Yamaha Triples – Technical Updates

Yamaha Triple Bore Distortion – During the 1996 season, we at Group K, have bored (and big bored) many Yamaha 1100 cylinders. Very early on, we found that many low hours cylinders seemed to be noticeably out of round…particularly the rear cylinders. After many hours of examining the situation we found one common theme…the bores were being seriously distorted (out of round) when the exhaust manifold was torqued on. We immediately assumed that either the surface on the manifold or the cylinder was not flat. Lapping both surfaces showed no “bowing” or irregularities of any kind. When the freshly lapped parts were once again torqued together, the bores were once again distorted. After repeating this process with many different cylinders and manifolds, we found another consistent theme. The two front cylinders were always distorted about .0015″ out of round, and the rear cylinder was always distorted .003″ – .004″ out of round. We don’t know why this happens, we just know that it does. It bears noting that .0015″ of distortion can be eliminated during the honing process of doing a bore job…but .003″ – .004″ cannot.

As a result, Group K does all the boring and honing of 1100 Yamaha cylinders with an exhaust manifold torqued on. This process makes for perfectly round “installed” bores. However the bores do distort again when the manifold is removed. This causes a strange situation when we ship a bored 1100 to a customer with bore measurement equipment. We will invariably get a call back saying, “You guys bored my cylinders out of round?”…well, not exactly.

High Speed Smooth Water Wave Venture Handling – The stock (52 mph) 1100 Venture handles very safely and predictably in both rough water and glass water situations. Even the (57 mph) Sleeper kit equipped Venture is manageable under all conditions. However, when horsepower is increased to levels that permit 60+ mph speeds, the smooth water handling of the Venture changes significantly.

In our testing with the 1130 cc big bore kit on the Venture, we started to notice a problem. The boat would accelerate swiftly up to the 6700 rpm range…then stop accelerating in a way that felt premature. The 1050cc Sleeper Venture has an rpm peak in the 6700’s. We knew that the 1130 had the power to run faster…but for some reason it wasn’t. During a later (smooth water) radar test session with the 1130, the hull encountered some rippled water at peak speed. Instantly the tach numbers escalated 200 rpm into the middle 6900s. The radar speed instantly jumped 2.5 mph at the same time. At first we were quite proud. We assumed that we had reached some sort of planing threshold that allowed the extra rpms and speed by simply lifting a little more of the hull out of the water. What had actually happened was that the hull got going fast enough to lift up on the second set of chines. It bears noting that this hull was never intended to go this fast, nor was it ever intended to get up on this second set of chines. Either way we were happy about it…for a while.

Getting our 1130 up onto these chines (on glass water) takes some doing. While stuck at 6700 rpm, the rider can rock the hull around to sort of “coax” the hull up to the second set of chines. After our test riders mastered this, we were able to make long sustained runs at 60 – 61 mph. During these runs we found the down side of this “higher” ride. The hull would spontaneously “turn” itself in any given direction without notice. The rider had little or no control over the steering of the hull at this point. More than once, the shore based radar gun holders ran for cover as the massive Yamaha barreling towards them seemed to develop a mind of it’s own. The steering problem became so profound that we chose to cancel any further smooth water radar tests until the problem was fixed.

The most effective solution (to date) has been the fitting of a Jet Dynamics concave ride plate. Unfortunately Jet Dynamics (at this time) does not make a ride plate for the Venture. We re-drilled and fitted a plate that was originally made for the Blaster 701. Not only did this plate eliminate the high speed “self steering” problem, but it also eliminated the need for the rider to coax the hull past the 6700 rpm barrier. We experienced a quantum improvement of overall acceleration, along with confidence inspiring control at 60 mph on glass water. Jet Dynamics has told us that they will make this plate (soon).

All these tests were conducted using a stock scoop grate with the center bars cut out. This scoop would be the best choice for grudge racers interested in smooth water peak speeds. However for high speed rough water riding, The Pro Tec grate offered much improved hook up. This Pro Tec grate did scrub off 1.5 mph on the Venture hull riding on the second set of chines (glass water), but the rough water speeds would likely be much increased over our modified stock grate. For more info about this ride plate, contact Jet Dynamics…we will not make reproductions of our prototype.

