Project 90 : Detailed Performance Upgrages

[90GTVert]

Project 90 : Phase 1 Review

[/color][/size]


I've completed all of the modifications that I wanted to try one at a time on a stock 90cc Minarelli. Now I'd like to give my thoughts and opinions on some of the best mods. To do so, I've put together a top 3.



#1 : Cylinder Porting And Adjustment


This modification topped the charts for max speed and did as well as any other mod for cruising speed. I may have been a little more aggressive than some would be or should be for a daily ridden street scooter, but excellent results could be had keeping the setup more mild as well. A mild setup can be done for the cost of gaskets. If you push it a little more as I did, you may need to add on about $30 for a high compression cylinder head. Even with the head, you should come in at $50 or less.


If you don't feel comfortable porting a cylinder, a big bore kit like the Hoca 54mm could be a good option as well. With a price of $150 though, I would choose #2 above the big bore kit for starters.




#2 : Hoca Performance Variator


The Hoca variator came in a close second. It showed as much of a gain in top speed as any mod other than the cylinder porting. Cruising speeds increased as well, on par with some of the other top mods. One of my biggest reasons for liking this one so much is that the scooter was really much more pleasant to ride. Accelerating well, staying in the powerband, and feeling smoother in general. This one will set you back about $50, but it is well worth it IMO. I just don't think the stock variator does the 90+cc engines justice.




#3 : Performance Exhaust


An aftermarket exhaust would be my next choice. The stock cylinder may be a little mild to make the most of any of the performance pipes, but you can still see great gains. This class is next in line to the previously mentioned mods for top speed and cruising speeds are right there with the rest of the top mods. Acceleration is also comparable to the rest of the bunch. Check out each individual test to see which pipe you like the best.




Honorable Mention


I have to mention the other free mods. Removing the airbox snorkel and modifying the torque driver both showed gains and they cost nothing. You do need to upjet with the airbox mod, but for the purposes of this project I leave out tuning supplies in costs. Neither of these things showed big improvements at this stage, but it's hard to argue with free.

[/center]

[90GTVert]

Phase 2 On Hold


I began testing and tuning with the Hoca variator along with the ported cylinder to use as a baseline for the second phase of Project 90. When I pulled the stock variator off, the threads on the crankshaft were stripped. That really didn't surprise me, because I've been seeing metal shavings every time I remove the nut from the time I got the engine. It looked like it had been re-threaded already.




I threaded it to the largest size that I could get good threads with, which turned out to be M9 x 1.25.




I continued testing and was hoping to have a couple of updates to post today, but the threads on the crankshaft stripped on a ride yesterday. The new threads only lasted a bit over 100 miles.




The threads stripped and the nut backed off enough to grind the fan against the cover and kickstarter.




The key broke as well.




Not only are the threads gone, but the splines on the crank were stripped some as well at the end.

[90GTVert]

Project 90 : Phase 2


After the threads on Project 90's crankshaft failed, I picked up a used stock crankshaft from a 2003 Dinli Helix 90 (90cc minarelli powered ATV) on eBay. It came complete with bearings in good condition.




I tore the engine down and removed the old crankshaft.




Unfortunately, the previous owner had machined the 90cc crankshaft down so the cases would need less machining.




That meant that my unmachined stock crank would not fit into the cases as-is.




Machine work isn't cheap, so I decided to try to clearance the crankcases myself rather than having a machine shop do that or cut down the replacement crank. I figured I couldn't do much worse of a job than the previous owner did.






I spent A LOT of time using a Dremel and an electric drill along with grinding stones and 60-120 grit flapper wheels to open up the crankcases. The job could have been completed faster using large carbide burrs, but I chose to use the slower working tools to reduce the risk of removing too much material or having the carters look as they did from the last builder. The flapper wheels and stones did a nice job. The tricky part was trying to cut the larger case even all the way throughout, since it's much thicker. It was a very time consuming process that I'm not sure that I want to do again or would recommend for anyone to try. I did eventually end up with a set of crankcases that would fit the new crankshaft, and the finish turned out surprisingly well considering electric hand tools were used.










While I had the cases apart and right in front of me, I wanted to match the cases to the cylinder. Before I did that however, I wanted to port the base of the cylinder out. I never touched the base of the cylinder when I did the previous port work to it, because the case ports were too small for it to matter. Here's a look at the cylinder's base untouched.




In this picture, you can see half of one transfer port opened up to give you a good look at the changes made.




Here's the cylinder after porting. It's opened up as large as I felt comfortable with to leave enough of a sealing surface at the base. I didn't work very far down into the ports. I just tried to blend the large openings into the rest of the port.




Once the cylinder was ported, I started out by matching it to the boost port in the crankcases.






Then I opened up the transfers in the case to match the cylinder.




Again, since the cases were apart and right in front of me, I decided to go just a little bit farther. I did some mild "trenching" on both case halves. Trenching is cutting trenches into the case to allow the fresh intake charge coming from the reeds to be drawn directly into the transfer ports. Here's the large case half before any trenching.




These pics show the same case half after a mild trench was cut. This was a very quick job and the trenches were not cut very deep in an attempt to avoid cutting through the case and needing to repair it.






I went a bit far on the other carter and ended up with a hole.






I filled the hole from the outside using JB Weld after roughing up the surface to help the JB Weld stick.




Once the JB Weld cured, I sanded it down to complete the repair and case porting.




I installed the crank and assembled the cases, sealing them with ThreeBond 1194.






After installing the cylinder with the same amount of base gaskets as I used after the porting and adjustment described at the end of the first phase of this project, I noticed that the piston was traveling both higher and lower in the bore than it was with the previous crankshaft.






Adding up the stroke I could measure within the cylinder with the distance the piston traveled above deck, I discovered that the new crankshaft has a 45mm stroke instead of a 42mm stroke like the last one. That means it will travel about 1.5mm higher and lower in the bore, so I would need to set the cylinder up a little differently and recheck port timings. The additional 3mm of stroke also means that the 89cc engine is now 96cc.




I used more base gaskets and swapped to a cylinder head that was recessed in order to get port timings, compression, and squish clearance where I wanted them. I'm not going to go into details about these numbers, for reasons you'll soon discover.




After installing the cylinder, I installed the stock reed assembly and intake and did a leakdown test. Somehow I failed to take a picture, but I used a reed spacer (as shown earlier in the project) when I installed the reeds. I had my doubts about the cases sealing after seeing some scars from the previous owners on the sealing surfaces of the case halves, so I was pleased to see the cases held pressure very well.




When I went to install the flywheel onto the new crank, I discovered that it was too large. I was unaware that 90cc crankshafts could come with a small taper for the flywheel, but I learn something new every day. Forum member 2strokd was kind enough to hook me up with a small taper flywheel very cheap. Thank you 2strokd! Here are the large taper and small taper flywheels side by side.






I was glad to learn that the crankshaft did have large splines for the variator, as I originally expected it to, so I didn't need to acquire a new variator for the project and the CVT went together without a hitch.




I finished up the engine, bolted it into the scooter, and ensured that it was in tune. Later that night, I headed out for a ride on the new and improved Project 90. Everything seemed alright for the first few miles. As time went on, it started to feel like the engine was getting progressively weaker. When I tried to tuck down and top it out, I was still just barely topping 50MPH. By the time I got to the town that I was headed to, just shy of 10 miles from my home, I was hearing a strange noise. I decided to turn around and head back in hopes of avoiding a breakdown.

Shortly after turning around, I noticed that the noise I was hearing was getting louder and I also realized that it sounded like a compression leak. I pulled over to investigate. It definitely sounded like the noise was coming from around the cylinder head. I was hoping that perhaps the spark plug was loose, so I shut the scooter off and removed the spark plug cap. The spark plug was tight, so I wanted to get back home and check the head gasket and cylinder nuts. Well, it's not a great idea to shut off an engine that's low on compression. The scooter wouldn't restart and I ended up going home in the back of a truck that night.

