Put an O2 in each primary and you will see that “balancing” the flow by sucking on individual runners with the others blocked off isn’t the whole picture.  Scott has a point that I have to agree on.  Looking at Mr. Joe’s flow data on a stock and a modified RPM manifold, notice the #6 and #7 runners in both cases flow more than the #1 and #4.  If you tune up #1 and #4 to mid 12 AFRs at wide open throttle, you will find out that #6 and #7 are usually lean enough to scare you.  Mid 11s on #1 and #4 will generally get #6 and #7 to a tolerable range.  The lower side of an RPM has very  good distribution, but the high side isn’t as good.  Usually they will make more power at WOT if 1 and 4 are a little too fat, which still leaves 6 and 7 a little too lean.  Part throttle cruise, lightly loaded, will show nice distribution to all eight cylinders.  Fooling around with various spacers can help the high side distribution, or screw it up worse than polio, depending on the path taken.

In terms of distribution, the early 60s iron manifolds, Sidewinder, 428 CJ, and Ed RPM, in that order, have the most consistent AFRs measured in the individual primaries.  The Ford PI and Blue Thunder are the WORST for distribution.  Both of those will cook #6 and 7. You can’t fix it with jets, because 1,4,6, and 7 use the same piece of the carburetor.  I discovered the problem working on Stock Eliminator FEs that are bound by rules to use specific manifolds.  After that, I started looking at all of the dual planes.  You can fix the 2x4 intakes with staggered jetting.  The other neat thing when you start smoothing AFRs is that BSFCs can smooth out, and there is also a by-product of more power and torque.  The closer it gets to “right”,  it also starts trending towards helping the “terminal” runner design problems.  I think pooling of excess fuel in the manifold improves as the overall tune gets better.  You can wear one engine with one manifold out trying to correct distribution issues.  Making some windows in the other plane into 6 and 7, and using different cam lobes on 1/4 versus 6/7 have shown trends in the right direction.

There is more going on in there than porting and dry flow testing will show.  I can say for sure that most manifolds have design characteristics that can’t be fixed without cutting them in two and changing their mind in places you can’t reach too good otherwise.

I will say that it is a plus to have runners that “flow” more and more equally, but the distribution is a different situation altogether.  If you look at AFR in collectors, it looks pretty good down there, where two high runners and two low runners mix together to create the number.  An old C4 iron low profile manifold isn’t the best for power, but it is the best for distribution. 

My two cents

I think I have a set or two of 4.035 bore stroker pistons on the shelf, and also some forged 4.062 bore CP pistons for the 3.50 stroke.

Pretty easy to see……..Van Cleve’s has the most power, and Paquet’s has the most power per inch.  Both of those engines have conventional FE cast cylinder heads.  At 440-ish inches, and Super Stock legal, with 715 carbs, Ray’s car is awesome, and unmatched in terms of power per inch.

We didn’t go overboard on camshaft in Jim’s engine.  It is 501 inches, but intended to be low maintenance.  A Super Stock camshaft would make his car even faster, but would require very frequent valvetrain maintenance. 

The non traditional heads should make more power than the inline, stock configuration, 13 degree valve angle heads.  I’ve done one engine with four digit power.  Numerous with 900+ power.  About 15 years ago, I did a Super Stock 427 that went 8.60s, which was fast at that time.  That “legal” engine made 892 hp at the time.  Ray goes 8.40s now, so it’s safe to say his probably makes 940-ish power these days. 

We are getting ready to run a 540 inch FE with billet heads.  I am sure it will make really big power, but not apples to apples with conventional design heads. 

The Cammer is king on OE design FE heads, in terms of flow, but the rocker ratio hurts the potential of the engine.  There are really three categories of FEs now.  Conventional wedge, SOHC, and modern redesigned stuff.  The modern redesigns should dominate, but so far, the quickest FEs on earth in 2025 have conventional wedge heads.

those shorty's didn't lose too much to full headers, 18 horse for the cool factor is worth it on a 62-61 Galaxy .

Blair how much more power would a single 4 make over the 3x2 in this application ?

Eric,  I don’t think a single carb would be as good in this combo.  The 3x2 manifold is sized pretty well for an engine this size, and the injectors in those three throttle bodies are tuned individually.  We checked AFRs in all eight cylinders, and we were able to get distribution pretty darn nice in this set up.

It only has 222 duration at .050.  It’s designed to run DAILY.  If it was a toy, I would like to see what would happen if we threw some camshaft at it……

This little engine was so neat I had to post it.  It is a daily driver, 352 + .062 overbore.  Low lift, short duration, 16” idle vacuum.  3x2 EFI on top of the new BBM heads, cnc 2.100 intake ports. Static CR is 9.2:1 and built for 91 octane California gas. 

