Very interesting comments.

Batz, I can't comment on the negative pressure wave.  I just don't know enough about the acoustics involved, and I personally am of the opinion that the acoustic wave pretty much only pertains to open exhaust systems.  I can't fully explain why I feel that way, but my gut tells me that if you are dumping this big pile of hot, expanding gas into a can with a bunch of teeny weeny holes, and baffles, and attenuators, the acoustic wave is toast.  You need an abrupt change in diameter and pressure to create the acoustic wave.  All we are getting when the gas enters the muffler is sudden stoppage.  I'm sure that there is a science to all this, and folks with proper credentials have it all figured out, but a shade tree mechanic like me doesn't have the skills or the means to figure it out.  So I have to keep my plans grounded in concepts that are more easily adapted to the shade tree stuff.  For instance, exhaust systems that are significantly less restrictive rather than "tuned".

Regarding your comment about removing material from 3&9 o-clock positions, all I can say is that I've been reading a lot over the last few months.  Pretty much everything I see from the so called experts says you want to increase the radius on the long side to improve the angle of attack, and only smooth the short side radius.  The experts seem to agree that you don't want to remove a lot of material from the short radius because the port volume increases too much and the gas velocity is adversely affected.  What makes this particular discussion interesting is that all the literature I am referring to deals with the roof and floor of the port, not the sides.  But I also must add that the experts rarely discuss the exhaust port, they generally talk to the intake.  The huge differential pressure across the exhaust sort of makes it a non-problem, at least up to a point.

Your comment regarding the converging condition being "bad news" is interesting.  I agree that its a bad design, but I consider it sort of good news since it should be easy to remedy.  Just gotta be prudent with the grinder.


Lancer, I like your question about the header pipe.  Every piece of literature I have read and every formula I have tried indicate that the stock header is grossly undersize.  But on the same topic, I can find formulas that indicate that a 1.75" ID pipe would be ideal, and others that indicate 1.5" ID is the cat's meow.  To complicate the problem, the user of the formula has to come up with the rpm that he or she is shooting for.  That simple question becomes very difficult to answer.  You might think you want max torque somewhere around 4000 rpm only to find out you miss that low end punch.  And then of course that formula stuff is theoretical.  I can tell you this, I installed a MAC header that has a 1.79" ID.  That seemed to me to be way bigger than necessary, so before I bolted the header up, I shoved an 18" section of flexible exhaust pipe down it's throat.  The flex pipe has a 1.5" diameter ID, but its sort of corrugated.  The corrugation feature makes the problem complicated since it adds an unknown that the formulas don't address.  The bike feels great with that set-up, and I've had to drastically increase the jet sizes (usually a good indication).  But honestly, its seat-o-da-pants.  That seat-o-da-pants stuff has bit me in the butt on more than one occasion.  I believe Rutley is running a 1.75" ID header with a 97mm setup, and he seems to like it a lot (Yo, Rutley, I apologize in advance if I got the numbers wrong).  So a trip to the dyno is in order.  But I just can't justify paying for dyno time every time I make a small change.  So the next time I get some dyno data, it will represent a number of changes, and as such, will be impossible to correlate to any one mod.


Regarding smoothing out the sharp edges and opening the outer ring, my opinion is that sounds reasonable and proper.  It seems logical.  I will probably do it unless I find some good reason not to do it.  Smoothing the sharp edges and getting rid of things that upset laminar flow are non-threatening modifications, there is little risk of screwing up the head.  Opening that port exit has some risk but it simply seems too logical.  It makes no sense to neck it down like that.  I will most likely enlarge it to match my final header pipe ID.

Now I have a question for you.  As I recall, you are located out in Henry, Oklahoma.  As far as I know, that's hot-rod country.  You got wide open spaces and plenty of opportunity for the WOT stuff.  I've heard a rumor or two that you have a pretty swift Savage, and of course you have lots of experience with this beast, and what seems to be the inside track on performance parts.  What size header do you run?  Have you tried several size headers, and if so, why did you choose your current header size?

