SWCDuke Senior Member 4482 posts [100%] Redondo Beach, CA This may not be for everyone, but if you are interested in the low level details of cam design, you might find this comparison of the design features of the Cosworth BD-3 cam with the SB LT-1 cam interesting The BD-3 cam is an early design for the DFV F1 engine and has been cut on Cosworth Vega blanks. The lobe is very similar to the production CV cam though about 10 degrees longer in duration. The narrative is from an e-mail I sent to some members of the CV community who are collecting data on the CV cams in an attempt to optimize valve timing. As you will see, the difference in detail design is so great that any comparison of "high level" design parameters such as duration and overlap between a pushrod cam and a DOHC cam are essentially meaningless. The narrative also points out some of the features that make the LT-1 cam such a well thought out design - the best design ever conceived for a street high perf. SB IMO where the design objective is maximum torque bandwidth with an absolutely bulletproof valvetrain using off the shelf production components. .................... I have some cam pro curves of the LT-1 SB cam, and they make an interesting comparison with the BD-3 cam data taken by Carl. The LT-1 data is a curve of displacement and acceleration. Unfortunately it doesn't include velocity, but acceleration is the best indicator of a lobe's "aggressiveness". Data is in inches/deg. squared times 10**-5 and is estimated from the data scatter. (The first number "32" would be .00032 inches per degree per degree and so on for the other data.) ........................................................LT-1 cam.......................BD-3 cam Max accel. on opening.........................32.................................70 Accel over the top................................-17................................-20 Max accel on closing............................25.................................68 Note that the LT-1 cam has much lower max acceleration on opening and closing. This is where the pushrod is under compression, and it can't take anywhere near the acceleration of the direct acting CV valve train. Acceleration over the top is not that much different as this is where valve motion is controlled by the spring. Also note that the LT-1 acceleration on closing is milder than the opening acceleration. This is to unload the pushrod compression force as mildly as possible and is why the LT-1 lobe has noticeable asymmetry. If the pushrod is unloaded too quickly on seating the valve can bounce. Such "soft action" of a pushrod cam is necessary due to the flexibility of the pushrod, so a DOHC lobe with its more aggressive acceleration can have more area under the lobe for a given duration, and long duration is not necessary. Also, this is why I believe the most aggressive action - roller lifters - should be chosen when simulating on DD2000. My cam files for the OEM CV cams are now 258 degrees duration with a roller lifter. Another noteworthy observation is that the BD-3 lobe reaches max velocity at about .050" above the base circle on opening and doesn't slow from maximum negative velocity until about the same point on closing. The LT-1 peak velocity occurs at about .150" off the base circle due to the modest acceleration. These major differences in acceleration make "overlap" comparisons meaningless. Due to the quick acceleration of the CV lobe, EFFECTIVE overlap in square-inch degrees would be much greater at the same overlap in degrees compared to a pushrod engine cam - another reason to use a conservative overlap figure (degrees) on CV street engines. Pushrod flexibility is a limiting factor on a pushrod design. This and poorer head flow means that a pushrod engine must have much more duration to achieve anywhere near the specific power output of a DOHC engine, but the longer duration kills low end torque because of the excess overlap that results from long duration cams. Bottom line is that a comparison of cam specs for a pushrod engine and DOHC engine is apples and oranges. I couldn't make any sense out of the CV production cam data for velocity and acceleration because the scales are offset and don't show a proper origin (zero). Mark - maybe you can get these corrected. BTW, Carl's BD-3 data has some incorrect signs on the acceleration values. Duke [Modified by SWCDuke, 1:10 PM 8/27/2003] Tyler Townsley Senior Member 523 posts [100%] Nichols Florida Re: How to design a cam (SWCDuke) 9:04 PM 8/27/2003 Would also like to add some more food for thought taken from a post on another list. The M/S refers to Mega Squirt a DIY EFI setup found here. http://www.bgsoflex.com/megasquirt.htm This M/S thread is something that I know something about - especially since I have the technology to look at the valve lift curves in a way that 99.9% of the motorsport enthusiasts CAN'T look at them. "Once you are at the point where you have to use solid rather than hydraulic followers then plenums can start to cause issues....." This statement is inappropriate in the impression it gives and here is why. Effectively, the only difference between a hydraulic and mechanical cam is the method by which lash is assimilated before the valve moves. Realistically, once the valves start to move, the plenum has no conception whatsover as to whether the lifters are solid or hydraulic. The issue with respect to how "big" a cam is does relate to idle quality and performance and there are a lot of preconceived misconceptions out there relating to hydraulic and mechanical cams - here is the straight story. Until/unless one references the duration of ANY cam to another at a common NET VALVE lift point, cam timing comparisons are totally meaningless. This is why it helps to look at the "cam running duration" defined in SAE J604d and also to cam lift area (degree-inches or degree-mm) at the same reference lift points. Trouble is, you CAN'T do so because most cam companies won't publish specs in a fashion that facilitates this standardized process. However, after looking at every cam sold by TRW (and many other private label parts while I worked at TRW for ~10 years) plus literally HUNDREDS of cams sold by other folks (Crane, Comp, Isky, GM, Ford, DCX, etc) since then, I can assure you that the following relationship pretty much holds true. The NET RUNNING duration of a mechanical cam with the SAME 0.050" CAM duration ends up roughly 10 degrees SMALLER than a hydraulic cam