Topic: Exhaust design question

[Cammando750]

Greetings to all, I am new to your forum! Honestly though, I've been lurking in the background for some amount of time, listening to all the wise advice and helpful hints.

I've got an oddball question that I'd like to ask. . . It's about exhaust design and maximum flow. Correct me if I am wrong but most exhaust systems pair up opposing cylinders (that is, cylinders that fire apart, not next to each other, in the firing order). For instance, in order to increase exhaust flow on a 1-2-4-3 firing order (the firing order on a SOHC 750), a 4 into 2 system needs to link header pipes from cylinders 1 & 4 together, and 2 & 3 together. This would create a vaccum that each exhaust pulse could follow, making for better flow.

However, when I look at aftermarket exhaust systems I find that they pair up 1&2, 3&4. Why is this? Is my information on the firing order wrong? Or are they designing these pipes for form over function? Are there any advantages to pairing adjacent pipes together other than convienence or a unique exhaust note?   

The reason I need to know is that I am building my own custom exhaust system and would like a balance of torque and high rpm power. The 4 into 4 delivers good torque off the bottom, but has a weak top end. The 4 into 1 has killer top end, but seems to have a dip in torque at low revs (lack of back pressure). My solution is to go with a 4 into 2, keeping the same diameter headers and leaving these headers somewhat long before using y-pipe to link them together. This should (i think) give me slightly better flow than the stock 4 into 4, and generate better mid-range. I was also toying with the idea of using an H-pipe or X-pipe to connect the two end pipes together. I haven't decided yet.

Any imput/advice would be greatly appreciated!

[TwoTired]

However, when I look at aftermarket exhaust systems I find that they pair up 1&2, 3&4. Why is this?

Pipe routing neatness.  I have a CB500 with an old crossover exhaust and 4 into 2 you are contemplating.  The oil filter is kinda in the way and the exhaust looks ... bulky up front.  The crossover is under the engine, reducing ground clearance.  Center stand is gone, too.

Is my information on the firing order wrong?

Don't think so.

Or are they designing these pipes for form over function?

Yup.

Are there any advantages to pairing adjacent pipes together other than convienence or a unique exhaust note? 

Routing neatness.  Ground clearance.  Sales popularity.

Then there is the question of how many people want ultimate power from their 30 year old machine.  It's not that active on the racing circuit is it?
I love the SOHC4 design.  But, power wise, nobody on a modern bike is going to notice the extra gains you've made with a trick exhaust.
Not trying to discourage you.  I would certainly admire it.  But, it's hard to imagine a large marketplace for a trick exhaust like this, for this engine, in this time period.  Only if you sell it cheaper than a Mac type will you have any kind of volume marketability.  There are exceptions, but most of these bike's owners are more interested in cheap and looks rather than actual max performance.   Then there is the carb rejetting required which is far beyond the skills of all but the most dedicated home wrenchers.

I could be wrong, though...

Post pics if you do this.  I'd like to see it and the performance figures!

Cheers,

[Cammando750]

Two, many thanks for the useful reply. I've always loved the feel of the older bikes with the skinny tyres and picked up my CB750 relatively cheap. So I have some money to burn. I own a few modern bikes as well, and sportbikes no less. I know I'll never get this old grampa bike to burn through a quarter mile like a current-crop plastic rocket, but I love to tinker and mod and tweek things until they work better than they did before i touched em.

Anybody have suggestions for jetting? I've heard a 120 or 125 main will work good with open pipes. . . I plan on installing some homemade adjustable baffles as well (turnable bolt with a washer welded onto it), so I can adjust for more or less back pressure.

Thanks again yall!

[Einyodeler]

I have a 2-3 & 1-4 crossover on one of my 500`s , there`s just enough room to get the oil filter off and I have the centerstand.
 No idea who made them and the baffles are about gone,but she sure sounds good reving to redline.
Also gives it a whole different sound at idle than the other 500 with the 1-2 & 3-4 pipes.