We made 438 hp/456 torque with dyno headers.  We swapped on a set of Kugel Komponents cast shorties at the end, as that is what will be used in the ‘65 pickup it is going in.  It didn’t hurt it much, and would likely be even closer, but they had a hard bend in the 2.5” pipes between the headers and my exhaust tubes.  Still had 415 hp/ 437 torque!!   

Daniel put a short dyno vid on YouTube @captcobrajet.




Performance Summary:
      Cubic Inches:    362          Dyno brand: Stuska
      Power Adder:     None           Where dynoed: Blair Patrick Enterprises
      Peak Horsepower:  438 at 6120 rpm
      Peak Torque:   456 at 3800 rpm

Horsepower and Torque Curves:


Engine Specifications:
   ‘64 390  3-web block   4.062 bore
     
   OEM 352 iron crank 3.50 stroke
     
   Molnar rods 6.540 FE big end

   CP full skirt heavy duty custom pistons.  1-1-3 mm

   King main and rod bearings

   Mahle. 1-1-3

   M57B with BPE mods

   Aviaid Cobra pan modified for ‘65 F100 chassis

   Camshaft: hydraulic roller .520 lift, 222 @ .050, 110 lobe separation

   Lifters brand, type: Morel short travel

   Timing chain and timing cover: Cloyes double roller

   Cylinder heads:  BBM  C8OE-BP castings  2.100 cnc intake, 1.665 as cast exhaust

   Cylinder head flow in cfm at inches of lift (28" H2O pressure drop):
      Intake               Exhaust
      .100    105      .100     70
      .200    170      .200   135
      .300    225      .300   180
      .400    280      .400   210
      .500    305      .500   225
      .600    320      .600   235
      .700    330      .700   238
      .800    335      .800   238

   Flow bench used, location:  Quadrant Scientific,   BPE

   Intake valve:   2.100 REV 11/32

   Exhaust valve: 1.665 REV 11/32
   
   Valve springs brand, part number, specs: PAC beehive

   Retainers and locks brand, part number, specs: 10*

   Rocker arm:  Rocker Arms Unlimited adjustable

   Rocker shafts and stands, brand, material:  HD  Rocker Arms Unlimited

   Pushrods brand, type, length:  Manton  3/8

   Valve covers, brand, type:  CJ plain fin reproduction

   Distributor brand, advance curve information:  Holley/MSD dual synch

   Harmonic balancer brand:  Powerbond

   Water pump brand, type (mechanical or electric):

   Intake manifold brand, material, porting information: 1961 OEM Ford 3x2  BPE light mods

   Autotrend EFI 3x2 throttle body injection with Holley HP ecu

   Dyno headers and Kugel Komponents cast shorties

All of these examples posted are all different combos.  Just looking at where peak torque happens on each one, you can conclude that Mike has the smallest cam, and Mr Woody has the biggest one. Curious how many of these engines had a dual plane manifold……..that also can affect it.  If the cam isn’t big, and the engine is, it will run out of plenum past peak and taper off.  The strength of the dual plane is it will have nicer power where you want to drive a street car.  Happy and peppy below peak torque. 

I'd say Mike's data is pretty accurate.  The power curves correspond pretty well to the flow he witnessed.  The iron EMC heads I sell have nice  11/32 stem valves.  The contest heads did have a rather expensive set of 5/16 stem valves with some back angles that I don't sell, just in case I ever go racing again, lol.   The valves size is 2.150 in the iron.  Bigger than that gets pretty thin around the spark plug hole.  It would respond to a 2.200 valve, but it wouldn't live long before it would leak beside the plug hole.

That back-up in the TFS head is a real thing.  I have noticed over 20 years of flowing on my bench, that it is more finicky about turbulence than a Superflow.  Mine seems to expose the turbulence and it affects the flow more than the Superflow benches.   I am not smart enough to know why, but I know it happens.  The crutch for the turbulence is most likely wider lobe separation, but it will take torque away through most of the curve, as it helps it hang on up top.

Mike's more than likely had a tighter lobe sep than Brent's examples, and the torque was there bigger and sooner, and then it ran out of breath.  Overlap will aggravate the sonic problem in the TFS head.  If the exhaust isn't pulling it through as hard, the turbulence is less pronounced.  The downside is the loss of usable torque by going wider.  Torque is king unless you plan to drive one around at 6000 rpm, so chasing a peak power number that loses you torque at lower revs isn't how I would do it.   This has turned into an interesting thread.  It brings to light the fact that different heads have different personalities, beyond just flow numbers, and they all won't like the same cams.