I removed the cylinder shroud and found that one of the cylinder nuts was not even finger tight anymore and the rest weren't as tight as they should have been. I assumed this due to the base gaskets compressing more with heat. I re-torqued the cylinder nuts and did a compression test. I found that compression was significantly higher than it should have been from the gaskets compressing and moving the piston closer to the cylinder head.

I removed the cylinder and found that the small sealing surfaces on the cases and base of the cylinder were causing the gaskets to compress more than normal and even allowing the gaskets to push outward enough that I was surprised it stayed sealed.

I decided to use a base spacer instead of a stack of base gaskets to eliminate the problem. Forum member 190mech donated one spacer that he had already made during one of his builds and some aluminum of varying thicknesses so that I could fabricate my own spacers. Thank you 190mech!




I set the degree wheel up on the engine and started making and checking the specs with spacers of different thicknesses. I ended up choosing a spacer 1.2mm thick. When the cylinder was originally ported and installed with the 42mm crankshaft, I set it up with 186° exhaust and 130° transfer durations. The 1.2mm spacer gave me 188° exhaust and 130° transfer durations. That's close enough for me and should still work well with the exhausts that will be tested.





The 1.2mm spacer also put the piston in a good position at TDC and BDC.




I used a thin coating of ThreeBond 1194 on each side of the base spacer and installed the cylinder. With 1.1mm of squish clearance I saw 180psi on a compression test, just where I wanted to be.




I finished up the engine's assembly and put it back in the scooter. I started the scooter and thought I heard an exhaust leak when I took it around my yard for a lap. I wasn't sure, because something in the stock exhaust now rattles and sounds a lot like a leak at times. I decided to verify that both exhaust bolts were secure, and snapped one off in the cylinder. I removed the exhaust and attempted to remove the bolt with no luck so I removed the cylinder. I tried a bolt extractor and that broke off inside the cylinder.




After a little more thought and effort, I decided to take the cylinder to a machine shop to have it extracted. I was afraid I'd end up ruining the cylinder, but they have more equipment to ensure they're drilling straight and to cut into the hardened extractor. I got it back about one week later for $30 with a helicoil already installed. The machinist told me this was the toughest extraction he's had to do in a long time, which made me feel better about having someone else do it.




I got the scooter running yet again and took it for a 79 mile ride. It ran well, and seems to have a little more torque than it did before the new crank and porting. It was completely dead at low RPM without the new mods. While it still wasn't strong at low RPM with the latest setup, it would pull through the low revs and get going faster. Nothing drastic, but an improvement. Cruising speeds were 48-50MPH most of the time. I was surprised to see a top speed of only 54MPH @ 9,150RPM with 6g sliders. That's 1MPH slower than the ported cylinder on the stock bottom end from phase 1 of Project 90. Gas mileage was good, averaging 58.6MPG.




I checked the drive belt and found that it has worn down to 16.05mm wide, so I replaced it with a new belt that measured 17.1mm wide.




The new belt did the trick, and I saw a new max speed of 57MPH @ 9,120RPM.




I never had any intention of keeping the stock variator for this stage of the project, I just wanted to provide more data to compare it with the Hoca variator that I have been so pleased with. I installed the Hoca variator and settled on 6 gram Dr Pulley sliders in it after some testing and tuning.




Even with the Hoca working well, I didn't see an improvement in top speed. I was still seeing 57MPH, although it was at only 8,660RPM. Lighter sliders could get the RPM up, but speed didn't increase. Acceleration did show a dramatic improvement switching from the stock variator to the Hoca. As I said in phase one when I used the Hoca variator, it does a much better job of keeping the engine within it's powerband. The only negative would be that low speed cruise RPM increases just a little versus the stock variator. It's really not a big deal though, and the increased acceleration is well worth it. After riding with the Hoca, I wouldn't want to go back to stock.




I decided to modify the fixed half of the front pulley in an attempt to gain more top speed. I started with a drive face that had been filed flat in the center to allow the front pulley to close tighter, as you can see earlier in this project. I then filed it in order to create a linear angle from center to outer edge of the drive face. The image below shows the difference in the angles on the drive face before and after filing. I felt that the steady angle, as shown on the right side, would help to keep RPM more consistant when acclerating.




I left a small flat section in the center of the face for the drive boss to sit against. If you choose to try this mod, take care to keep all angles uniform and smooth. After filing was completed, as shown below, I sanded the face up to 400 grit paper to smooth it out.




The first image below shows how far the stock fixed half and the Hoca variator are able to close. The second image shows how far the newly modified outer drive face and the Hoca variator are able to close. Closing tighter should allow the belt to ride higher in the front pulley to drop RPM and/or increase speed.






The modified face also creates more space for the belt to ride lower in the front puller for a deeper "first gear" without using spacers on the drive boss. I used a slider weight as a reference in the pictures below to allow you to easily see the difference in the stock (first pic) and modified (second pic) faces.






Before installing the CVT cover, I drilled a 1" hole just in front of the clutch in an attempt to reduce heat a bit more within the CVT cover.




I took a 70 mile ride once everything was back together again. Acceleration RPM at low speed picked up just slightly with the modified drive face compared to the stock fixed half, but RPM dropped just a bit at higher speeds. Top speed increased by 1MPH to 58MPH at only 8,460RPM, 200RPM less than it took for 57MPH with the stock outer pulley half. Cruising speeds ranged from 48-52MPH, depending on conditions. Fuel economy was 55.8MPG, just shy of 3MPG less than the ride with the stock variator. Revs stay a little higher on average with this setup though, and I was riding a bit harder. The Hoca variator is definitely staying as part of the baseline for phase 2 of Project 90.





Project 90 : Phase 2 : Baseline Specifications

Bore : 52mm
Stroke : 45mm
Displacement : 96cc (95.56747199999998cc)
Cylinder : Stock Ported w/1.2mm Base Spacer
Exhaust Duration : 188°
Boost/Transfer Duration : 130°
Cranking Compression : 180psi
Squish Clearance : 1.1mm
Cylinder : Stock Ported w/1.2mm Base Spacer
Cases : Port Matched & Mildly Trenched

Carburetor : Stock 49cc/90cc
Main Jet : 90
Pilot Jet : 22.5
Needle : Stock. Clip set in next to richest position.
Air Filter : Stock Airbox w/o Snorkel

Fuel System : Stock tank, 1/4" hose, 1/4" fuel filter, manual petcock.

Oil Delivery : Pre-mix. I removed the stock oil injection system in favor of pre-mix in order to provide consistent oiling with any modification. I am using AMSOil Sabre Professional 100:1 Premix, mixed at 50:1.

Intake : Stock
Reed Block & Reeds : Stock w/Reed Spacer

Exhaust : Stock 49cc

Ignition System : Stock CDI, stock stator/flywheel, Bando coil, NGK BR8HS spark plug.

CVT Belt : 788 17 28 Bando
Variator : Hoca Performance Variator w/Modified Fixed Drive Face
Roller Weights : Dr Pulley 6g Sliders
Clutch & Bell : Stock
Contra Spring : Stock 90cc
Clutch Springs : 1,500RPM

Final Drive Ratio : Approximately 10.125:1


Max Speed : 58MPH
Cruise Speed : 48-52MPH
Fuel Economy : 55.8MPG


Other Info For Reference

Elevation : ~40ft Above Sea Level
Rider Weight : 250lbs
Fuel : 91-93 Octane (Premium)

[90GTVert]

Phase 2 : Leo Vince SP3

The first test in this part of the project is the Leo Vince SP3 50-70cc exhaust. This exhaust was tested in phase one of the project, so you can check that out HERE to see pics, video, and installation details.