[Jim Shea]

Interesting subject. I have just been talking o a guy building a custom exhaust for a car and the 'tuned length' is crucial for the gasses to escape properly. Not relevant maybe to the bike question, but equal lengths of headers to where they meet the silencer is important: has anyone actually measured this?
Jim.

[dusterdude]

my 4 into 2`s are made 1-3 and 2-4.

[sohc4]

I have put the little knowledge I have about exhaust systems (designed for power, not for looks - these old irons still rock  ) in a page on my site:

http://www.satanicmechanic.de/exhaust.shtml

A 4-2-1 which connects 1-4, and 2-3 is probably the best system for street use. For racing, (i.e. more top end at the expense of bottom) you may go for a 4-1. All headers should be equal length.

Axl the Satanic Mechanic
www.satanicmechanis.de
www.seeley-regsiter.org
www.easyridertrip.de

[HondaMan]

Oh, boy! Someone to get DEEP TECH with!

You didn't mention (or I missed it) which CB750 you had, so I'll pretend it's a K1 for the sake of discussion, for carb jet number references.

The pipe question: above 5000 RPM, it makes no difference which 2 pipes join in a 4-2 pipe system, because the distance is wrong, anyway. Period. If anyone wishes to see the evidence, look back at Hooker, Action Fours, RC Engineering and the rest: all different combinations of 4-2 were used, and the only thing they really did was improve ground clearance or appearance, one over the other.

You're on the right track, though: the 4-1 pipes improve things for stoplight-to-stoplight racing. The 4-2 versions, with proper jetting, all work so much the same that I'd go for looks or ground clearance before worrying about the power.

If you want the best performance overall, with the excitement of the full-throttle rocket these "big fours" have, here's the full poop:
Use 4 pipes, joining together the two on each side near the ends (Honda wasn't kidding when they did this, honest). Inside the engine, make this change: get a cam with 5 degrees more duration than stock, and advance it 3-4 degrees from the stock hole. Modify the spark advancer so that it can get 4 degrees more spark, but add some tension to the springs for a later advance (I have a post about this somewhere...). Richen the main jets about 10 or 15 to start, unless you're running velocity stacks, then make it 5-10. Drop the countersprocket to 17 teeth, raise the rear to 48 to 52. Make sure you have a good tires and an empty road, 'cause you're gonna need both....

In 1973 we tested several exhaust pipes on a 1972 K2 model, running on a rear-wheel car dyno (dropped the dang bike twice doing it). NONE of the 4-2 pipes increased the on-ground HP over the stock Honda pipes, and these were the baffled stock pipes, not the glass-packs. They DID make more output at lower RPM, though. We jetted and jetted, to no avail. The on-ground HP, stock, was 49 with the Honda pipes (at 7500 RPM) and 49 with Hooker, RC, Action fours and one from JC Whitney that a local liar said gave him 11-second ETs in the quarter mile. They all peaked out at around 6000-6500 RPM. Gearing was: 18-tooth counter and 48-tooth rear.
For stoplight racing, I'd recommend a 17-tooth front and 50 rear with these pipes.

We tried the RC 4-1 pipe and got 50 HP at 5500 RPM, but only 46 at 9000 (2nd gear, all tests).  Same gearing.

Then, we tried the Yoshimura 4-pipe tapered megaphones. They required rejetting from the stock 110 mains to 135 jets, but put down 61 HP at 9500 RPM with no other changes. The 5500 RPM point was over 50 HP, but my notes don't show exactly whether it was 51 or 55, cause I was too excited: just "50+".

In a later test of a stock K4, we got 41 HP at the rear wheel, but did not change pipes. (Didn't drop it, either.  )  This had the same gearing as above, but with about 5K miles on the chain and sprockets.

When I roadraced, I did it at 15k-16k RPM peaks (yes, the 750 will do this). The Yosh pipes just never did seem to quit...but, that's a whole different world of tuning, not for the cheap or weak-willed....

