The Honda PCX / Forza / ADV150 / SH Forums

No other component on a "scooter" determines it's personality and overall driveability more than the CVT. For those who are completely satisfied by the Forza's stock performance, read no further....there's nothing of interest to follow, but if you would like a faster 0-70 time and a quicker passing gear and are willing to accept a small penalty in fuel mileage, read on.

The Rollers

The total roller weight "governs" the engine, setting and limiting the actual rpm that can be reached during acceleration but there is another feature that is often overlooked....the roller diameter determines your "low" gear. By careful weight selection, you can affect the starting gear ratio, the upshift rpm, the downshift rpm, fuel consumption (both up and down!) and most importantly, your cruise rpm.

An important part of the reasons I bought a Forza was the CVT, more specifically the roller sizes used.....23x18.
This is the same size weights used on the earlier Reflex/Foresight/Big Ruckus MF04E engines and I have acquired a ton of information tuning variators for this weight size over the last 8 years.

What exactlly is meant by 23x18?





not exact, but close enough.......
The actual diameter of the rollers will vary with manufacturer and design intent.

All 23x18 Honda rollers are very uniform in diameter......

here's a 31 gram Reflex roller


here's a 26 gram Big Ruckus roller


here's a 21 gram Reflex roller


very uniform, all ~ 23.27mm

But the Forza rollers are sourced from another vendor....
they measure from 22.99 - 23.25, which is more variation in diameter than I've seen in the past with Honda rollers.


Aftermarket rollers are usually slightly smaller in diameter.....





And for the lowest possible starting gear, you have DR Pulley sliding rollers....





At 22.37mm, sliders start you off in a very low gear.

The pictures above show another feature of sliders and rollers.....
Honda rollers and DR Pulley sliders have "swappable cores".
They will interchange....if you look at the pics, you'll see there is a blind side and an open side to the roller or slider jacket.
Press one core out andd another in.

Most aftermarket rollers do NOT have removable cores.

But this weight range has another surprise, courtesy of the Honda Helix....


The older Helix used a slightly larger roller called 24x18, but it's really 23.80.....
that's 1.5 mm larger than the smallest choice, the DR Pulley sliders.
!.5mm may not sound like much, but remember the entire usable life of your drive belt is ~1mm of wear.

Many aftermarket variators come with tunable bosses....the spacer the variator rides on has been shortened a bit and it comes with 3 shims to let you set your starting gear ratio by selecting the shim.....they 1mm, 1.25mm and 1.5mm, again giving you ~1.5mm of adjustment on the fixed face side, while roller diameter selection gives you 1.5mm on the moving face side.



No other variator that I've worked with has the range of tuning options present in the plain old stock OEM Honda 23x18 variator. Although the Forza's variator is different from the MF04E version, it still uses the same basic design criteria....hold rpms under 6000 until the scoot is moving about 50mph.

Here's an example:
with easy acceleration, revs climb to 5400-5600 rpm and are flat out to ~50, then gradually increase.
with hard acceleration, revs climb to ~6000, then drop back under 5800rpm until 50mph and they gradually climb.
yours probably works the same way....

I call it the 5800 rpm limit or (governed limit) and it is not new....
all of Hondas 250s used the same design parameters.





The graphs show that for total weights over 120grams, acceleration revs are held under 5800 (I think that's the case with the Forza, but we see a little needle overshoot).

When I finally get the Veypor harness installed, I suspect the Forza will look very much like this BR running 126gram rollers...






The Veypor harness has arrived but I've decided to wait until I replace the rear tire, since I have to remove the swingarm to install the wheel sensor. The new harrness is more complicated with connections for EGT, ammeter, CHT, voltmeter, tach and speedo.




But I have been reducing roller weight in 3 gram steps every 100 miles or so.

This is the setup I'm currently running...





105 grams total weight, 3x20 gram DR Pulley slider and 3x15 gram Adige CarbonFibre rollers.
This combination will do 0-60 in <10 seconds and will accelerate from 55mph to 70mph in < 5 seconds.

To break 6000 rpm during launch, you need to go under 120 grams....
to stay over 6000 during acceleration, you need to go under 110 gram.

Mixing roller weights and types is useful for this sort of tuning.
Mixed weights are less stable and thats a good thing in this situation....
it takes longer for them to move out, giving good acceleration and the lighter rollers release quickly at rolloff, giving a strong forceful downshift.

The 3 heaviest rollers used will set your normal acceleration rpm, while total weight will determine your hard acceleration rpm....
what this means is that if you drive conservatively, you can still get good fuel mileage, even using much lighter rollers.

