in pure stock trim (the blue line is a chip I made which has some nice spark advance--red is stock calibrations)
.
The 3.4L DOHC, especially the 1996-1997 models seem to be a fish
out of water,
in my opinion. Reading about classic hot-rodding techniques, certain
themes become
ingrained as they are repeated over and over: Smaller throttle
bodies, skinny, long runners,
longer, skinnier headers = good low end torque. Good for developing
air velocity at low RPMs.
Not so good at developing the CFMs and velocity needed at high RPMs.
In this excercise we are
talking about naturally aspirated (NA)air induction, not turbos and
superchargers.
And I won't even discuss using factory iron exhaust manifolds, or non-equal length (non-tuned) headers.
So, the converse, is true also: large throttle bodies, fat, short runners,
medium length headers,
even 4 valves per cylinder, better for higher end power--developing
big CFMs at high RPMs
and an accepted sacrifice of good-ol' American low end torque
because the short fat runners cannot develop the squirty, quick air velocity
needed for quick power at very low engine RPMs.
Imagine two anaolgies:
1] the garden hose. Turn it on half way--the faucet being like RPMs
on a motor. In this case, the
RPMs are rather low because the facuet is not turned on full-blast.
. A fairly large volume of water pours out the end, rather haphazardly,
but, there's a decent amount of water. Now, put your thumb over the hole.
Open it slightly... the tighter you make the opening at the end of the
hose, the farther and harder the water squirts out the end. It's the same
volume
of water. It's just how you treat it, that makes a difference.
That same volume of water is forced to move faster, and get to it's destination
quicker. This is analogous to the skinnier, longer runners and headers,
& smaller throttle body. Low RPM torque, without alot of
VOLUME needed. A skinny/ long-runner motor is more like the faucet at low
volume. You can modify it's velocity by squeezing the water, and making
velocity happen without alot of volume AT the facuet (without having to
rev high). Quick torque--without having to rev high.
2] If you turned the water, way way WAY up, you might be able to meet the same velocity of water as before, but at this point, if you tried to force the water to move even faster by shoving it thru your half-opened thumb (IF you were even able to), you'd be choking the water off--restricting the huge volume now trying to get out. What if you left that hole huge, and wide, and open? A whole lot of water would flow thru at a very high rate. The best way to slow the velocity is to turn the water down, at this point. Less water means, less power, more water means, you must move alot of it--with no restriction, which translates to more power. A large runner car is more like a garden hose's faucet on full blast. You have to turn it UP to get velocity--this time (rev it high--and move air without restriction). You turn the faucet down to slow down, and in turn, get less velocity and less power at low speeds. No low end torque, since the water is literally slowly-- rolling out the hose.
So, "fish out of water"? Well. Consider this: The 1996-1997 DOHC has a 2.5" throttle plate. That's huge. Simply huge, for a stock V-6. The runners are fairly long--which hints of mid-range torque, but rather fat--and we all know fat usually means, no air velocity at low RPMs. The ports, intake and exhaust, simply huge. The valve area= bigger than alot of DOHC V-8s.
Have you ever played at trumpet, and then tried to blow into a tuba? If not, forget it... but those who have will know exactly what I am talking about.
So. big TB, fat runners, large ports, but... and this is a big but... long runners--long runners being the trademark of a motor that wants to try to develop low--or at least--midrange torque--certainly not high end. Only 1" shorter than the previous years (from runner end to valve seat). Yet, when dynoed, this motor has peak HP and torque a full 1000 RPM higher, than the previous years. (and is capable of higher peaks, as we'll see) Why the long runners? The motor design seems confused--compared to contemporary hot-rodding rules.
It's a compromise. I feel GM needed to attempt to support the idea that with these supposed Hp improvements, the motor was still a street motor. (all this trouble--a complete re-design, and they gain, maybe 5 more rated HP!!--seems silly) And they did not take advantage of the technical marvels of the Ford Motors, like the V-6 SHO, or the new DOHC V-8 in the Mustangs. They have two lengths of intake runners, and a butterfly which flips between them at a certain RPM, to maintain proper velocity at any torque range. More support for the rule that a large-breathing motor has to be regulated with the RIGHT length runners to develop torque and HP to it's potential. I feel the long runners do help develop some mid-range torque for the street, but actually in turn, hinder the motor's ability to scream and develop HP-- as high as it can rev. It's a nasty compromise. The improvements they made literally did not match up with common hot-rodding sense in the strictest sense of the word.