[Lumbee]

...opening up the exhaust def. seems to perk these bikes up.  I have MAC drags on my chop (78 750K motor) and have jetted to 120.  I'm still "feeling" it out so I havn't even got above 7K rpm.  Seems to run crappy below 3K rpm, but really jumps after that.  Have some 125 jets on the way...gonna try that and see how the performace is.  Plugs look OK, but I fear I may be a little lean...

[MRieck]

Oh, boy! Someone to get DEEP TECH with!

When I roadraced, I did it at 15k-16k RPM peaks (yes, the 750 will do this). The Yosh pipes just never did seem to quit...but, that's a whole different world of tuning, not for the cheap or weak-willed....

It is easier turning those revs with 5 mm valves. That is a bit high (15,000-16,000.... cough) for CB rocker arms and stock valves. 

[ofreen]

Oh, boy! Someone to get DEEP TECH with!

When I roadraced, I did it at 15k-16k RPM peaks (yes, the 750 will do this). The Yosh pipes just never did seem to quit...but, that's a whole different world of tuning, not for the cheap or weak-willed....

It is easier turning those revs with 5 mm valves. That is a bit high (15,000-16,000.... cough) for CB rocker arms and stock valves. 

Not to mention piston speeds in the 6400 feet per minute range at 16,000 rpm....

Greg
'75 CB750F

[Sam Green Racing]

Mgp Hondaman, what sort of power was your motor making at 15/16000rpm ?

What size was the motor and was it over square ?

Sam.

[MRieck]

Oh, boy! Someone to get DEEP TECH with!

When I roadraced, I did it at 15k-16k RPM peaks (yes, the 750 will do this). The Yosh pipes just never did seem to quit...but, that's a whole different world of tuning, not for the cheap or weak-willed....

It is easier turning those revs with 5 mm valves. That is a bit high (15,000-16,000.... cough) for CB rocker arms and stock valves. 

Not to mention piston speeds in the 6400 feet per minute range at 16,000 rpm....

Greg
'75 CB750F

  Right. I have spun CB engines up to 10,500 on a dyno and found no power after 10,000 secondary to the poor cylinder fill. Even using Titanium intake valves as large as 36mm and ports that don't resemble anything like a stock CB power started to go away at about 11,000. The cam chain tensioner system is poor also and does not tolerate high revs for long either.

[MetalHead550]

You guys seem to know your #$%*.  As an aside, I was talking to a guy I met at the Christmas gathering we had.  He races midgets(no, not little people  ) And he was saying that the SOHC4 750 was/is a popular application for their cars.  He said this was because they could be cheaply turned into 836s, which checks out, and that happens to be right at the limit for displacement for midget racing.  He said the modded 750 hes running now puts out 160hp at the tires w/ overbore, domed pistons, rods, cam, valves, ignition, and running straight alcohol.  He also quoted some outrageous redline.  I had just read an article posted on this site about the RC builds and so I asked him if they had probs with the crank, or timing chain at those revs.  He said the late 70s engines had better cranks that could handle it.  But damn, 160hp?  Is he streatching it or can it be done?....safely and reliably.

[MRieck]

With enough compression and exotic fuels, boost or nitrous oxide you can see 160HP. Unfortunately you will see a ton of heat and break parts. The crank is pretty stout but the cam chain tensioner system is lousy. A new unit like McGrew at M3 sells can solve the tensioner problem. You also get case shift problems which will kill the bearings- this is one reason old roller crank engines are better at handling large amounts of horsepower. The CB engine wasn't that overdesigned to handle 3 or 4 times the original horsepower. You can get it but at a equally big price.