Installation of the SP3 required moving up to a 95 main jet. Needle position remained the same. I made a slight adjustment to the mixture screw. No changes were necessary in the CVT.


I went for a 58 mile test ride and observed 47.3MPG. Full throttle cruising speeds increased to 50-53MPH in most conditions. Max speed improved by 1MPH over the baseline to reach 59MPH. Max RPM observed was 8,560RPM. I could feel increased power throughout the revs with the SP3. As setup to work with the cylinder, revs rarely dip below 6,000RPM and are most frequently above 7,000RPM. Peak power seemed to be in the 7,500 to 8,200RPM range. The whole RPM range seemed very smooth and part throttle acceleration was even pretty impressive. I feel that the main reason max MPH didn't increase more is that the SP3 is just not wanting to rev higher. Even without increasing RPM, there is an obvious increase in power with the SP3.




I was not expecting the SP3 to work as well as it did, since it's intended for 50-70cc engines, but I was pleased with it. I think future testing will show that it's not the top dog for even the stealthy street pipes, but it was certainly a nice improvement over stock. The powerband didn't feel peaky at all, which made it very pleasant to ride in town or on open roads. Add in the facts that the SP3 is quiet, easy to install (as is the other Leo Vince pipe I have, the ZX), and required only a main jet swap and a tweak to the mixture screw and I can't say it's a bad choice or at very least don't hesitate to try it if you happen to have one left over from a past 50-70cc setup or can get a great deal.

[90GTVert]

Phase 2 : Down Again


I was in the process of testing the jetting on the next pipe upgrade today, when Project 90 died on me. I was doing a WOT pass to check jetting when it died. My first thought was a soft seize, but I realized I had no compression when I tried to kick it over.

I pushed it home to disassemble the top end and see what was going on. Upon removal of the cylinder head, I saw this...






I removed the cylinder, and immediately noticed that one circlip was pretty much completely gone.




On the other side, part of that circlip was missing as well.




Perhaps missing isn't the right term. It was still in the piston, just not where I wanted it.










The circlip chunks did a little damage to the cylinder as well.




I have a new piston kit on order and I'll go over the cylinder. It should be OK after the new piston is installed. This is the second time a circlip has failed on Project 90. I've never had a circlip failure before this project, and I've swapped quite a few top ends and put a lot of miles on similar setups. Both failures were the same type of circlip. I was already certain after the last failure that I would like to use a different style of circlip, but I used this style again because I had them around. I never expected another set to fail so fast again, and I knew this build wouldn't stay together for thousands and thousands of miles, so it didn't seem like that bad of an idea at the time. Oops.

[90GTVert]

Phase 2 : Repaired And Ready To Continue Testing


After the circlip failure in the last update, I picked up a new piston kit.




The kit came with the circlips shown at the top of the image below, but I used a set of circlips as shown in the bottom of the pic for assembly. It would be a good idea to use a bit of loctite on either style for added peace of mind. I used nothing. As much as I don't want another engine failure, I'd like to see if I have better luck with this style of circlip installed just as the others were.




I used a ball hone to hone the cylinder. As the pics show, I still have gouges in the cylinder wall, but it should be alright.






The head suffered some damage, but I cleaned it up and it can be re-used.






I put everything back together and verified that squish clearance, timing, and compression were all where they should be. Just as I thought I was ready to resume testing, I started having problems with the electrical system. To make a long story short, I ended up making a complete custom wire harness for this scooter and all is well now. Project 90 is back on the road.

[90GTVert]

Phase 2 : Phongeer Exhaust

The next pipe on the list to test is the Phongeer exhaust. This exhaust was tested in phase one of the project, so you can check that out HERE to see pics, video, and installation details.


Tuning the scooter after installing the Phongeer exhaust with the Phase 2 setup required a bit more effort than past pipes or this pipe on the otherwise stock setup. Jetting did not change. It runs well with the 90 main jet, just as the baseline did with the stock pipe. I spent quite a lot of time swapping up and down on main jet size and changing needle positions though. I did this because I had a rich stutter or bog at times trying to get to WOT. I ended up swapping from 1500RPM clutch springs to 2000RPM clutch springs. This helped the engine get past the dead low rev range faster, that I think the pipe is accentuating. No change was necessary to the roller weight. After installing the clutch springs I settled on moving the needle to the next to leanest (from next to richest) position for best results and re-adjusted the mixture screw. There is still a slight sputter, but nothing to be concerned wit. Essentially if I open the throttle all the way, quickly, from a stop, I get about two to three quick sputters (pop pop pop <please forgive my sound effects) and off it goes. This happens in a split second now, but before the clutch springs and needle position change it was quite a severe bog at times.


I took P90 out for a 31 mile test ride and managed 50.11MPG. Wide open cruising speeds ranged from 50-55MPH in most conditions. The faster you go with limited power (like a scooter), the harder it is to maintain a constant speed. Max speed increased 3MPH over the baseline to 61MPH. Project 90 has finally broken the 60MPH barrier! Maximum RPM observed was just shy of 9,200RPM. The Phongeer exhaust definitely likes to rev a little more than the previous setups, but power is increased anytime above 6 or 7,000RPM greatly. This pipe, coupled with the other modifications for Phase 2, is the first time that Project 90 is really beginning to feel like a 90cc should in my opinion. Strong power throughout the throttle. There's no need to go WOT or anywhere near it to get a strong start and it doesn't take much throttle to cruise unless you want to cruise fast. I believe this is why the gas mileage showed a little improvement over some setups.




I really enjoyed riding for this test. I believe this could be an excellent, reliable, setup for someone that enjoys riding reasonably quickly. If you ride even slower than I do, you could probably see better than 50MPG, which isn't bad for a 90cc 2T. It's plain enough and quiet enough that it doesn't draw much attention as well, so it's a great choice IMO. The porting on Phase 2 of this project is actually tailored for a bit higher revs than this pipe is meant for, so with porting to match I think it could be even stronger.

[90GTVert]

Phase 2 : Leo Vince ZX


Next up, the Leo Vince ZX. I tested this pipe in the first part of the project as well, so click HERE to see more details, pics, and video of the Leo Vince ZX 100cc in action.


I moved up to a 92 main jet and put the needle next to the richest position. It felt a little bit too rich at part throttle at times, but I got a lot of detonation at part throttle at higher speeds and the richer needle setting seemed to help that just a little bit. Throttle response was good, so I had no problem setting it slightly rich at part throttle. I stuck with the 2,000RPM clutch springs, as used on the last pipe test with good results.


I noticed a lack of RPM compared to what I was expecting. I had been seeing the revs hanging around 8,500RPM with the 6g sliders the previous setups used. I used to use this pipe with my first 90cc setup and it ran strong up near 10,000RPM, usually seeming to run strong until about 9,700 or 9,800RPM as I recall. I moved to 5g sliders and saw just over 9,000RPM, but it felt like nothing was there after 8,800 or so. I moved to 5.5g sliders and the revs would hang around 8,600 to 8,800RPM most of the time I was full throttle and that seemed to be the best option.


I rode 34 miles and saw only 39.7MPG. This is far below some previous tests. I must admit I was quite hard on the scooter, spending the majority of the ride at WOT or close to it. WOT cruising was good for 50-55MPH. I saw a max speed of 58MPH. I found this pretty disappointing considering the baseline for Phase 2 with the stock pipe was 58MPH and I've seen as much as 61MPH so far with the Phongeer exhaust. I might have been able to tuck down for a little longer and find an extra MPH, but the power just isn't there when I ask it to rev a little higher. My RPM was around 8,900 at top speed.