IMHO. It's dumb. They should have left the DOHC alone. Or: they could have followed the larger port design philosophy properly and developed something like the Honda S2000. No, and I mean NO low end power--screaming top end that begs the car to be driven "like you are stealing it". The 1991-1995 motor is nicer to drive and torquier on the street. Although, in a slightly comic way: my 1996 motor in a Fiero will beat my pals' 1995 motor in his Fiero at the drags, every time. I just need 3/4 of the track to catch up to him and beat him. He has so much low torque, he gets a jump on me. (this is 10 out of 10 races) I catch up because, in any gear, I have more HP than him, but ONLY at a higher RPM. We are both forced to shift at almost the same points. But: I can hang in a gear a bit longer, and actually get something from revving to 7000 rpm, for instance. But my car still feels restricted somehow. Imagine if I could be getting even MORE HP at high RPMs and rev to 7500 and have it mean something?
Well, the 1996-97 DOHC is fixed runner length, and decidedly low-tech in comparison to the Ford motors (and Porsche, and Honda and --oh, the heck with it). But what it lacks in polish, it makes up for in the brute hardware supplied by GM. It just needs some "corrective surgery" if it is to remain a fixed-runner-length motor, and see it's maximum naturally aspirated potential. You should by now, be able to predict what would happen if someone were to match a huge throttle body, with 4 large valves per cylinder, short, fat runners tuned to some medium length, fairly large exhaust headers. Class?? Anyone?? That's right, you'd get no low end power, and almost no LIMIT to the upper end power available, since the motor has all the right orifices to pass a large volume of air--WHEN it's got the high piston velocity to drag it in. At high RPMs, it's potentially capable of dragging in alot MORE air than it's 1991-1995 predecessor--with the earlier model's smaller TB, ports, exhaust porting, everything--and capable of exiting that exhaust a whole lot more efficiently.
The idea, when talking to Don Jeffers, who is an automotive design engineer,
was to experiement, and see how much upper end, N.A. HP I could get from
making short intake runners, and tuned-length headers to match those runner
lengths, and exploit the already great hardware the motor had to offer....
(like... a motor like this 96 needs a head port or valve enlargement--how
ludicrous!!) We both expected the results to be a peaky, higher revving
DOHC. We chose higher PRMs, because we knew it could generate more HP than
you would if you were to run it at lower RPMs as natural aspirated, and
had skinnier, longer runners, and a tuned exhaust for a low-end torque
kind of system. The higher-than-stock revs, and cam-settings would help
devlop power for a properly tuned system--since we were not expecting to
add much torque.
Whether it was suitable for the street was secondary to the task:
If it drove like shit; who cares?
Don Jeffers' rx:
1] 10.5" runners, equal length, plenum opening to valve seat.
2] 37.5" header primaries, equal length, valve seat to collector.
3] big-assed air box plenum (well_I got carried away--it's way bigger
than he called for)
4] cam settings to be determined--most of the tuning to match the system
up is done with cams, since the hardware is defined for upper-end power,
and the cams are the last thing to need to be adjusted to fit this
objective. Don actually has a cam grind I may try someday (more duration
on intake and less on exhaust, basically--certainly no more lift)
5] fuel and spark seem to take care of themselves with the 1993 speed
density system, with some (but not dramatic) room for improvement.
Objective was to get about 220 hp at the wheels at about 6500 RPM--
up from the honest 174hp at the wheels I dynoed below,
in pure stock trim (the blue line is a chip I made which has some nice
spark advance--red is stock calibrations)
.
Once I have a dyno sheet with the numbers using the intake and exhaust
and cam settings, and
probably a bored TB and spring shims, I'll post it. (10/2000)
Another design objective was to give the motor an ideal volume of air
available at the higher RPMs (ideal-- being something so extreme,
no normal car manufacturer would dare do it).
I literally built an air box/plenum as large as would fit under
the hood, and let the runners sit inside this enviroment, free to draw
air from a box about 700ci in volume. About 3x the volume of the 208ci
motor itself. Throttle response above 2000 rpm is suprisingly good, counterintuitive
to the notion that throttle snap might be mushy down low. Up high, throttle
snap becomes positively neck-snapping.