[HondaMan]

When I helped guys modify the CBs into midget racers, here's what we did (circa 1973-4):
1. Cut off the tranny, made a new plate for the open side to seal it.
2. Drove the oil pump off of the right side of the camshaft, direct drive. We made an adapter for it. Then, it produced about 30% more flow and pressure.
3. Bored to 836cc, 11:1 CR, running alcohol and in demonstartions, nitro with gasoline.
4. The valves (intakes) were changed to 30 degree faces, the intake ports opened into a "pocket" above the valve as deeply as possible (too deep once, had to heliarc it closed and start over), passages left UNPOLISHED for better mixing, but valves polished like a mirror. The cams were real radical, but made by welding up and regrinding the stock ones. Roller rockers were fitted for the lifters, and they were lightened considerably. Overlaps were in the 40 degree range, which made them a real bear to start.
5. Using stock points, but with doubled springs and a hole drilled into the ground-side pad, then slight mods to the opening ramp on the points cam and extending the advance angle to about 48 degrees, we ran 16,000 RPM redlines. (i.e., no electronic ignitions were allowed).
6. Yes, it can be done reliably to a bike engine, if you consider 3000 miles between rerings reliable. But, the tranny mass tries to limit the redline. With a LOT of tranmission work and clutch work, the bike version can reach 16k redline, but if you limit your aspirations to 14k redline you will have much more streetable power. You need a 7 or 8-speed gearbox to cope with the powerband above the 14000 range, because it drops to less than 1500 RPM of useable bandwidth. The midgets used a torque converter and 6-speed gearbox (1/2 midget size) to cope with this.

It's radical. I don't remember having any crank problems with the 1971-73 engines we used (K1-K3 series). Oil was critical, though, because synthetics were almost unheard of back then....

[MRieck]

.
3. Bored to 836cc, 11:1 CR, running alcohol and in demonstartions, nitro with gasoline.
.

  11:1 seems real low for alcohol. I know a well known old timer who built TQ's with Offy engines and they were up to 18:1. Now I know those engines did not use head gaskets but I'd think even bike engines running Oringed copper would need 14:1 or so for good heat and power. It is difficult to get real compression out of the CB engine unless you resort to things like positive .200" of deck height and counterbore the head. I've seen this done for a NA CB drag bike that was very radical....and delicate.

[MetalHead550]

Thanks for the response honda man.  I was a bit sceptical...wow!  Seems like the more I learn about mechanics and engineering, the more I realize I dont know. 

[jtb]

Einyodeler,  you have my dog.     this is the hoppy dog

[HondaMan]

Yep, the 16k RPM thing took a while to figure out.
It ran with (now you'll think I'm lying) 28mm Keihin carbs. They didn't look like the ones in your bike on the insides, but the outsides were stock. Had to be for the class of racing.

Part of the tricks: serious "pocket porting" above the intake and exhaust valves. And, yes, we went too far on 2 heads, reaching air in the process on #2 once and #3 another time. Hemi head. We removed all of the shrouding around the valves. This system directs the flow into a circular path so the compression stroke pushes the richer part of the mix to the walls for better cooling in a stock engine. Limits the flow almost 25% though...Compression: the bores were stock except for 1st oversize pistons (undersquare engine), head milled .010", .020" and .040" in various versions. We heliarc-ed little ridges across the center of the piston crowns, them ground them to clear the valves while reducing volume as much as possible. The quench area was likewise modified down to .007". The intake seats were narrowed and the floor above it narrowed as much as we could. Then, like Ford did on the legendary 427, we added an apparent "restriction" on the floor of the intake ports to cause the air to compress itself in the "pocket" over the intake valve. The result was free turbocharging above 8000 RPM.

Cam chain issues were a bear. We finally used Yosh chains with case-hardened steel rollers on tensioners and idlers - not rubber. The rockers were ground and ground again (yeah, we bent and broke a few) to get lighter, then case-hardened for strength. The valve stems were narrowed in those portions where they did not contact guides. The valves looked like mirrors, all the way down the shafts, but the ports were just mildly hand-sanded for a smooth finish. Brass inserts were used in the guides. Stock valve keepers were used for lack of something lighter. Yosh made lightweight retainers and rocker clearance adjuster screws. The opening ramps on the cams were gentle start, then agressive lift, then gentle beginning-of-close and almost snapped shut afterward. Stock lift or .010" over was used, but we saw little difference between them. We sandblasted the valve springs until they started to weigh less than stock by a few grams.

