I  love it ! Everybody remember his words , lmmfao. We all said that. But you build it , cause we all love a good build. What type of racing ? On or off road ? Build it and just maybe I’ll see you at Four Corners on a crowded Sunday morning. Do they still do the breakfast run ?

Ok , the names haves changed , The Sunday morning ride was originally called The Breakfast Run , big fun cause sometimes you would ride with rich or famous people but they were just another rider to greet. And Alice’s is four corners and on Sunday morning that area was like the Concourse De Elegance, some of most exotic bikes in the area all lined up and there too you never knew who you might meet. My X stomping grounds learned how to ride in them woods , from Hwy 92 to Santa Cruz. Gas up in Santa Cruz race to four Corners. But yeah I bet riding there ain’t much fun anymore due to recent squid population explosion. Hell I’d go just to see bikes & hang out , maybe see an old friend or familiar face.

After reading DragBikeMike's technical articles I decided I should pick up a spare Savage head to inspect.

Peering into ports (the cylinder head people I've known usually don't look, instead they insert a finger and feel around in the port) can be a difficult way to tell the shape of them.  What is helpful is to pull a mold from the port.  2 part rubber compounds are often used.  What I have is a product called "Vinamold" which is vinyl based and can be melted and reused many times.  The white compound is of medium stiffness.  It also works well without additional powdering when doing a 3D scan.

Last week I ran the spare head through the ultrasonic cleaner many times to get the ports as clean as I could, heated up the Vinamold and poured it in.  The intake mold wasn't too difficult to pull out, but the exhaust port took a lot of work as I didn't want to pull too hard and tear it.

I also have some Aprilia/Rotax RSV 1L twin port molds to compare the Suzuki molds to.  The Aprilia is a 500cc cylinder making 61.5hp @9500 (per cylinder), and 37.5 hp @ 6500 RPM.  The Suzuki is reported to max out at 31hp @ 5400 RPM.  The Aprilia is 123 hp/liter, the Suzuki is 47.7 hp/liter.  Power will vary with RPM, but the 50% increase in RPM of the Aprilia gives a 2.6X increase in power over the Savage, so it is breathing much better.

Valve sizes (Savage from DBM) and horsepower/RPM:
Suzuki
94x94 (single)
IN 33 x 98 x 6.96mm
EX 28 x 91.5 x 6.95mm
31hp @ 5400 RPM for 47.7 bhp/liter

Aprilia
97x67.5 (twin)
IN 35.89 x 90.33 x 5.97mm
EX 30.95 x 90.56 x 5.94mm
123hp @ 9500 for 123 bhp/liter
from a dyno chart, 57hp @ 5400 RPM (Savage power peak)
75hp @ 6500 RPM (Savage reasonable redline)
But those are best cut by 50% to compare a 500cc cylinder to the Savage's 650cc.  The Savage is tuned to make more power down low compared to the Aprilia.

I was surprised to find the Savage head had the longer valve stems.  I had expected them to be short and contributing to the port needing to bend.

First the intake ports (the molds with a valve in them are the Suzuki ports):




The Savage doesn't benefit from the sidedraft intake, putting a bend in the port, but the shape doesn't look bad.

Now the fun stuff, the exhaust ports:







I showed photos of the Suzuki exhaust to a knowledgeable friend who remarked "It is amazing just how bad manufacturers can make things.  It looks like the performance chaps had a reasonably good port design, very steep upwards but then the valve chest designer came along and put his dents in it.  I sometimes saw this back when I was gainfully employed, two designers at odds with one another or that they just never got together for a chat."  Another friend remarked that "good intakes/bad exhaust" has been a combo he's seen on various engines.

Unfortunately, a 7mm thick layer of aluminum can't be easily added in the floor of the trough above the exhaust port because the valve springs are in the way.  A quick measurement with "finger calipers" doesn't seem to show the floor of the trough being very thick, though there's probably enough metal to shave a bit off the bottom but a person would want to be pretty cautious when doing that.

At this time in the project I'm not going to worry about it.  DBM and others have probably found all the reasonably easy gains.  I'll still plan on  digging out my Audie FlowQuick test equipment and test this spare head, but I'm not going to be in a big hurry on that.

The Aprilia ports show that much nicer ones can be had, but it would probably take a redesign of the Savage cylinder head to get them.  I'll put my time into the chassis first.

Gung ho!
I pushed a dead cylinder head thru the bandsaw.
Really shows how bad the exhaust port sucks.

What with the recent (not uncommon) clutch problems experienced by several other people (exploding throw-out cam and worn shaft) I decided to look at the parts in a spare cover I bought.  Yes, there's wear on that shaft too.



The edge of the flat where it meets the outside diameter of the shaft is smeared, there's a visible ridge of erupted metal where the left side of the cam was pressing, and at the right side of the cam where it is up against the shoulder from the flat to the top there was enough of a burr that the shaft wouldn't come out of the case without lightly going over it with a diamond EZ-Lap file.

I decided that while I had the parts in hand I'd start modeling them to capture dimensions (like the difference in the hole in the cam compared to the matching area on the shaft).  It may be difficult to see but the outline of the part is in red.



