,500 to 5,300 rpm the Holley won out by up to about 5 lbs-ft and from there on up it was anyone's game. Peak power with the Holley was 501 while the Demon never made 501, but consistently hit 500.5!

From the output aspect the $523 Demon looks like it is every bit as functional as a $1,400 Winston Cup carb. The Demon also returned exceptionally good BSFC figures with numbers running from 0.37 at 2,800 rpm to 0.42 at 5,800. The part throttle cruise test showed even better fuel economy than the Holley's already excellent figure. On this test the Demon showed 0.5 lbs/hp/hr.

One might suppose that the Holley could be at a disadvantage in this test because it was built to be optimal on an engine operating in the 450 to 780 hp range. Certainly from the point of view of airflow capacity there is little to choose between the Holley and the Demon. And yes the Holley could well be at a disadvantage. For instance booster atomization characteristics would probably have been optimized for best output within 50 hp of peak power where as the Demon almost certainly would favour operation in the lower ranges exactly as was found. The extent such possibilities we could not establish because out test engine only just reached into the Holley's range by 50 hp. However, it says reams for the Holley in it produced excellent results even when operating below what its normal range would be and at the same time shows the Demon to be every bit the top-notch carb that it is.

Final Tuning
Once the 500 barrier was broken, a new goal was set. Could we make 505 hp? An increase in spark intensity might help. With Performance Distributors HEI unit plug gaps up to 0.75-inch can be successfully bridged. The wider the plug gap used the more intense the spark. Our gap was already at 0.050-inch so not a lot was expected here. However some experimentation here proved worthwhile.

By making an already big fat spark a little bigger the motor rewarded us with a new peak of 503.6 hp but try as we might it would not, at least in this round of events, go 504. The final curve is on page 164. Here I have also put the average of a number of regular, non-custom 392 crate motors that Kenny has tested. Note our Custom 392 beats out the regular crate motor by 55 hp and 55 lb-ft of torque. So how come the big difference? In a sentence it is build quality and component compatibility. The build quality is what you get for the extra cash it costs to do a custom crate motor as apposed to a regular crate motor.

As far as parts compatibility is concerned the biggest influence on the end results would be the heads and the cam. In Part 1 you read how the head port cc/velocity was matched to the peak power rpm. Note that this computation was less than 100 rpm off the projected figure. Next compare (page 162, bottom) Ford's crate motor cam spec with the Motor Machine computed cam spec. The cam probably accounted for half the difference between the RDI motor and a stock Ford crate motor. Probably the most interesting point here is the cam in our Street Pro 392 is 8 degrees less duration on both the intake and exhaust than the cam in a Ford crate motor. Even though shorter on duration this cam allows a higher peak power at the same rpm and delivers far greater torque.

All this points to one thing--more accurate cams out perform cams that are simply bigger in every respect period. Many pro engine builders are finding having the cam spec computed often delivers as much as twice the increase that is seen by just guessing the cam spec from experience or buying an off-the-shelf cam just because it has worked in the past. In our case it looks like RDI's Preston Miller added some 30 hp to the final result by simply making a phone call and shelling out the $40 fee it cost to get the cam spec computed. Putting this into prospective 30 hp for $40 represents three times the power increase per dollar spent of even the cheapest nitrous kits.

Total Seal Test
After four days of rigorous dyno testing it was time to check the cylinders for compression pressure and leak down. Remember four cylinders of this engine are equipped with regular rings and four with Total Seal rings. The chart at the bottom shows how each type of ring fared. The leakage measure on each cylinder is shown within the bore and the compression pressure above or below. The cylinders with Total Seal rings in can easily be identified, as they are the ones with a zero leakage rate. Not only did the leak down prove to be far superior but, as you would expect, so did the compression readings. The cylinders with the Total Seal rings delivered, on average, 11 percent higher cranking pressure. With results like this why would you want to use anything else?

Conclusions
Although our RDI motor did very well we did not actually see the best it would do if you were ordering a replica. First in anticipation that the motor would make about 475-480 hp I elected to test on 1 5/8 headers. At the power level achieved we would probably have had better results on 1 3/4 inch headers from about 4,500 rpm up. Also, the fact we had regular rings in four of the eight cylinders meant four cylinders were not being sealed up as well as they might. It would not be unreasonable to suspect our motor could be 2-3 hp down on that account.

Second all our testing was done on the mineral break-in oil. From past experience I know changing from a quality mineral oil to Mobil 1 synthetic is consistently worth 3-5 hp. What this means is, as good as the figures achieved were, a customer for one of these motors could be looking at 510 hp plus. Although this motor would be completely at home in a working truck or a Crown Victoria because of its excellent low-speed manners (what a stealth motor this is) we are going to put it into a 5.0 street/weekend warrior so look out for the subsequent story down the road.

Shown here is a comparison between the cam used in a typical Ford 392 crate motor and RDI's Custom Crate motor. The key points are the superior output of the computed cam came from shorter duration but more appropriate valve opening and closing events. Along with this the computed cam also called for a 1.7/1 rocker on the intake as apposed to the 1.6 used by the factory motors.