The carb bodies were bored and slides from 30mm Keihins (specially welded and modified) were installed. We used the standard 4-carb throttle mechanics. Jetting was an obsession....the rubber hoses to the heads were hand-sanded to match passages all the way in. The port inlets on the head were polished and widened only a little.

The real trick was the hand-built spark advancers. They went to full advance at 3000 RPM, then a second device that fit over the (modified) "normal" unit had tapered pins that moved down toward the weights on the "normal" unit. At about 12,500 RPM, these tapered pins engaged and pushed back the advance weights about 4 degrees or so to get past a "wall" we kept hitting right there, due to volumetric inefficiencies. This retard let the heat build up a bit and it would then "slip" into the 13,500 powerband start range. The tapered pins then retracted because of their advance weights and the full advance returned. It was 48 degrees, by the way... Today, this would be a piece of cake with computerized timing, but all we had was the hand-built (and designed, by me) transistorized ignition, following Ford's lead from the 427 again, that was switched by the points. Double-springed Nippon Denso points. Honest! Although, we did ease the opening ramps on the points cam, and welded up a higher ridge in the open portion. (That's something I learned from the CB450, running it to 12000 RPM).

Oil had to be preheated, then during the race it was too hot. So, we got some power steering coolers from some full-sized Fords, they look like a U with radiator fins. Then we cut into the return hose from the engine and piped up to the front, just below the oil filter, where the cooler was mounted on the frame. We made copper head gaskets. And, we cross-drilled 1/8" holes in the fins on the oil filter housing: lots of them, for better cooling.

The pistons: we shortened the skirts about 5mm, then cut the piston pins' retaining grooves for a slightly larger circlip (the "button" versions weren't reality then). We dimpled the higher faces of the piston skirts with a round-ground 1/8" drill bit, .005" to .010" deep (deeper up higher) in a fairly random pattern, using this to also help match the piston weights. The skirts looked like mirrors otherwise.

The rods: Yosh made some real nice ones, but they required crankshaft rebalancing. We lightened the rods wherever possible (less than 1/2 ounce overall, though), then shot-peened them with glass beads. Same thing on the crank weights. Then we trimmed the alternator so it weighed about 1/3 as much, and it only generated about 130 watts, just enough to keep the slightly-oversized battery alive thru a race. The whole assembly then had to be rebalanced, an expensive deal. Oh- the piston pins were Yosh high-strength units, very light. They cost $8 each in 1972! Must have been titanium or something, but were really light and never broke (or wore!).

Big surprise: the stock oil pump was up to the task. We just beveled the intake holes and matched passages all thru the engine to allow free flow.

By 1974, when I was getting out of the semi-pro thing, someone had started making roller-rocker kits. I wish I'd had some of those.

[rucorey]

This thread has more info than I can process at this point 'Great Stuff, though. What kind of power profile would a stock 750F get with a Dunstall style configuration?

[8 Track]

I fell asleep during exhaust engineering class, but I don't see any scavenging advantage there, so don't look for a boost at the top end.  (I hope I'm wrong, because it looks like it would look good on the bike!)

[Lumbee]

...good look'n pipes ru...r those yours???  where'd u get'm from?

[HondaMan]

This thread has more info than I can process at this point 'Great Stuff, though. What kind of power profile would a stock 750F get with a Dunstall style configuration?

Dunstall's always made a big splash, but I've not had opportunity to use it on Hondas. Worked good an a couple of Norton 750 Commandos that we "converted" to Dunstall-like bikes with a bunch of those parts, though (1970). Both of them outran my K1 afterward..   

[rucorey]

hmmm, interesting comment Hondaman. Maybe I'll give them a chance. These are for sale at http://www.cyclexchange.net/Exhaust%20System%20Page.htm

These guys seem to have some knowledge of sohc's and they have some interesting bits...