I've got a good match between my measurements and a scan of the cam.  Note that this doesn't show a few small radiused areas where instead there's a point inserted to show what I presume was the original "design intent".  This made it easier to get dimensions off the drawing to doublecheck against the part.

As DirtBikeMike has pointed out, there's a lot of lost motion in the action of the cam due to the loose fit between the cam and shaft.  He's shimmed his cam, but I'm thinking I might try a couple of different approaches.

One is to set up the inside of the cam like a Snap-On "flank drive" socket.  You can find the patent document for that here:

https://patentimages.storage.googleapis.com/ca/ea/14/4941e2f44fd72e/US9718170...

The idea is to move the contact of the cam away from the corner on the shaft.  Even if the fit is loose, that should help to prevent smearing that corner off the flat on the shaft, and it seems like something that might be done with some filing/stoning on a stock cam.  Ideally, a new cam would have tighter clearances to the shaft and also the flank drive.

I've also ordered some M10 linear shafting.  I've used linear shafting for making spacers to ride inside of swing arm bushings and also rocker arm spindles on some Honda vintage racers.  They are case hardened but once you get under the case they machine OK.  The case might help keep that corner of the flat from smearing (if the flank drive doesn't fix that) but it seemed it would also be helpful in preventing the wear I found on the curved outer diameter of the shaft.

Now a new shaft will need to have a lever attached on it in the right spot/angle, but I'm not worrying about that at this time.    I want to see if I can successfully machine the flats into the linear shaft.  Milling a 10mm thick piece of steel with tiny end mills may be an issue for making a new cam, but using the flank drive concept helps with that.  There still might be a small bit of hand filing/rotary tool work to clean up some leftover machining cusps, but that might not be a big deal.  I think the AR500 used by TheSneeze in his cams may be overkill (but there's nothing wrong with that) so I'd probably use some 4140 or 4340 that I've got on hand, and that should be plenty strong enough to avoid the breakage seen with the sintered stock part.

The shafting was pretty reasonably priced, so it might not get used if I can make a tighter tolerance flank-drive cam and have that make life easier on the stock shaft which I suspect is not a particularly hard/strong steel (but I no longer own a Rockwell tester to check and see what the hardness is).  Using a stock shaft lets me avoid the extra machining/lever  making issues.

If I make something for the throwout I'll use this, a Tree 425 (TOS FV30) with a Centroid control retrofit and 4th axis.  The mill has a lot of Z travel in the knee but it is "old school" and only has 6" in the quill, though it has the "high speed" pulleys on the spindle and will do 6000 RPM instead of 4000 RPM:



I think that drilling at the corner of the flat and OD in the cam with enough offset to cut the flat back on the cam and then putting a bit of radius to create the flank drive might be good enough.  I need to read the patent document to see if there's something magic at work.  But it will be small OD cutters with a lot of stickout so having CNC to drive them at the correct speed is a big help.



I put the cam and shaft in the case to check for lost motion between them and there is a lot.  I'm not sure what the short "leg" on the cam is supposed to do as by the time it hits the case the lever has moved much farther than the cable will move it.  Perhaps making that section thicker would allow it to be drilled/tapped/split for a small pinch bolt so there is zero clearance between the cam and shaft?  The pocket for the pushrod and the flat surface behind it that bears on the hard stop in the case seem to be the main critical features.  I don't know how much clearance there is between the body of the cam and whirling parts of the clutch.  

Here's a photo of the cam with some clay on it with the lever rotated until the short leg hits the case.  That would give room to thicken the section there on the cam, but they seem to break in the thinnest sections 180* apart from each other.  Stronger and more ductile material is in order.

One of the things I find interesting about projects is trying to understand why the manufacturer/designer did things the way they did with parts I'm thinking of changing/replacing.  Were they following a standard practice, did they have a bright idea (that did or didn't pan out), were they aiming to improve performance or save production costs, did they just want to do it their way and be different, or ?????

I'm not understanding the thinking at Suzuki on the Savage (and before that the SP600) clutch.  The regular motorcycle clutch with the pressure plate to the outside of the basket that is pushed or pulled outwards to relieve spring pressure has been a pretty standard item for the better part of a century or longer.  Why did Suzuki decide to reinvent the wheel (or clutch) with their "backwards" clutch?  

I first thought that they might not have wanted to have the release mechanism in front of the transmission output pulley (pushrod through the mainshaft) because of the larger size of the belt pulley compared to the usual sprocket, but the earlier SP600 was chain drive so that doesn't seem a likely reason.  My 1971 TM400R motocross bike had the throw-out mechanism in the clutch cover but it had a standard clutch and the mechanism pulled (via rack and pinion) the pressure plate outwards.  Suzuki used that kind of arrangement on a variety of bikes (including the DR650) for decades.  It is a proven part of their design knowledge.  I'm having trouble seeing a big savings in money with the Savage release mechanism since it requires a different clutch design, and in any event they could have grabbed some standard rack and pinion mechanism out of one of their other bikes that was already proven and in production.

Maybe they bought some new sintering process equipment and were looking for parts to justify that purchase so someone designed the exploding clutch cam to make use of it?  Beats me.

I like to modify things, but I prefer to do that where I think I'm improving a decent part instead of having to try and improve problematic parts up to the level of "decent".