About Carb Synchronization – Every one understands the importance of carb synchronization on the triple engines. However we have observed some important variables related to synchronized butterfly opening.

It’s important to understand that perfect butterfly synchronization at all throttle opening is always a plus, but not always attainable. Because of slight variations in the throttle shafts, and the clearance between all the shafts and the carb bodies, very few carb “racks” offer absolutely identical butterfly angles all the time. Having identical butterfly angles at full throttle is not totally crucial because the butterflies are all usually within the cross section of the shaft itself. However maintaining identical butterfly angles from 0 – 30% throttle is absolutely crucial. Particularly the butterfly angles at the idle position.

During our modification of 3 cylinder carb racks for our triple kits, we reassemble and synchronize each rack on the bench. However we have seen several situations where those bench “synced” racks offered visibly uneven butterfly opening angles at 0 – 30% throttle “on the boat”. After much inspection, we found the reason for this inconsistency. On the bench, our technician checks for perfect synchronization by rotating the throttle shaft end that connects to the cable. However on the machine, there is no rotating action. The throttle cable pulls in a straight line from one direction. As the cable is drawn, it first pulls out all the clearance between the throttle shaft and carb body in the number one carb. This slightly changes the fulcrum point where the #1 and #2 shafts meet. This shaft/body clearance, and changed fulcrum, can slightly delay the opening of the #2 and/or #3 carb butterflies.

On the water, the engine can still start well and idle okay. However it will often be idling from the opening of the 1 and 2 carb openings…the #3 butterfly is often still closed. As the throttle is snapped open, the front two cylinders (which already have fuel and air flowing) will react, while the #3 cylinder stumbles momentarily because it was passing no air or fuel. In other words, #3 takes a moment to catch up. To the rider, these uneven butterfly opening angles cause, what feels like, a low speed bog or possible lean condition. But it’s not a fuel mixture problem. You can tune the low speed screws til the cows come home…and you won’t fix it.

The only way to assure “absolutely perfect” butterfly synchronization in the 0 – 30% range, is by actuating the installed carb rack off of the idle stop screw using the throttle cable (that is “installed on the boat”). Turn the idle stop screw in and out, in half turn increments, to assure that you can see identical butterfly angles all the way down to full closed. Then assure that the #1 carb is not leading when the cable is initially drawn. We have found the Keihin carbs of the Kawasaki triples to be much more susceptible to this effect than the Mikuni racks on the Yamahas.

While it sounds a little complex, this inspection will pay big dividends in initial throttle response.

GP 1200 Water Speed Variations – During the autumn of ’97, the varying peak water speeds of GP1200’s had become a significant controversy among owners and rec sport jetski regulars. To be sure, there were many inconsistencies and many unanswered questions. For our part, we are primarily interested only to the point of meeting the needs of Group K kit owners…not all GP 1200 owners. Just the same, this following information likely holds answers for all of the above.

We received our brand new production GP1200 test unit in February ’97. Our pace of testing was immediate and full time. Weather conditions were perfect, and air density (in western AZ) commonly ran in the 100% – 105% range (this makes for uncommonly good performance and radar results). After one full day of break in, we radared the machine at 58.3 mph @ 6830 rpm (close to the speeds forecast by Yamaha). Given our long experience with big bore kits on the 1100’s, the prototype testing of this 1200 project ran uncommonly quickly. After only a few days of constant on water testing (a total of 13 operating hours), our finished Sleeper GP1200 ran 64.4 mph (@ 7000 rpm w/Solas J). By this time, the boat was exhibiting some “unfriendly” handling characteristics in very rough water at high speeds. Since we hoped to make these machines a prime endurance racing boat, we immediately went on to the testing of various handling components before pursuing more water speed.

Since no aftermarket parts were available yet, numerous prototypes of trim tabs, ride plates and scoop grates were fabricated and tested. Since our stock trim tabs were destroyed during the removal process, a set of aluminum plate duplicates were made to replace them. Of all the different trim tab shapes we tried, these aluminum duplicates of the stock tabs proved to net the best of handling and water speed. All in all, our handling testing had not brought about any big improvements in high speed rough water handling. We decided to carry on the rest of our performance testing on glass water only. With less than 25 operating hours on our GP, we had completed our Type 2 testing that yielded a batch of 68 mph (@ 6970 w/ Solas X1) glass water passes. We concluded that no more horsepower work could be done until the handling issues were resolved. At this writing (October ’97), we still feel the same way.