Acceleration felt good. Throttle response was good. Cruising speeds were good. All of that said, it just feels like something is missing. I know this pipe has more in it. At this point I am inclined to believe that without a larger carburetor this engine just isn't going to run as it should when attempting to rev it into the 9,000RPM range. When I ran my other setup in the mid 9,000 rev range all of the time, it had a 24mm carburetor. That's a full 10mm larger than the stocker being used on Project 90 right now.


The ZX still made a nice improvement in acceleration and power up till close to 9,000RPM, but if I had to choose a pipe for this exact setup it would not be the Leo Vince ZX.


Since I was able to get out for a brief ride during the day for a change, I took the GoPro along and made a little video to show you how Project 90 is doing at the moment. Check it out.

[90GTVert]

Phase 2 : Peace Pipe


This time around I'm trying out a Peace Pipe, made by our very own 190mech. For anyone that doesn't know, John builds pipes for the big minarelli engines. Peace Pipes power the world famous Skeedr (owned by forum member 2strokd) that has taken a lot of trophies in NASRA's scooter drag racing events. I also have one of John's custom pipes on T1, pushing it to over 70MPH. He does excellent work and he's always evolving designs. Thank you to John for letting me play with this pipe as part of Project 90!


Alright, time to get down to business. It was immediately obvious that this pipe would not clear the center stand. I'm not a fan of this feature, because this scooter only has a center stand. John makes no claims of this being a convenient and friendly street pipe though. Other than that, it fit well and I had no problems getting it to seal. John uses nice thick flanges and large header pipes that match well with the exhaust ports on most 90-100cc minarelli cylinders. Rather than grooves, holes are drilled for the bolts to mount to the cylinder. I suppose this makes installation a little more difficult, but the reason it's done is to be sure the pipe's flange aligns perfectly with the cylinder.


It's a cool looking piece with the big beefy outlet and custom bracket.




One more drawback of this pipe is that it is loud. This pipe was made for power, not for good street manners. It sounds like none other. I really love the idle. A lot of people think the Leo Vince has a great deep sound, but this thing sounds like a much larger engine to me at idle. It sounds good revving too. The volume really ruins it for a real rider though. It's hard to get that in a video, but it's quite a bit louder than any other exhaust I have tested.





Normally this is where I'd tell you what jetting I used and what I had to change in the CVT. I won't be doing that for this test, and you'll find out why if you keep reading.


I attempted to get Project 90 in tune with this exhaust, but I never could get to a point that it would be suitable for any real ride distance. It appears that this is too much pipe for the little stock carburetor. Even when I run it rich on the main jet, it heats up and mild detonation quickly becomes severe and not long afterward it's trying to die out and I have to let off. It's hard to make it hit as well, because I was trying to run it rich to keep it cooler, so that makes it sluggish hitting the powerband.


Sounds like it must be slow, right? Wrong. I never did any more than short little test trips, but it could hit 60MPH sitting up whenever I wanted to on flat road. Think about that. The best pipe tested in this phase so far maxed out at 61MPH with me tucked in and stretching it out to squeeze whatever I can out of it on flat land. I didn't even manage to get the scooter properly tuned. It was overheating, surging, and sputtering at times at WOT. Even then, it went just 1MPH shy of the previous best with me not even trying for max speed. That alone should tell you the kind of potential this pipe has. Here's a GPS pic from one test run.




When John says this is a race pipe, he means it. It's loud, it requires removal of the center stand, and it kicks you in the ass when it hits and makes all of the other pipes I've tried look pretty weak. I was seriously impressed to see how it stacked up against everything else so far. I knew it was going to be a strong runner, but I didn't expect it to that degree. It's biggest drawback right now is that it just won't work with this little carburetor. It needs a larger carb to help feed it and to deliver more fuel to cool the cylinder. The engine would try to rev to a little over 9,100RPM, but it started acting up and just couldn't hold on for long with the heat building up.


If you want a race pipe for a 90-100cc minarelli build, I would look no farther. For a street pipe, it fails the test IMO. Again I must point out, this is intended as a race pipe though. As I said earlier, John is always working on new designs and ideas and he can custom build pipes, so don't be surprised if you see a version better suited for street use at some point. Thanks again John, I'm truly impressed and I can't wait to see what this thing will do later on in the project!

[90GTVert]

Phase 2 : YMS V8

The next exhaust tested is another new one to the project, the YMS V8. It's easy to find for a pretty low cost and seems reasonably popular, so I thought it would make a good addition to Project 90.

It appears to be similar to the Phongeer exhaust that I am fond of, so I thought it would also be interesting to see how they stack up against each other. The YMS starts off with a nice taper made up of many pieces, where the PG exhaust uses only a 3 piece pipe to taper into the can, which I thought might give it an advantage.




Here's what the YMS V8 should look like when installed.




Unfortunately, mine doesn't look like that installed. The picture above is without the header bolter up, just to show how it should look. This pipe has been a pain.


When I started trying to install the V8 exhaust I encountered a few problems. First off, I'm not exactly sure what's up with the flange, but the only way I could get it to seal well was to use the thick gasket supplied with the exhaust and coat it with a bit of ultra copper RTV.




Once sealed, it sat in this position.




Just to be clear, the shroud was not keeping this from bolting up, that's where it had to sit for the header flange to line up well enough to seal.




Even if you don't think sticking way out and pressing on the fan shroud is a big deal, it won't bolt up. The pics below show how the bolts want to line up. The top bolt is so far forward that I would need to cut part of the shroud away in order to get a bracket to fit (shroud is unbolted and pushed forward a bit in the pic). The bottom bolt isn't too great either.






If you're thinking, no problem... just cut that bracket off and make another from scratch, there's more. It also ran into the center stand with the scooter up on the stand. I had to put the scooter on jack stands to prop it up for installation so I could make it line up right to finally seal. The stand can be used, but it's definitely not an ideal fit.






I've heard of enough other people using this pipe that I'm thinking I must have got one with something off about it. Either that or there are a lot of scoots out there with poor fitting pipes. Regardless of what the deal is, I decided to make the YMS V8 fit as it should so it could be tested.


My goal was to modify the pipe to bolt up easily and clear the centerstand without affecting the way it performs. I started out by cutting the pipe a few inches from the flange.




Then I bolted the flange and small length of pipe to the cylinder.




I tinkered with turning the pipe different ways and ended up turning the pipe ever so slightly and grinding one side so it would connect at just a slightly different angle than before. I don't believe I removed more than 2-3mm total from the length of the pipe, which should have little to no effect on performance. Once I got it turned and angled how I needed it, I propped it up and tack welded it together.




Here you can see that there isn't a lot of clearance, but it's enough.




Once it was tacked into place, I took it to the bench and finished welding it back together.




The mounting bracket was much closer to working then, but not close enough.




I chose to cut down the original bracket and only leave a small portion of it in place. I drilled a hole in what was left of the bracket and used it to attach the YMS V8 to a bracket that used to belong to one of 190mech's Peace Pipes.




Once it was complete, I gave it a coat of flat black to keep it from rusting where it had been ground and welded.









I used a 92 main jet, with the needle next to the leanest position, and re-adjusted the mixture screw. As with the Phongeer exhaust, I get a little bit of a rich sputter very briefly on takeoff at times. The only change in the CVT from the baseline of Phase 2 are 2,000RPM clutch springs.


I headed out to take a test run with the new exhaust. Roughly 5 miles into that trip the scooter started getting louder and louder. Eventually it sounded nearly as loud as if it had no exhaust. When I investigated the cause, I found that the little bit of the stock bracket that I left behind had came off of the pipe and left a large hole.




I patched up the hole with some scrap metal and welded part of the old bracket on and drilled it to work with the mounting hardware.