With 105g, on a long ride I'm getting about 66-67mpg, but here's a picture of a recent test result.....the gas station is exactly 10 miles from my house.
Staying on the speed limit.....



Thats 6 miles of 45 mph, 3.5 miles of 35 mph and 1/2 mile of 25 mph.


More later.

The Variator



The Forza 300's variator is different from the early 250 variators, but it still uses the same clutch as earlier 250s and the Helix. The main reason for this is the larger output shaft.
The Forza variator has longer and steeper ramps, with a noticeable cutout at the end. Honda learned this from scooter tuners who dremelled or machined the ends of the ramps to get a little taller topend.
Several aftermarket variators also use extended ramps.

One of the things I noticed on disassembling the Forza variator was how tight the thrust plate sliders were to the guides. I usually had to thump the boss to push the thrust plate out.



This is a real variator slider (as opposed to sliding rollers!)
This "slider" was called that years before DR Pulley invented sliding rollers and co-opted the name sliders. Now it's used for both. Just be sure you know to which slider someone is referring when discussing sliders.

The guides taper inward and the thrust plate gets tighter when the variator is open (rollers retracted). On mine, I removed the sliders and checked that all of them moved freely on each of the three guides.
On the rear of the sliders ther was a bright spot showing contact and I sanded the rear of each slider at that spot to allow it to slide deeper into it's seat.
I just polished eah one until the thrust plate moved smoothly from top to bottom without drag.
Normal wear will eventually accomplish the same thing, but it will take about 10,000 miles.
A set of variator sliders (3) cost about $10.



The above picture shows the variator assembled without rollers....
you can see the thrust plate sits right on top of the ramp rails.




The above picture shows the variator assembled with rollers....
this "gap" sets your low or starting gear and also maintains belt tension.
As a belt wears, you can correct for wear by increasing the roller diameter (if you choose).



The above picture is very important if you use DR Pulley sliding rollers.
Almost without exception, every scooter I've found them in had them installed wrong. They go in tall side up, as pictured.
If they lay stable in the tracks, they're wrong!
If they perpetually want to tip over forward, then you got it right!



The odd thing is, the scoot will accelerate like a bat out of Hell with them installed incorrectly....
usually, it will rev limit before it can make topend.



The above picture shows the configuration that I've found through testing to give best all around performance....3 sliders in the 19-21 gram range combined with 3 rollers in the 15-17 gram range.
With weights heavy enough to fully close the variator, 6 sliders will gain about 1000 rpm during the upshift process and 6 rollers will lose about 1000 rpm upshifting.

A combination of sliders and rollers give a smooth upshift with a flat rpm curve, something like this.....




more later....

Forza Noise Makers


The Forza makes some interesting sounds when you really open it up and let it roll.....
most of these come, either directly or indirectly, from the CVT.





The above picture shows the air intakes. The upper snorkel is the engine/FI air inlet and the lower opening shows the cooling fins on the variator used to push air through the CVT.
The engine air inlet is mostly a roaring rush of air at higher speeds, but the variator fan has a distinctly turbine-like sound that follows engine revs.





This is the air inlet cover....the large opening is engine intake and the small series of ports is CVT intake. These face forward, just behind the left side panel.



Here is another Forza noise maker....



This picture shows the CVT cover and the clutch housing outboard bearing.
Under the clutch nut is a bushing with 2 O-rings that fits into this bearing.
This bushing must be regreased at each removal....this is a clue to how it works.

The outboard bearing has two modes of operation....
at low rpm, it is an idler and the greased clutch bushing spins freely in the inner race of the bearing.
at high rpm, just as the variator fully closes and the rear pulley compresses the contra-spring, applying maximum force on the drive belt,
it becomes a loaded bearing and spins up to speed.

That's what we hear.....this transitional state as the bearing spins up.

With a stock Forza, this occurs at ~65mph through 70mph, at which point the bearing is loaded and turns in sync with the clutch.
if you listen closely, you can hear it disengage at a much lower rpm during gradual deceleration.

As you lighten the roller weights on the Forza, the Forza rumble moves to higher speed....
lighter rollers close the variator at higher rpm and the "rumble" moves higher with it.
105 grams moves it well above 70mph.

It works the other way too....

As a test I intalled 6 24 gram sliders to see what would happen if I forced ithe CVT to high gear sooner....
the Forza rumble started at 55mph through 60mph.

With 144 grams of weight, the Forza accelerated like a slug (0-60 in 19 seconds) but it got excellent fuel mileage, consistant mid to high -80s...for as long as I could stand it.