Through the course of ’97 we have seen a very broad variation of the peak speed abilities of other production GP 1200s. Very few, that we have seen, have equaled the 58 mph we saw in February. However very few were swiftly broken in, and then radared in 100% density air. For several months, we simply wrote off the speed variations as a reflection of “less than ideal” air and weather conditions. We justified this with the fact that very few, if any, other stockers were turning 6800+ rpm in stock form.

This perspective changed when we started to see customers, with Sleeper kits, whose water speeds and rpms varied greatly from our test boat. The only common thread among these machines was that they all had a lot of hours on them before the modification, and none of them had been tach or radar tested before the modification.

In time, we were fortunate enough to get access to one of these machines that was running speeds (and rpms) much lower than projected…this machine provided many of the answers we were looking for.

The first “gremlin” we uncovered was very unexpected. This machined had previously had a flush kit installed in one of the lines running from the head to the exhaust pipe. When this flush kit was later removed, the separation in the hose was mended with a two ended PVC barb that had a .235″ inside diameter (much smaller than the oem .310″ aka 8mm fittings). Apparently the GP1200 does not respond well to having the water to the exhaust pipe restricted (at all). The net gain in peak rpm, from this correction alone, was about 120 rpm. Unfortunately, this 120 rpm did not regain our lost speed, it only helped to uncover the next culprit.

During a glass water pass, under full power, this particular GP 1200’s rear end suddenly swung a couple of feet to one side. The result was an instant (and unwelcome) 45° turn at 60+ mph. Later examination led us to suspect the stock trim tabs, which appeared to be lifted very slightly away from the hull in a couple of spots. We had heard that “less than perfectly sealed and fitted” tabs could cause high-speed instabilities. With this, we installed a set of Riva tabs, which are identical replacements made of aluminum plate. The difference in radar readings, tachometer readings, and high speed handling, was immediately apparent. However even with this improvement, we still noticed the tail of the boat lifting (at around 64 mph) causing the nose to be driven downward. This causes the hull to “steer on the nose”…a very undesirable effect at 64 mph.

In this case, the problem was easily traced to the aftermarket sponsons attached to this GP. While these sponsons make for outstanding rough water turning, they caused an unacceptable amount of lift at the rear end of the boat at high speed on glass water. Installing the stock sponsons restored even more of our lost peak speed, not to mention the same high speed stability we remembered from our February test boat. All in all, our little amount of testing resulted in an 8 mph increase on this machine…without ever touching the engine adjustments. We suspect that machines like this one are less the exception, and more the rule.

Since this test scenario, we have advised numerous GP 1200 Sleeper kit owners to install the aluminum trim tabs. These customers have commonly experienced 4 – 5 mph increases with the tabs alone, to say nothing of the increases offered by going back to stock grates and sponsons. We are not about to defend the magazine tests that claimed stock GP 1200’s can run 60+ mph (we have never seen it). However we firmly believe that big measures of peak speed can be easily (and affordably) recovered from so-called “slow” GP 1200’s.

About Glass Water Scoop Grates – As most glass water grudge racers already know, the fastest scoop grate is the stocker with the center bars cut out. However for GP 1200 owners there is an additional option.

While the 1100 and 1200 Yamaha grates are identical with respect to mounting points, they do differ in intake area. The stock 1100 (Raider) scoop grate has a water intake opening that is measurably smaller than that of the stock 1200 grate. This smaller opening on the 1100 grate is a disadvantage in every area but one…high-speed glass water. On well modified GP1200s, the stock 1200 grate can actually over-stuff the front side of the pump during very high speed glass water runs. By installing the stock 1100 grate (with the smaller opening) this over-stuffing effect is reduced…thus increasing peak water-speed ability. Be forewarned that the rough water hook-up of this setup is “poor”. However the “glass boys” don’t usually care about that.