Finally, I went for a 46 mile ride. I got 44.76MPG, cruising around 40 most of the time. WOT cruising speeds ranged from 49 to 53MPH on flat road, sticking around 50-52 most of the time. Max speed was 57MPH. This is the first time I've lost speed versus the baseline in either phase of Project 90 to date. 57MPH happened around 8,700RPM. It felt like anything above 8,500RPM was pretty flat. Even though the max speed dropped by 1MPH, don't think this pipe offered no benefit. Acceleration definitely improved. The YMS V8 seemed more comfortable from 7,500 - 8,000RPM than any other pipe I've tested. It seemed to me that peak power would be somewhere between 8,000 - 8,400RPM. I have a feeling this pipe would have shown better results in Phase 1 with the stock porting, unfortunately I didn't own it at that point.




After maxing out at 57MPH, I came home and swapped up to 6.5g sliders, from 6.0g. I thought perhaps the V8's ability to pull at lower RPM and lack of power above 8,500RPM would suit the heavier sliders a bit better. It was a bit more sluggish getting the RPMs up to the powerband, but I was impressed with how well it operated at a little lower RPM otherwise. There was no improvement in in cruising or top speed though.


I looked over the CVT while swapping weights to see if anything seemed wrong. I checked the spark plug. I even did a compression test. I didn't see anything wrong anywhere, so I have to assume this exhaust just can't create a lot of speed because even the stock exhaust seemed more willing to rev. It could turn out to be a function of the smaller carburetor, but I think it is the pipe.


All in all, I have not enjoyed testing this pipe. Elongating a bolt hole or using some spacers is reasonable to me when installing a new exhaust. Cutting it apart and re-positioning it is not. Even without interference with the centerstand, the flange is a bit harder to seal than the other pipes I've tested. It's not a big issue, but I found it needed a gasket and RTV for a good seal every time. I have to assume that I just got a bad one since others have used the YMS V8 and I have not heard these complaints, but I don't know for sure.

I think this might be a good choice for a setup with porting closer to stock, but I have a hard time recommending this exhaust for any purpose after my experience with it. You'll just have to decide if you want to try it. I picked mine up from Taiwan. That and the amount of time it sat waiting to be tested are the reasons I didn't return it. I would suggest finding a seller in your country and trying it quickly after it arrives if you wish to use this exhaust, just in case.

[90GTVert]

Phase 2 : MRP Exhaust


The last pipe I have around to test is the MRP 50-70cc exhaust. I had no plans of testing this exhaust, but I was asked to try it. I'm not a big fan of using tuned pipes for 50-70cc engines on the 90-100cc engines.


My first 90cc engine had trouble with overheating and soft seizing when I used this pipe. Initially the compression was a bit high on that setup, but even after lowering it I had issues with soft seizing over time. I added a 19mm carb as well, and that didn't stop it. It seemed to be a gradual heat buildup. I could take short rides and it was fine, but on a longer run of maybe 15 miles or more when I was cruising fast it would eventually start showing signs of overheating and ultimately soft seize. I got a 90-100cc pipe and the problem went away.


I have always thought the problem is that the smaller stingers on some 50-70cc tuned pipes are too restrictive for the 90-100cc engines. Here is a quote from A. Graham Bell's book, Two-Stroke Performance Tuning.

"You may find that a minor reduction in pipe diameter will raise the power
output, but do be careful. A stinger pipe smaller in diameter or longer than specified could easily result in engine overheating and seizure."


Bell goes on to specify that 50-80cc engines should have a stinger inside diameter of 17-19mm and 100cc engines should have a stinger with a 19-21mm inside diameter. The MRP exhaust's stinger has an 18mm outside diameter. So there's my reasoning for preferring not to use this pipe with 90-100cc engines, on to the test.


One thing that I like about the MRP exhaust is the 2 piece design. The short header section makes it easy to bolt the exhaust to the cylinder without holding up the whole exhaust. It also allows the header and pipe to be at different positions relative to each other, which can help with mounting it on some setups.




Everything fit well without issue. This pipe has been through hell and back, being the first performance pipe I ever owned. I've crashed with it on one of my scooters, had the bolts back out of the mount and dragged it home, and who knows what else I might be forgetting. It's not nearly this ugly stock and comes with a clear coat and blue silencer.




Here's a sound clip of the exhaust. This pipe has been repacked since it was stock. In stock form, this pipe was quite loud. After a couple of months it got really loud. I remember riding around and I could startle some people by revving it when I was riding in town. Right now, it's packed with FMF's 2 stroke muffler packing. It seems to work much better. The pipe is still loud, but nowhere near as bad.





Main jet size did not change from the baseline 90 main jet. The needle clip was set in next to the leanest position. I reduced slider weight from 6g to 5.5g. The MRP pipe is not revving any higher than other setups that used the 6g sliders, it just seems to lack the power to get the RPM up without the lighter weights. I stuck with the 2000RPM clutch springs installed for previous pipe tests.


I went on a 43 mile test ride and observed 41.37MPG, being pretty liberal with the throttle. WOT cruising speeds ranged from 48-52MPH. Max speed was 58MPH. Power felt pretty similar to the baseline test with the stock pipe. The cruising speeds and max speed are the same as the baseline. Max RPM was around 8,800, which should be a bit below where this pipe is meant to perform. Upper 8,000 to 9,000RPM seems to be the rev limit for this phase of the project with any pipe though. As I have mentioned in past tests, I think the 14mm stock carburetor might be the restrictor here. Based on just the performance data here, I see no real reason to use this exhaust on any tuned 90-100cc. I know the pipe made some power on some of my 70cc setups, but it's not impressing me with the "big" engine and small carb for sure.





Unfortunately, the lack of a performance increase isn't the only reason I wouldn't run out and buy this pipe for your big minarelli build. After the lack of performance for the last couple of exhausts, I've been checking some things out to make sure it really is a fair test and nothing else is wrong. I found a very small leak at the base gasket, but nothing that I feel would be noticeably detrimental to performance.


When I pulled the cylinder off to reseal the base of the cylinder, I found another problem. I was hearing detonation at times on the test of the MRP exhaust and, considering my past experience with this pipe, I was expecting a seize at some point. I never had any soft seizes though, so I figured everything would probably be fine. Not exactly. As you can see in the image below, part of the piston crown has been melted away. This is aligned right in the center of the exhaust port and toward that side of the piston.





Here's the underside of the crown.





Here's another quote from A. Graham Bell's Two-Stroke Performance Tuning. This is from the same paragraph as the first quote I gave you, explaining what to expect when the stinger causes engine overheating.

"Engine overheating, in its early stages, is indicated by the presence of oil burnt dark brown under the piston crown. On the next stage the burnt oil turns black, until finally 'death ash' appears. After this, the piston can be holed at any time."


Normally, I don't remove the head often. I leave everything alone as much as possible. Luckily, at least for the sake of this information, I started a spark plug a little off angle when I was tuning this exhaust. The threads weren't bad, but I opted to remove the head to run a tap through it to clean the threads up and hopefully avoid any future issues. Since I had the head off, I took a look at the piston and cylinder to see if everything looked alright, and the damage to the piston was not there at that point. So the piston managed to make it through multiple exhaust tests with no damage, but when I use the exhaust with the small stinger it begins to melt.


I feel as if this supports what I've been saying for a long time... don't use a 50-70cc exhaust on a 90-100cc minarelli or at least make sure you know a bit about the pipe first. Some exhausts will have larger stingers, so they aren't all bad. The milder street pipes that are not designed as a true expansion chambers are probably alright. I've had no issues with the stock 49cc exhaust or the 50-70cc Leo Vince SP3.

[90GTVert]

Phase 2 : Results & Review


For full details on all of Phase 2's specs and reviews, please see the detailed posts for each in this thread. I'll start out with a couple of tables that may help make the results easier to review quickly.


Prices may vary depending on seller and shipping charges. All prices assume no installation fees and account for no additional parts.