I currently run a header with 1.75”OD & 1.65” ID, all one piece and essentially the same length as the stock header.  I do intend to try a 1.5”ID header at some point though.
Why did I choose this size ?  I had spent a LOT (years) of time shortly after getting my new 96 model Savage in the year 2000 (Fathers Day gift from my son) looking for information on “how to size a header”.  There was lots of discussion but actually finding someone willing to offer guidance seemed to be harder to find.  I ran across a website that sells primarily header parts for those wanting to build automotive headers for whatever purpose they were in to.  “Headers by Ed” was the site.  He has built headers since the late 60’s and his criteria for choosing a particular size & length in the header tubes and collector is solely ET & speed in the 1/4 mi. which is what he built for the most, but’s also included seat of the pants feel when appropriate.
He has a LOT of info on his sight about theory and practical use, one of which is a listing of cylinder size and hp/cylinder being produced by the engine,  and then suggestions of the tube size best to use.  The suggested size for our cylinder is 1.5”-1.65” ideally.  
His rational for what he suggests just made sense to me.  I did not understand everything in the theory but his practical approach was understandable.

LANCER wrote on 11/22/18 at 04:37:36:


Lancer, doesn't your bike have a 97mm piston? If so, 1.5" is going to be a tad on the small side for higher rpm use. That size will put gas velocity near 300fps at 5000 rpm which is a little below where peak HP occurs.. Ideally you would want to be at 290 fps at the peak hp point. What you will gain at lower rpm you will lose at higher rpm with that size. 1.5" or 1.625" ID is a good size for the stock bore. For the 97mm, 1.625" is the smallest you want to use and 1.75" ID is as big as you would want to use.

For the pipe length, that is determined by your cam timing. The formula is (180 + your exhaust valve opening BBDC x 950)/ rpm of your shift rpm=pipe length in inches.

So I got out the inside calipers, a dial caliper, and some copper wire and mapped out the port.  The copper wire works great.  You form it to hug the surface of the port, then lay it on a sketch with several data points mapped out and trace the wire with a pencil.  Here's what I managed to sketch for the exhaust port.

I hope the sketches are useful to all.  I'm already thinking about a few improvements that have potential.  If you have some ideas, mark up a sketch and post it.  Looks to me like the exhaust is even more choked off than I imagined.  Maybe I should reevaluate my comment regarding the big differential pressure.

DragBikeMike wrote on 11/23/18 at 23:11:07:



I have 3 kids & 11 grands, a son and 2 daughters.  I was intimately involved in my kids lives daily from the day they were born and have been blessed with great relationships with all of them.  Daily involvement, consistency in teaching and discipline when needed, and all that wrapped in lots of love will yield wonderful benefits for the rest of my life.  They are 43, 42 and 39 now.  I am blessed.


The engine runs very strong, from just off idle up to redline.  It pulls hard enough to require some effort to hang on, you cannot just sit back as if on a stroll.  I get a bit of a wobble above 80 so I need to address that still but I don't typically go there very often.  I think the pipe is sized well for the engine.

Fast650:  Yes, I have a 97 piston in it, Stage 3 cam and a 36mm round slide with UFO, DAJ & TW.  I also have the Barnett clutch pal though don't think it is really needed unless doing a lot of 0-60 & 1/4 mi. runs.  The stock clutch did well until after 8-10 such runs before it began to slip a bit.


These are the basic charts Ed has that are designed to help with making choices on pipe size.  

but I am also OK that my Savage is not 50HP.  I have always enjoyed the challenge of riding slow bikes fast.


Making riding a Savage a near orgasmic event.

Dave said,

"I often wonder if the flow of the exhaust port can support (or requires) the extra valve lift of the performance cams?  Currently my bike runs really strong in the low to middle rpm ...then the engine begins to feel like it is working hard to make power as it gets up toward 6,500 rpm."

  The fact that the performance cams have extra exhaust valve lift is probably the reason they work so well. The extra lift keeps the valve head farther out of the exhaust stream, and the lift adds to the duration ,giving added time to extract exhaust gases. At 6500 rpm that is about 18.4 milliseconds, and our ugly port needs all the help it can get, but perhaps that isn't enough. It maybe that the intake ,with 36 VM carb  is too effective, or the fact that the domed 97 mm piston which protrudes into the combustion chamber(making it smaller) ,causes a higher percentage of exhaust gases to fuel/air  mix,(there are exhaust gases left behind anytime the motor exceeds peak torque 3500 to 4000rpm ) It may pay to try decreasing intake valve lift or to try wave tuning the exhaust,  shortening the Exhaust header from the stock length of 32 inches ,toward the limit of 30 inches may increase scavenging rate for the motor at those higher RPM's.(the wave reverses at the exhaust's first enlargement , the wave travels at the speed of sound-mach 1 - exhaust gases travel down the pipe at about 350 feet/per second) . removing the baffle in the muffler may have little effect.     All this is of course OPINION .