Exhaust Pipe Options – Yamaha triple owners have waited a long time for the aftermarket to offer an affordable and effective replacement for the stock exhaust pipe. In the meantime, GP 1200 owners (willing to run 110 octane race gas) do have an option. The rear body section of the 1100 triples has a rear stinger exit that is considerably smaller in diameter than the same stinger diameter on the 1200 rear bodies. This smaller stinger has two basic effects. The first is a noticeable increase in peak horsepower output in the higher rpm range. The second is an equally noticeable increase in piston crown operating temperature. We do not recommend this modification for anyone using pump gasoline’s. However for those “race gas” grudge racers out there who are looking for a little more top end power, this works.

1100 Sleeper Cylinderhead and Jetting Update – When we first developed our 1100 and 1130 big bore kits for the Yamaha triples, the styles (and specific gravity) of oxygenated pump fuel was varying widely. Since then, the consistency of fuels has become much more uniform (within the continental USA).

Our original 1100/1130 big bore kits were jetted to accommodate this wide range of prevailing fuels. More recently, we have revisited that carburetor calibration (for the Group K modified 38s) used on these Sleeper kits. These tests resulted in a new jetting arrangement that offers much better low and middle acceleration along with (in most cases) an increase in peak rpm. While there are changes to all the circuits, the most profound change is the use of much larger pilot (low speed) jets along with leaner mains. Because of this particular combination of changes, we noticed a slightly escalated operating temperature of the rear cylinder. To abate this problem, we further reduced the (already staggered) compression of the rear cylinder head dome. This reduced compression does not reduce power output in any way…however we consider the head mod to be a mandatory partner for the jetting change. This update would apply to all 1100/1130BB sets (both pump gas and race gas) modified before Dec ’97.

We are offering this upgrade to the current owners of these Sleeper kits at a cost of $50 for the re-jetting, and $50 for the head modification. Please call first to make us aware that the parts are coming our way.

1100/1130 Kit RPMs and CDI Boxes – We recently had a customer bring us his 1130cc Sleeper equipped Venture with a complaint of low rpm ability. He told us that no matter how he jetted the machine, his digital tach would never show more than 6700 rpm (high 6900’s is normal). We conducted every imaginable test and part replacement with no visible result. Even a new oem Yamaha cdi box did not appear to make any change. Throughout this entire battery of tests, the machine exhibited a “running over rich” kind of sputtering sound (yet big steps in leaning out made no change). This machine had our attention because we had received a couple of calls from 1100 owners who had reported this same high speed sputtering, and lower than normal tach readings (on both Group K and other 1100’s).

During one of our test outings, we had another 3 seat endurance machine that was radaring at just over 60 mph. After doing a few back to back tests on the two different machines, the test rider remarked that our problem Venture “felt” every bit as fast as the good one. We decided to run a quick radar test of the problem Venture. At it’s current 6650 – 6700 rpm (with a Solas J), we expected 56 – 57 mph (this is what all the normal stock bore Sleeper Ventures run). All four radar passes were over 60 mph…a speed that can only happen with 6950+ rpms.

In time, we realized that all our digital tachometers were not indicating “actual crankshaft revolutions”, but rather the number of observable spark pulses being delivered to the plug. We further realized that it would be possible for a weakened ignition component to permit rough operation up to the 6900’s while only providing about 6700 observable pulses for our tachometers. Initially we out ruled the possibility that a new oem cdi box could be the culprit. However it bears noting that the stock cdi box is only required to deliver a maximum of 6400 – 6500 rpm against the load of the very mildly pitched stock impeller. The load increases induced by the steeper Solas J, and the higher rpm/output of the Sleeper kit, could very possibly be enough to cause a new “stock spec” cdi box operate at the edge of it’s envelope.

The apparent solution is an aftermarket (Riva) cdi box. In the past, we have discouraged the use of these boxes on pump gas 1100/1130 kits because they could permit rpms that were not pump gas safe. However the most current Riva boxes do have a setting capability for 7100 rpm (a peak limit that can still be 92 octane safe for “average” recreational use). In the meantime, our problem Venture is getting the Riva box, and we will post the results.

Solas J Impeller Pitch Variations – The Solas “Super Camber” series of impellers was introduced shortly before we started development on our first Yamaha 3 cylinder Sleeper kits (on the 1100 Raider). We quickly found that the Solas “J” pitched impeller was a perfect match for the 1100 Sleeper kit (allowing rpms in the high 6900s). As the 1100 Venture and GP 1200 models became available, we found the “J” to offer the same ideal results on those models (as well as on our big bore Yamaha 1100 kits).