And here's why: (4/2001), I finally went back to the same chassis dyno shop in San Jose, the Mustang Ranch. I had .030" spring shims for the valve springs--allowing me to rev to about 7600 RPM before valve float, but no bored out throttle body. Stock cams are set to about 8* retard for intake and 18* retard on exhaust. 10* of overlap. The intakes felt best at this setting, power really does not feel strong until 5000 RPM! But when tweaking the cams...it locked in... I could just feel it. Anything less felt wimpy. Anything more, and my motor could not rev high enough to complete the peak hp curve!! Right at 5000 RPM there is just this push that is gone if you try to go higher or lower. So I figured the runners were telling me something. This was just barely enough to allow me to shift and rev from 5000 to 7500 in each gear and have a strong/long powerband. Go up to 7500 in first, drop to second, right smack at 5000 RPM. Rev on up to 7500, drop into third, right smack at 5000 again. If you're in a drag race, the only time you suffer the lag is for about 1.5 seconds in first. The rest of the trip is WOT-power at it's best--full-tilt hp straight up--no chaser.
Exhaust cams were tweaked until I felt alot of power from overlapping the intakes. Idle was slightly lopey, but totally fine for daily commuter driving. Overlap is really most effective if you have truly tuned headers. Otherwise, you'll never scavenge much during the overlap period, just dirty up your emissions. Overlap is no substitute for longer duration on the intake cam, which I have not tried (imagine..!!). Also, retarding your exhaust has the nice side-effect of quieting it down, but you have to be careful you don't induce too much pumping loss.
And here's the results, comparing just the old horsepower reading with the new:
If this project was supposed to garner around 220hp, it certainly
was a success!! I acheived 240hp @ around 6400 RPM!! At the Wheels!!
I had added 65hp over stock!! That's about 38%!! For you Crank-hp
folks, this is at least 270hp!!
I am pulling strong to 7500 RPM in 4th gear and the dyno software cuts
me off because 165mph is as high as it goes!! On this run, the wire was
a little loose, but all my 4th gear runs were the same. The guys were getting
nervous about my Z-rated tires holding up to many more runs in that gear.
And here's the old torque reading versus new:
(the shop had the assistant help me that day and he didn't change the name on my runs, so I edited it to be "MODS")
I added a good 10 ft-lbs of torque. Not bad considering I had not added
any displacement to the motor.
Notice how far higher the peak of the torque curve had moved... Almost
to 6300 RPM.
Now, The Tuned Intake Box Giveth, and it Taketh Away. This is
NOT a suprise. Notice the huge torque (25 ft-lbs) dip after 3500 RPM--also
evident in the hp curve. Yes, you can feel it. But this is WOT from 1500
RPM. You can feather it by not going to WOT until about 4500 RPM, and it's
actually smoother, but I did not get a chance to try some dyno runs like
that.
Here's the intake box:
I left the design "open-ended" literally. I can remove the top and view
what's going on
make inspections (clean up oil from the PCV which does not feed directly
into the runners)
and perhaps, someday, even make the runner length adjustable (with
sets of different bolt-on stacks)
to literally "tune" the motor until it sings like a violin, exactly
in the RPM range I want.
the K&N filter is not a conceit...I'd be the last to throw one
on for looks.
They are loud, they attract attention, they look dumb. They hurt my
ears.
If you use science to determine how you should make these
decisions, like
making some runs with Diacom, for instance, and measuring your vacuum
at
WOT, you might be able to prove whether you really do
have any restriction
up at say, 6000 RPM which might be causing you some power loss. With
this intake,
you worry about power losses at 6000 RPM and above, believe me. I was
seeing
30 or 29 Hg dropping down to 28 or 27Hg when this motor was stock.
When I put the
intake on, it got worse, down to 25Hg at WOT and 6000+ RPM. This
was actually
good news!! Why? The intake was doing it's job: sucking
more
air at higher RPMs.
The Fiero air cleaner and water separator had finally shown their weak
side. Fully
adequate for a 2.8L pushrod motor, they restrict the air a motor like
this can pull.
I put a K&N filter on it, and not ONLY did my pants
seat feel better, I could measure
(MEASURE--folks) a good 30-29Hg at WOT at 7000+ RPM.
Same thing happened to Terry Kelly on his SC Fiero. His response after
trying my idea was something
like, "EEEEEEEAAAAAAAAAH!!!", I believe.
THIS is a situation where you NEED to actually do something about restriction.
Measure it first, then do something.
since this picture was taken, I radiused the edges of the TB intake
hole
and also fully polished the runners and the lower intake manifold.
(did it make me fly? no...
of course not.... did I feel better about it all..? sure--I
guess.)
It's sinfully easy to remove and work on DOHC lower intake manifolds,
considering all
they have to worry about is a little coolant (no lifters or rods),
so I could not
say "no" to doing the lower manifolds also.