Modification	Max Speed	Comments	Price
Phase 2 Baseline	58MPH	96cc, Ported Cases & Cylinder, Hoca Variator	N/A
Leo Vince SP3 50-70cc Exhaust	59MPH	Some improvement in power. Worth bolting on if you have it around. There are better choices if you are buying a pipe for this setup though. Stiffer clutch springs, lighter rollers/sliders, and carburetor tuning may be beneficial or required.	$150
Leo Vince ZX 100cc Exhaust	58MPH	Noticeable power increase, but I feel this one needs a larger carburetor to work as it should. Stiffer clutch springs, lighter rollers/sliders, and carburetor tuning may be beneficial or required.	$250
MRP 50-70cc Exhaust	58MPH	I advise against this pipe for this setup. Severe overheating risk for an engine this size that can lead to engine damage. If that's not bad enough, I really didn't pick up much performance with this pipe.	$200
Peace Pipe 100cc Race Exhaust	60MPH	This is a race pipe and it was really too much for this phase of the project. Get a larger carburetor before using this one. MPH is sitting up, unlike the others at full tuck, and I wasn't even able to get it to a proper state of tune.	$250
Phongeer 100cc Exhaust	61MPH	Nice gains throughout the current powerband. Great choice at this stage. Stiffer clutch springs, lighter rollers/sliders, and carburetor tuning may be beneficial or required.	$150
YMS V8 100cc	57MPH	Despite the lack of top speed, this pipe did offer improved power. I believe it's better suited to cylinders a bit more mildly ported however, making peak power below 8,500RPM. CVT and carburetor tuning may be beneficial or required.	$150

Carburetor Jetting Changes With Modifications


The jetting changes were significantly smaller in Phase 2 than in Phase 1. I don't suggest using these numbers as any sort of guideline, but I've put them together for reference anyway. The baseline main jet was a 90.

Modification	% Change	# Change
Leo Vince ZX	+2%	+2
Leo Vince SP3	+6%	+5
MRP Exhaust	+0%	+0
Phongeer Exhaust	+0%	+0
YMS V8	+2%	+2

There is really one very clear winner among the exhausts tested in this phase of the project. The Phongeer exhaust worked better than all of the rest, hands down. It showed very nice gains that made acceleration, cruising, and pegging the speedometer more enjoyable. It's got an inconspicuous look and sound that make it a great choice for a street scooter. My only real complaint is that it needed some work to bolt up.


The Peace Pipe was very clearly capable of the most power, but it was also bery clear that it was too much at this stage. While I feel most of the exhausts tested may benefit from a larger carburetor, I believe the Peace Pipe really needs a bigger carb. I think when the project gets to a point where I'm testing exhausts with a new carburetor, the Peace Pipe will really shine.


The YMS V8 appears to like less RPM than the current setup is meant for. While it didn't impress me here, it could be a good choice for milder or stock 90-100cc engines. I strongly suggest researching this one before purchasing though. I had to invest a lot of effort into making this exhaust fit. I think I just got a bad egg, but I would never buy another without looking for reviews from others about the fit.


The two Leo Vince pipes showed gains, but they weren't impressive at this point. The ZX was definitely not making it to it's peak power RPM, again I believe this to be a function of the small 14mm stock carburetor.


I must absolutely advise that you stay away from the MRP 50-70cc exhaust for 90-100cc engine combos. The design causes too much heat buildup and it started melting the crown of Project 90's piston. Leave this one alone.


The baseline with the stock exhaust was not bad at all. It could make an excellent street setup. Obviously it looks and sounds stock, because it is. From what I've seen here, I would lean toward keeping the stock pipe at this point for a strong running budget build.


That should wrap it up for Phase 2 of Project 90, since I'm out of exhausts to test. Stay tuned for Phase 3!

[90GTVert]

Project 90 : Phase 3



Since I partially melted the piston crown testing the 50-70cc MRP exhaust at the end of Phase 2, I replaced the piston and checked everything over. I have noticed that the compression is down to 175psi from 180psi when setup exactly as it was before. That's not bad for a cylinder with 2,000 miles on it in this project that has suffered multiple soft seizes, circlip failures, and again the piston crown melting a little.





I went over the CVT as well, checking belt width and cleaning things up to ensure everything is working properly. I went with 2,000RPM clutch springs instead of 1,500RPM clutch springs that I used in Phase 2. Either works well.


Phase 2 was all about exhausts with a ported cylinder. I believed that the small stock carburetor was holding back some of the exhausts in Phase 2. Naturally, Phase 3 is going to be all about carburetors with a ported cylinder. My plan is to test a range of carburetors from the stock 14mm all the way up to a 28mm DellOrto carb. Hopefully we will see what size or range of sizes works best at this point.


I wish to keep things as even as I can for all tests, but I will have to use a different intake for the largest of the carburetors. There isn't much I can do about needing the larger intakes. You just aren't going to fit a big 28mm carburetor in a stock intake. If you could, it would be a major restrictor.


The images below show that the stock intake manifold has a sort of ring cast in that reduces it's inside diameter. The stock intakes have a marking which says they are 20mm. I measured 19.5mm ID at the ring.







I used a dremel with a drum sanding bit and ground away this ring to make the intake path a uniform size. This creates a 22mm ID. Motoforce, and probably others, make a 22mm stock style intake. I own one and as far as I can tell it's just a stock intake made without the ring inside. Save yourself a little money if you need a slightly larger intake and do this yourself if you feel comfortable with it. Now the intake for the carburetor up to my 21mm can be the same. Here's a look at the intake after the mod.





The other thing that simply won't work with some of the larger carburetors is the stock airbox. I know a lot of people would probably opt for an open air filter, like a UNI or K&N, but I am not a big fan of these for street scooters. I plan to use s modified stock airbox that I currently use on a 103cc setup with a 24mm carb for the largest carburetors in Phase 3. It has a larger inlet and outlet. I also added three UNI filter vents to farther increase it's ability to draw in as much air as it needs.





If you would like more info about that airbox, have a look at the links below.

49ccscoot.proboards.com/index.cgi?board=carb&action=display&thread=1346

49ccscoot.proboards.com/index.cgi?board=carb&action=display&thread=2387&page=1


In order to keep things as fair as I can when using the stock airbox on the smaller carburetors, I did the same mods, just leaving the stock carburetor connection in place. These first images show the stock inlet vs the newly modified inlet. This can be cut away with a razor knife or filed or ground away easily.







Next, I moved on to installing three UNI filter vents just like the other modified airbox has.





I suggest plotting your locations from the inside so you can avoid all of the internals easier. UNI suggests installing these pre-filter. That means they would be installed on the opposite side of this airbox, facing engine heat. I don't like that idea too much. This is exactly how they are installed on the other airbox and it has been working well for me, so I kept it the same. The vents require 1" holes.





It's probably not necessary, but I opted to add a little hot glue around each vent since they are at the bottom of the airbox to keep them in place on those rough roads. The blue filter elements can be removed without removing the whole vent, so it's still possible to clean them when you clean your stock air filter.







Unless you are sitting or kneeling, you really don't notice the vents. There is just a little more sound from the airbox now, but it's not much different than just removing the snorkel. The plain, stock, appearance and the low volume levels should help to keep this one under the RADAR.





With carburetor mounting and air supply figured out (hopefully), I moved farther down the intake path to the reed valves. I don't believe that the stock reed block will do a large carburetor any justice. A well-flowing airbox, large venturi, and big intake will do no good without more area to flow through the airbox IMO. I replaced the stock reed block with a Stage 6 VForce 3.







Here you can see the stuffer included, which helps to smooth airflow into the reeds. It has a 28mm opening.