During late ’97 and early ’98, we began getting reports from 1100 and 1200 Sleeper kit owners about rpms that were uncharacteristically low. The GP 1200 and 1100 big bore kits that were supposed to turn into the middle 6900’s, were only able to attain 6700 rpm. After getting several of these reports, we finally obtained a Sleeper equipped ’98 GP 1200 that was experiencing the same problem. We spent a week inspecting and tuning this machine in an effort to recover the lost rpms…but the tuning had no effect.

After exhausting all the other variables, we installed a 2 year old Solas J that we knew turned high 6900’s in one of our earlier test boats. With this older J prop mounted, our ’98 GP 1200 Sleeper turned 6980 rpm, and ran 2.5 mph faster. When we compared the “old J” against the “new J”, there were several very obvious design differences. These differences were the cause of the heavier engine loading (and lower rpms) experienced with these “newer J props”. We contacted Solas regarding the nature of the design change, however they claimed no knowledge of any change at all…yet the newer props are clearly different.

Our next tests were centered around re-pitching the new J in an effort to match the performance of the older J. After several on water tests, we were able to make a specification adjustment to all the impeller blades that recovered all the lost rpm. The unexpected surprise was that the “adjusted” new J ran about 1 mph faster that the conventional “old J”. The acceleration in the mid-range appeared to be better as well.

As a result of these tests, we now perform this adjustment to all the Solas J props used for our kits. We would urge any 1100 big bore, or 1200 Sleeper, customers (running a Solas J) to check their peak rpms with a digital tachometer (on glass water). If your peak rpms are under 6840, you very likely have the new design J that would require the adjustment. The cost for the adjustment is $48, and in house turn around is 2 – 3 days. For customers who purchase a new J from us with their engine kit, the impeller adjustment is done at no extra charge. Please note that we cannot weld of repair damaged impellers. This specification adjustment can only be performed on good condition impellers.

R&D Ride Plate Mod for GP 1200 – The ’98 GP 1200’s employ an exit nozzle that has a 5 degree upward angle to help lift the nose of the hull off the water. The net gain of this change is about 1 mph. Installing this nozzle on the ’97 hulls has the same good effect, and we recommend it.

The stock trim adjustment on all the GP 1200’s allows you to set nose ride height, however this is done by re-aiming the direction of the water exiting the pump nozzle. This re-aiming process compromises the efficiency of the thrush, and therefore peak speed ability. The more effective way to lift the nose of the hull off the water is by altering the bottom face angles on the ride plate. By changing these angles, the front of the hull can literally be lifted away from the water surface, thus decreasing water contact area (and increasing peak speed). It bears noting that, like everything else, it’s easy to cut too much angle on to the plate, resulting in uncontrollable porpoising.

We chose to start out with the R&D Pro Series ride plate (which already carries some multiple negative angles). After testing with many angle combinations, we found a specification for the ride plate that allows for maximum nose lift along with “porpoise free” operation on smooth water. While the cutting of accurate dimensions on ride plates is not common fare for most high performance machines, it is a mandatory modification for this Hammer 92 kit. While this mod still allows for reasonable rough water handling, it is intended to allow the best in smooth water speeds. Rough water racers may not benefit as much from this mod. We strongly recommend the R&D Pro Series scoop grate to go along with this modified ride plate (there is no mod for the grate). Group K price for this ride plate mod is $58.

Ride Plate Mod For XL 1200 – After having the afore mentioned experience with the GP 1200s, we decided to visit the same issue on the longer XL 1200 ride plate. The 98 XL 1200’s have a very “nose heavy attitude in the water. This nose-heaviness causes the rider to take excessive punishment at high speeds in rough water, not to mention a noticeable loss in peak speed ability. To reduce both these problems, Group K makes a machining modification to the stock ride plate. On our Sleeper XL test boat, this ride plate mod accounted for a full 1.5 mph increase in smooth water peak speed, and an instant increase of 80 rpm (on a digital tach). On pure glass, this mod did create a slow nose “bobbing”. However even with this slight bobbing, the radar numbers were 1.5 mph better. Once the XL was on anything rougher than glass, the bobbing vanished altogether. Perhaps the best benefit of this mod (on the XL hull) was the tenancy for the nose to “loft” over rougher water (as a motorcycle lofting the front wheel over rough terrain). The end result is a much less punishing ride through rough water, with no compromise at all in handling or turning ability. This modification makes a noticeable improvement in high-speed rider comfort in rough water (the speedo drive remains intact and functional). Group K price for this ride plate mod is $48.