of the same 0.050" cam timing once you take up the lash and get above the clearance ramps. This is what J604D does. Thus, a 230 @ 0.050 deg hydraulic cam will idle worse than a 230 @ 0.050 mechanical cam because the damn thing is about 10 degrees bigger at the point where the valves are moving in a reasonable fashion. Thus, there should be no surprise that a 230H would have more idle/low speed issues than a 230M. If you then compare the 230 hyd to a 240 mech, they act pretty much the same. This shouldn't be surprising because as far as valve motion goes, they pretty much are the same as far as the engine "sees it". The things that guys forget about carbs and big cams is that air reversion pushes some of the air/fuel mix back out the carb venturi. Keep in mind that a venturi mixes fuel as a function of "fluid flow" in either direction and the amount mixed is a function of fluid density. Thus, air gets drawn thru carb and fuel is mixed. As an a/f mix (which is denser due to the fuel already mixed) gets shoved back thru the carb - more fuel is mixed, again. Depending on the plenum/air box, the fluid could get sucked back thru again, mixing even more fuel and so on and so forth if the "stand off" mixture gets recycled over and over. A plenum provides a damper or collection chamber that can contain a "reversed/stand-off" a/f mix from getting shoved back and forth thru carb venturi. With injection, you can shove the a/f back and forth till you're blue in face cause you only get the 1 shot of fuel and no more unless you rotate 720 degress. But, by then, you already ingested the air so you have fresh, dry air to mix fuel with next time around. No wonder M/S's idle better than carbs - they only mix fuel once. Few people can really see what's going on with cams and the cam companies WANT/NEED to keep it that way. However, if you can get access to see what's really going on like I can, you'll see that it all ends up being a lot of bunk and a ruse to sell unwitting customers something they really may not need. This is a prime reason why I TOTALLY steer clear of the retail/DIY cam business. It is just too hard, too time consuming and too unprofitable to provide good, sound engineering logic on how to REALLY use the right cam to folks who've learned a bit too much bogus car magazine technolgoy. Simply put, you can't unring a bell when it comes to straightening out the preconceived misconceptions about mechanical and hydraulic cams. SWCDuke Senior Member 4482 posts [100%] Redondo Beach, CA Re: How to design a cam (Tyler Townsley) 10:59 PM 8/27/2003 I might add that SAE J604d calls for measuring duration at .006" VALVE lift as there is little effective flow at less lift. Some aftermarket grinders use SAE J604d to specify "advertised" duration, and this is the value that will usually give most accurate/realistic results as "seat to seat" duration on DD2000. BUT, the other wild card is rocker ratio. On a SB it starts out at about 1.37:1 and peaks at 1.44:1. It is never 1.5:1! This is why OEM lash settings on mechanical lifter cams must be "factored" to compensate for actual rocker ratio. Even GM did not take rocker ratio into account, so if you lash to GM spec your valves are too loose. Also, even if you use compensated lash the actual duration at .006" valve lift will be a little less than advertised duration. The point about comparing duration at .050" between mechanical lifter and hydraulic lifter cams is well taken. Because mechanical lifter cams have much higher ramps, much of this "lift" is clearance ramp, so you can't compare the duration at .050" of a mechanical lifter with a hydraulic lifter cam, and your 10 less degrees EFFECTIVE duration for a mechanical lifter relative to the raw spec is probably a good rule of thumb, but very few understand this issue. Also, hydraulic cams typically have less aggressive accleration than mechical lifter cams, so comparisons based on a few simple numbers can lead to erroneous conclusions. The real important part of a cam is the first quarter and the last quarter of the lift curve. This is where acceleration is a big determining factor, and it is also where all the overlap is, so comparing "degrees of overlap" is meaningless unless you KNOW that the cams have similar acceleration profiles. The LT-1 cam has 242/254 duration at .050" tappet lift versus 222/222 for the L-79 (151). The LT-1 sounds bigger, right? Maybe too big in fact, but consider that the top of the ramps are .016"/.020" above the base circle. If you compared actual VALVE lift at .050" you'd find the inlet lobes very similar. The exhaust lobe is a good deal longer on the LT-1 cam. It is also phased earlier, which is one of the secrets of the LT-1 cam. The early phased long duration exhaust event helps compensate for the SBs relatively restricted exhaust port. Actual EFFECTIVE overlap for the two cams in sq.-in.-degrees is very similar with a slight edge to the LT-1. It's too bad the aftermarket cam and hot rod publication industries don't do more to educate users, but I think both are more interested in hyping simple (but meaningless) numbers rather than educating their customers and readers. I've never seen a hot rod mag. or book that ever discussed actual rocker ratio in any meaningful way, but all you need is two dial indicators and a degree wheel to measure it youself. With all their talk of "degreeing in the cam" you'd think that some discussion on rocker ratio would also be appropriate, but I guess they've never figured it out in the nearly 50 years that the SB had been around. Go figure! To accurately characterize a lobe you need to specify the .006" valve lift timing points (preferably rocker ratio compensated) and peak acceleration and velocity on both the opening and closing flanks. Only then can you reasonably make even a subjective comparison. Two cams with the same basic timing numbers can have significantly different torque bandwidth and idle quality due to differences in acceleration and velcoity profiles. BTW, that Cosworth BD-3 cam at 268 degrees (.006" valve lift) duration appears pretty mild, but because of the aggressive acceleration the duration at .050" lifter lift is 240 and 236 at .050" actual valve lift! (The ramps are only .010" high.) The OEM CV cam is 258 at .006" valve lift, 228 at .050" tappet lift and 224 at .050" valve lift. Ramp height is .016". A 258 degree cam for a pushrod engine would only be about 205 at .050". Duke [Modified by SWCDuke, 8:14 PM 8/27/2003]