This reed block is actually manufactured by a company called Moto Tassinari, which is fairly obvious from all of the markings on the reed block. It has been brought to my attention that there are different versions of this block. Apparently there are the versions that they manufacture for Stage 6 and Top Performance and then there is a version that they sell themselves. I am told that the version they sell has a little different construction and better carbon fiber reeds, but you can expect to pay a little more for it.





I'll let the folks at Moto Tassinari explain the benefits they claim with the VForce3 design, since they do a better job of it than I will. Here's a quote from their website ( www.mototassinari.com ) :


"The unique design of the VForce reed valve system features double the reed tip surface over a conventional reed valve design. This has two major benefits: 1) The reed petals only travel half the distance as a conventional reed valve design to get the same airflow. 2) Because the reed petals only need to travel half the distance of that on a conventional reed valve design, there is a greater chance of them being fully opened as well as a better chance of sealing when they are "closed". A reed petal never truly "closes" in a classic sense, as it is always in movement. Consider this, an engine operating at 8,000 RPM results in a reed petal opening and closing 133 times in one second. This fact, sheds light on another VForce advantage. Since the petals are only traveling half the distance, they generally last longer are they receive less wear than the reed petals on a conventional reed valve design."


Sounds good to me. Just as in Phase 2, I installed a 5mm reed spacer under the block.





I wanted to add a little more data to this phase of the project. I was curious what, if any, effects larger carburetors may have on engine temperature. There's really only one simple way to find this out, so I picked up an Auto Meter cylinder head temperature gauge and a mounting cup to assure a clean install.







This gauge reads from 140°F to 340°F. I wasn't exactly sure what to expect, but I thought this would cover it.





One of the reasons for selecting this gauge is the sending unit. It uses a probe that threads into the cylinder head. You have to drill and tap the head to accept a 1/8" NPT fitting (and hopefully you'll drill a little more straight than I did). I have used this style of fitting as well as thermocouple rings that go under the spark plug before. Both options seemed to read the same, as far as I could tell, but I find the screw in style easier to work with. It requires a little more effort to install, but you don't have to deal with it every time you remove or install the spark plug.







There are gauges out there that are more compact and easier to install, but I am a fan of Auto Meter instruments and especially the Ultra-Lite series' styling. A big advantage over some gauges intended for 2-wheelers and ATVs is that the Auto Meter has it's own back lighting. My rides are primarily at night, so this is very important. Auto Meter includes a red bulb cover, but that provides a sort of pink glow with a yellow tint as shown below. If you want true red (or other color) illumination, I suggest picking up colored LED bulbs. I have these in my Mustang and they look much more manly with the red LED bulbs. ;D





With all of that tied up, I was ready to fire up the engine and get Phase 3 in tune. I plan to tune exactly as I have before. I will not be looking to the CHT gauge to aid my tuning. I will be tuning for the best performance and using spark plug readings if necessary.


Upon starting the engine and tuning the idle mixture, it became apparent than the engine was operating more smoothly than it has before. In the past it seemed that anywhere much below 1,400RPM the engine would struggle to run and want to stall out. Now I'm seeing that it will chug along happily down to about 1,050RPM. I have done a few modifications at once to begin this phase (Stage 6 VForce3 reed block, enlarged intake, airbox mods), but I believe this is largely from the new reed assembly. I still set the idle higher, to help keep the charging and lighting system working, but it's interesting to see the change. Here's a very brief video of the low idle.



www.youtube.com/watch?v=33bxgID97Co


I proceeded to check the jetting and adjustments, and was a little surprised to see that not much changed. I thought I would have needed to go a step up on the main jet after the airbox alterations, but the same 90 main as I used before the mods works well. The needle clip is set in the center position.


While tuning, I noticed cylinder head temperatures higher than I was hoping to see. So high that they were maxing out the gauge on wide-open runs.





I was hoping to see temperatures within the gauge's operating range, both to attain useful information, and because I felt it would be safer. The engine has likely been operating at these temperatures throughout the project, so I guess they are alright, but I at least needed to find a way to get the temperatures back on the gauge.


I looked to different cooling fan and shroud setups to accomplish this task. Project 90 uses the cooling fan that it came with. It's a typical minarelli cooling fan, found on 49cc and 90cc engines. There is a larger version that moves more air, and I happen to have one. I've only seen these on 90cc minarelli engines. The size difference is obvious.







I don't think these will fit under all cooling fan shrouds. Both of my long case Vento Tritons appear to have extra room under their stock cooling shrouds with the smaller fan in place. The larger fan fits perfectly. My other 2T scooter, the Keeway Venus, does not have this extra space under the fan shroud. Before you consider picking up a larger fan, have a look at your shroud.







I also decided to try out a cooling fan air scoop that I made a long time ago. The idea is that it "scoops" up air when riding at speed and helps to move it across the engine. I should also note, my "stock" fan shroud has had the grill removed to create the maximum open space.







You can find more info about the cooling fan shrouds at the links below.

49ccscoot.proboards.com/index.cgi?board=gentech&action=display&thread=394

49ccscoot.proboards.com/index.cgi?board=gentech&action=display&thread=393


I evaluated each setup with three bits of information. I began with idle temperature. I let the engine idle at 1,300RPM until the gauge reached it's minimum mark at 140 degrees. Once it was at 140°F, I let it idle for exactly 10 more minutes and noted the temperature.


I also used two different riding conditions to measure the performance of the fan setups. For each test, I rode for 2 miles at wide open throttle and noted the readings. I stopped at a landmark (driveway) to assure I traveled the same distance each time. I then turned around and cruised home at 40MPH and took note of that temperature. Here are the results.

Fan	Shroud	10min Idle	2mi 40MPH	2mi WOT
Stock	Stock	185°F	350°F	Max
Large	Stock	155°F	290°F	340°F
Large	Scoop	160°F	280°F	325°F

As you can see, there is a sizable temperature drop with the larger cooling fan across the board. Riding temps continued to decrease with the addition of the air scoop, other than at idle. The idle temp increase with the shroud could be a very slight restriction from the length of the scoop or it could just be a fluke. None of the idle temperatures concerned me, so I opted to use the large fan and scooped shroud for the rest of the project.


The down side of the larger fan is that it should create more drag on the engine and could have a negative impact on throttle response. I didn't notice either of these effects. I'm sure it makes a difference, but for a street ridden scoot I feel that the temperature drops are well worth a loss in performance that you probably won't even notice.


I got my large-finned cooling fan from a stock 90cc Triton engine. If you wish to pick up your own, I would suggest searching eBay or retailers of 90cc ATV parts. I have heard that some do not fit. I'm not sure if it's merely a shroud clearance issue, as I showed you before, or if there is more to it. There are aftermarket fans made as well. Parts For Scooters carries a Hoca "turbo" fan, part # 169-142, that appears to have a little larger fins and is available in different colors.


With the intake sorted out, the carburetor tuned, and the gauge and cooling covered, I could finally get out for the baseline ride of Phase 3. I rode for 32 miles and observed 56.95MPG, the best fuel economy thus far in Project 90. I will do my best to keep the same riding style for all of the tests in this phase, because I hope to see a good comparison of fuel consumption with various carburetors.


The idle, as mentioned earlier, is not the only thing that smoothed out. The entire rev range feels more smooth than it has before. Throttle response is very crisp. While neither of these give great gains in performance, they do make the ride more pleasant. I was thoroughly impressed with the results of the VForce3 reed block and smoothing/enlarging the stock intake.


WOT cruising speeds remained unchanged from the baseline of Phase 2, keeping me at 48-52MPH sitting up. Maximum speed increased by 1MPH, to 59MPH at just under 8,500RPM.





With outside temperature at 50°F, I observed 325-340°F cylinder head temperatures with long WOT cruising. Riding at 40MPH, I saw cylinder head temperatures of 275-285°F.