Skat Trak “D Pump” Testing – Once we finished all our testing with our 70 mph GP1200 Hammer, we had several customers ask about the potential improvements of the all stainless steel Skat Trak “D Pump” used by so many closed course racers. Skat Trak was kind enough to lend us a pump to gather the test data.

The “D” pump is actually a stainless steel insert that is fitted into the aluminum outer housing of the stock Yamaha pump. This insert consists of an all stainless steel vane body assy, and a sleeve for the impeller housing as well. The stock Yamaha impeller diameter is 155mm. The “D” pump impeller diameter is 148mm. The folks at Skat Trak have found (after “lots” of testing) that the higher horsepower triples seem to yield the best overall thrust with this smaller diameter. Since they are the experts, we don’t question their judgment in this area.

For our tests, we decided to retain the stock ’98 exit nozzle with it’s 5 degree upward angle. The Skat Trak “D” pumps are available with variable length bearing cover “cones” that extend rear-ward into the exit nozzle area. Using a longer cone increases the restriction (and pressure) at the exit nozzle. That is, a longer cone has the effect of using a slightly smaller diameter exit nozzle. Using a shorter cone has the effect of slightly boring out the exit nozzle. The available cone lengths are 2.5″, 2.75″, and 3.0″. We used the 2.75″ inch cone for all our tests.

We had lots of prop testing under our belt with the 155mm setup (on our triple pipe GP1200 Hammer) that had proven to us that the best results would be had with an rpm peak between 7350 and 7550. We had a broad range of 148mm Swirl props to choose from. At first we were very concerned about a loss of peak speed ability that often can come along with the swirl design impellers…but that concern didn’t last too long. In short order, it was down to two props, the 16/26°, and the 16/27°. With a peak rpm of 7530, the 16/26° was clearly the best accelerating and faster choice (although faster by only .4 mph). The 16/27° also accelerated great, but pulled the motor down to a more comfortable 7410 rpm peak. The milder 16/26° would have certainly been our choice for closed course racing (so long as the fuel octane was “spiced up” with a percentage of race gas). We considered the 7530 peak rpm to be a little too high stress for recreational applications…but boy was it quick and fast. The 16/27° in no way “tamed” the GP, but the ride did become a little less violent and abusive (but some folks like that). For sure, the crankshaft and other moving parts would tend to prefer the 16/27°. In the end, the higher revving 16/26 only gained us .5mph over our best setups using the stock 155mm pump (the 16/27° ran the same as our 155mm pump).

Unfortunately, our full battery of (glass water) radar tests never really uncovered the “D” pump’s best quality…that is hook up. However further testing gave a new dimension to the importance of that hook-up. During our many weeks of radar testing with the stock 155mm pump on our GP1200 Hammer, we learned that this hull delivered it’s best radar numbers on pure glass water, and significantly slower numbers on “anything that’s not glass”. While a mild 1 inch wind ripple didn’t noticeably effect the stability of our GP, it made for a 2 – 3 mph loss on the radar gun. As the water became “rougher” (3 – 4 inch chop), the radar speeds dropped 5 – 7 mph.

When we installed the Skat Trak “D” pump, all these rules changed. Water conditions that used to knock 2 – 3 mph off our GP, now made for losses of less than .4 mph…and the rough water that used to lose us 5 – 6 mph, couldn’t even take 2 mph off the “D” pump GP12. Equally important, our GP handled and hooked up like it was still “riding on glass” as we flashed through 4″ wind chop. With either pump, on glass, our GP could easily wax any boat on the water. However, with the “D” pump, we were no longer “searching for glass” to choose off some of the better rough water machines.

In short, the “D” pump may not pay big dividends on glass water, but we live in a world were most guys want to race longer than that first pass on the glass water. The D pump (with prop) is not an inexpensive modification. However if you want your modified GP1200 to be the absolute king of a wind chopped piece of water, there is no better place to spend your money than on a Skat Trak “D” pump.