But with lightweight rollers you can move it beyond your normal driving speed....
maybe?


Final Drive

One thing to remember if you ever bump heads with a SH300i.....
he's geared lower than you!
If you want to race him, make it a mile, not a quarter of a mile.
He'll smoke you in the short run.

Before buying the Forza I had researched most of the specs from reading about the SH300 engine/cvt.
I knew the SH300 had an 8.5:1 final drive ratio from the literature.
The first time I removed the cover, I was going to confirm this and spun the clutch bell while watching the tire.....7.5+.
Thinking I counted wrong I did it again....nope, 7.5something?

So I checked the service manual and there it was....7.537.

so......
SH300i Final reduction 8.571

NSS300 Final reduction 7.537


The SH300 is geared for bottom end and the Forza is geared taller for the highway.
Use that to your advantage.


more later...

great write-up. I've always been perplexed by the inner workings of the CVT. I've previously owned a Reflex and now own a Forza and couldn't be happier.

As a bigger rider I'm plenty happy with acceleration and have more top end than I ever hope to use. Is there a suggested combination of rollers/weights that I could use to possibly trade off a little top end for some extra mpg's? My commutes are almost always between 55-60mph and I'm averaging a rock solid 63mpg after about 4k miles of riding.

Thanks. Seems weird to mix sliders and rollers. Have you ever seen a locked vario dyno of the Forza so we can know what the power curve vs rpm looks like?

Forza uses the same CVT as PCX... um
was it built for Forza and handed down or built for PCX and... opps
I thought Honda never over builds anything.

I just put Dr. Pulley 17gr rollers and gliders in a few days ago. Wow, what a difference, the acceleration from 0-100km/h is almost 5 seconds better (14.7sec. with original rollers / 9.81sec. with Dr. Pulley), also the torque increased a lot, 60-100km/h and 80-120km/h is also much better than before..... topspeed is a little bit better but the difference is with 5km/h more not that much but it goes there much quicker..... I really love it, the whole transmission got also smoother with the Pullleys. Fuel consumption does look like the same but i need a bit more kilometers to find out if there is a difference.

Very Interesting. I wonder what the actual top speeds are with the stock and various changes?

original topspeed was 135km/h to 140km/h..... now it does 140 even with head wind..... and it goes there very quick.... im really happy with the result

Carbon fibre weights. I've seen it all now

I will play with the CVT on my PCX when my Forza arrives! All of the parts are already sitting in a drawer. Great work. Especially like the graphs.

I installed Dr Pulley 17gm sliders and moving parts and the Forza pulls like a train upto 120 kph at about 7.5K revs at that point. Goes upto about 130-135 max slowly after that. I have rechecked the variator and all sliders installed correctly. I am thinking maybe mixing 17/18gm sliders (total 105gms) or 19gm sliders with 16gm rollers may give me the original top end (140) and keep the increased acceleration through the band that I have got from the 17gm sliders but with slightly reduced revs. Any experience with this mix or suggestions e.g different mix or just install 18gm sliders to get my top end back to 140 - dont really want to lose the new accelaration though - I like it a lot lol!

You're describing the same scenario I found with sliders....
they gain rpm too fast during acceleration and you run out of revs before you run out of speedo.

Don't misunderstand me....
I LIKE sliders (a lot!) but during testing and graphing different combinations, I discovered that they work best in combination with lightweight rollers.
I think every mid-sized scoot should be using sliders in a combination, but that's for them to discover some day.

My suggestion to you would be to increase the weight of three of the sliders to 19 or 20 grams while leaving three at 17 grams....
an even better suggestion....3 x 19G sliders and 3 x 16G rollers.

If mixing weights and styles:
1) always use sliders that are 4-5 Grams heavier than the rollers (the sliders control the shift)
2) in the 23x18 weight size, shoot for total weight in 110-120 gram range for best acceleration with highest top speed
3) you need at least 93 grams to fully close the variator before 9500rpm (this observed on the Reflex 250 with it's dual-tracked variator)
4) in your mix, the 3 heaviest weights will set your easy acceleration profile while TOTAL weight will set your hard (WFO) acceleration rpm
5) the reason for carbon fibre jackets on the lightweight rollers is less drag (their roller jacket is much slicker, so it moves out faster with less weight during acceleration and it releases and retracts much faster during roll-off than standard plastic jackets, allowing a positive downshift on demand)

Bottom line...
the carbon fibre roller's slick jacket makes them act heavier during acceleration and act lighter during deceleration...
and of course they look so cool (whenever you have the variator open??)