Overall, this seems like a good street setup to me. With a strict budget in mind, I would leave out the Stage 6 reed block, because it smooths everything out, but doesn't give any major gains. The large cooling fan is now one of my new favorite mods.




Project 90 : Phase 3 : Baseline Specifications

Bore : 52mm
Stroke : 45mm
Displacement : 96cc (95.56747199999998cc)
Cylinder : Stock Ported w/1.2mm Base Spacer
Exhaust Duration : 188°
Boost/Transfer Duration : 130°
Cranking Compression : 175psi
Squish Clearance : 1.1mm
Cylinder : Stock Ported w/1.2mm Base Spacer
Cases : Port Matched & Mildly Trenched

Carburetor : Stock 49cc/90cc
Main Jet : 90
Pilot Jet : 22.5
Needle : Stock. Clip in middle position.
Air Filter : Stock Airbox w/o snorkel. Inlet enlarged and UNI filter vents added.

Fuel System : Stock tank, 1/4" hose, 1/4" fuel filter, manual petcock.

Oil Delivery : Pre-mix. I removed the stock oil injection system in favor of pre-mix in order to provide consistent oiling with any modification. I am using AMSOil Sabre Professional 100:1 Premix, mixed at 50:1.

Intake : Stock, Enlarged To 22mm
Reed Block & Reeds : Stage 6 VForce3 w/5mm Reed Spacer

Exhaust : Stock 49cc

Ignition System : Stock CDI, stock stator/flywheel, Bando coil, NGK BR8HS spark plug.

CVT Belt : 788 17 28 Bando
Variator : Hoca Performance Variator w/Modified Fixed Drive Face
Roller Weights : Dr Pulley 6g Sliders
Clutch & Bell : Stock
Contra Spring : Stock 90cc
Clutch Springs : 2,000RPM

Final Drive Ratio : Approximately 10.125:1


Max Speed : 59MPH
Cruise Speed : 48-52MPH
Fuel Economy : 56.95MPG


CHT @ 40MPH Cruise : 275 - 285°F
CHT @ WOT Cruise : 325 - 340°F
Ambient Temperature : 50°F


Other Info For Reference

Elevation : ~40ft Above Sea Level
Rider Weight : 265lbs
Fuel : 91-93 Octane (Premium)

[90GTVert]

Phase 3 : 19mm Arreche


First up in Phase 3 is the 19mm Arreche. I tested this carb in the first phase of the project as well, so click HERE to see all of the pics and details of the Arreche carburetor.


I started off by moving up to a 95 main jet, from the 90 used when tested in Phase 1. I thought that since the stock carburetor needed to be upjetted from 85 to 90 from then until now, the Arreche probably would need the same change. After testing various jets, I ended up with a 90 main jet once again. The needle was set in the center position. No other changes were made from Phase 1's test of this carburetor.


I went for a 34 mile ride, seeing merely 38.18MPG. I had a very similar result, roughly 1MPG more, the first time I tested this carburetor. The gas mileage is pretty disappointing to me. I rode the same way as in the test with the stock carbruetor to the best of my ability and the Arreche was roughly 19MPG more thirsty than the stocker.


Throttle response was still excellent. There seems to be more power at lesser throttle positions. That made leaving stops without much throttle a bit easier.


Cruising speeds at wide-open throttle remained the same, 48 to 52MPH sitting up. Maximum speed was 58MPH at just below 8,500RPM. That's 1MPH short of the baseline benchmark, but power at speed felt about the same. As I've said multiple times now, the faster you go, the harder it is to stay consistent. I had a slight crosswind, and that's probably all it took to knock off that last MPH.





The biggest improvement I noticed was a cylinder head temperature drop. Ambient temp was 45°F. Cruising along at 40MPH I was seeing temps from 225°F to 240°F. Cruising at WOT the CHT showed 285°F to 300°F. That's quite a large improvement over stock.


While I think a larger than stock carburetor is a smart move on these big minarelli engines, it's more about room for expansion and keeping the engine cool according to this test. There really seemed to be no power gain at this stage. The poor fuel economy is the most disappointing part of this information to me.

[90GTVert]

More Trouble Along The Way


I tuned the Dellorto 19mm and went to ride it that night. All went well until I tried to stretch it's legs for the max speed on the way home. It climbed to 57 pretty well, right around 8,400RPM which is right were it was for the others around that speed. Then it acts like all of the sudden it found a gear that's never been there before and RPM dropped off to about 7500 and it started slowing. I let off and hit it again and it revved higher and again dropped down. I tried this a few times with the same result. I wish I had this extra gear on a lot of my scoots, but thi sthing hasn't got the power to accelerate from that speed at 7500RPM.


My first thought was simply to clean the CVT and hope that got me a good result. I pulled everything apart and cleaned it up. I forgot to get a closeup pic, but there was a black ring right in the center of the rear pulley on one side where the belt had been slipping at speed. I scraped that off with a razor and sanded the surface with 220 grit. I sanded the front pulleys with 400 grit to assure they were clean. Greased the torque driver, polished the drive boss, cleaned the CVT cover and case, and everything in that area.




The cleaning seemed to smooth some things out just a tad, but really accomplished very little. The belt slip in the center of the rear pulley made me think of 2 things, the contra spring and the torque driver. I started with the contra spring. It was noticeably less stiff than it would be new. I replaced it and the problem seemed to lessen, but was still there.


Next I went back to something I saw when cleaning. The torque driver has worn worse in one area and now the paths are no longer smooth.







For some reason, this torque driver has grooves a little longer than my other stockers. It's pretty much on par with one that I've modified. I really didn't want to replace the torque driver for fear that it would change everything with the CVT and really throw off any future results for this phase of the project, but I needed to try something.


I swapped in a modded torque driver and RPM steadily rises as speed increases as it normally would. Then the issue became getting the RPM correct. How smoothly the replacement pulley works and the stiffer spring are throwing off the RPM. I've been swapping in all sorts of springs new and used hoping one would put me right where I need to be with the same weights and everything to keep results on par with no luck. It seems like everything either wants to rev higher or lower than before now. Seems to top out slower than it was no matter what I do.


I've been out looking things over, trying to figure out why I've lost a couple MPH and my max speed RPM has increased when everything else is setup close to before the CVT issues. While I find it frustrating at times, at least I can say my desire to keep the project consistent and the results on the up and up has taught me something else I probably wouldn't have noticed otherwise.


I have 4 fixed halves of the rear pulley here, and I was swapping the modified torque driver onto each of them to see if there was a difference. I found that two of them would allow the pulley to open a little farther. They would open just far enough for the belt to reach as far as it can in the center of the pulley, for the best top speed gearing.


Initially, I thought the issue must be with placement of the receiver holes for the little pegs that the torque driver rides on. Perhaps some are drilled at a different height, I thought. I measured base to center and top to center and they all came out with pretty much the same measurements.





Then I noticed it appears some of the bases are recessed farther into the pulley face, which would create less clearance at full open. Again, I took some measurements. Two of the fixed halves have the base recessed in about 4.1mm. The other two are recessed only 3.6mm. That makes an additional 1/2 mm of clearance when the pulley is fully open, and it's just enough to allow the belt all the way to the center.





I also saw that both of the more desirable parts shared the same marking on the back of them, while the others had no markings.





Now it sounds like I've figured out how to get this back on track, but there's a problem. Both of the halves that I have, the original from P90 and the one I had around, have damaged threads. I've damaged 2 of these in my history with scooters, and I just had to pick the 2 that are the best performers. For some reason the nuts that go on these are very easy to start just a little bit crooked. If you are in a hurry or not paying complete attention and hit it with an impact gun, that's the end of the threads. I already knew this, but I suppose I needed to learn again when I was doing a bunch of spring swaps and getting frustrated over the last couple of days.