And yes I realize this is unconventional advise, but it will only cost you the price of rollers and sliders and of course your time to test.
I'm currently running 3x20g sliders and 3x16G roller which yields 0-70mph in <12 seconds with extended highway mileage of 66mpg....
it will also top and hold 80mph in much less road than required in OEM stock configuration.

The stock 21G rollers are OK for excellent fuel mileage but they are a little too heavy if you're a roadrunner.

BRed wrote:You're describing the same scenario I found with sliders....
they gain rpm too fast during acceleration and you run out of revs before you run out of speedo.

This statement makes no sense. Unless you have installed the sliders the wrong way around. Or chosen a slider that is so light that it can't push the belt all the way out. Sliders always add gearing to both ends of the extreme. The short gear at take off is shorter which gives better jump off of the line. And then above the lock out speed, the sliders have traveled farther out the ramps than a roller can so they make the gearing longer. rpm versus speed above lock out is lower with sliders. And top speed goes up if you are still able to hit the rev limiter.
.
http://hondapcx.org/viewtopic.php?f=9&t ... &start=100
.

Sliders definitely do NOT add to both ends of the gearing range.

Sliders always gear the variator DOWN because of their lower profile.....
when they reach the end of the ramps, they always produce a lower top gear than can be achieved with rollers.
Every test I've done has confirmed this, in spite of whatever the advertising hype might say.

Anyway, that has been my experience with rollers and sliders in the 23x18 size range....
as I said earlier, with total weights capable of fully closing the variator, sliders gain ~1000rpm in the upshift.
Rollers lose about the same amount.

DR Pullley makes a variator with ramps designed specifically for sliders to do what you're suggesting, but I don't think they make a Forza boss yet.

The only variator that I've used that really lowered the low gear and raised the high gear was the J.Costa variator, but it uses a transverse pin system instead of rollers and ramps.

you should be aware that not all sliders (or rollers, for that matter) are created equal.....
in each size range, there are several different diameters as well as weights of both sliders and rollers.

BRed wrote:Sliders definitely do NOT add to both ends of the gearing range.

Sliders always gear the variator DOWN because of their lower profile.....
when they reach the end of the ramps, they always produce a lower top gear than can be achieved with rollers.
Every test I've done has confirmed this, in spite of whatever the advertising hype might say.

.

Wow! How could you be so wrong! Sliders always allow a higher gearing due to the control edge that stops their outward travel being so much smaller. If the front pulley haves are not mechanically restricted. Nor the rear pulley is not bottoming the belt on the boss. I don't have a Forza but many people with a PCX have reported gains on both ends of the range when switching to sliders. They move out the ramps farther. So they push the belt out the front pulley farther. This is the beauty of a very elegant design solution. Every surface dimension is optimized to do the job that is need. Not just a simple round profile as a roller is. It allows a further travel out the ramps to gain a longer final gearing. And, starts with a lower gearing from the slightly smaller thickness in that dimension. There are many owners on this forum alone that report the extension of the gear range on both ends when using sliders instead of rollers.

sendler2112 wrote:There are many owners on this forum alone that report the extension of the gear range on both ends when using sliders instead of rollers.

Which begs the question, how do they record and measure these extensions?

I will confess that I have no experience with the PCX variator, but I've tuned a few dozen Honda 249cc and more than a few Kymco 250 variators, all using 23x18 weights and the results from the Veypor tuning computer and it's graphs were fairly consistent.

I'm not disputing your statement but I would be interested to see any hard data that has been collected that can show that sliders actually produced a taller final gearing than rollers, even in a PCX.

I have to doubt it when it comes to the Honda 249cc or 279cc engines, in view of the fact that some 23x18 rollers are 1.5mm larger than sliders....
you can visually see they're taller at any point along the ramp.

I think we may be comparing apples and oranges?
Or maybe oranges and grapefruit?

Either way, I will qualify my posts by saying the information in this thread is intended ONLY for the NSS300 Forza....not for the PCX.
said information is my personal opinion based on my personal experience and is intended for educational purposes only.

and I'll eventually post some graphs of the Forza to validate some of this.....once the temps get above 0 again.

gn2 wrote:

sendler2112 wrote:There are many owners on this forum alone that report the extension of the gear range on both ends when using sliders instead of rollers.

Which begs the question, how do they record and measure these extensions?

Many videos showing an increase in top speed at reline on the stand and higher top speed on the road. Very easy to find here and all over the web if you look.

BRed wrote:I have to doubt it when it comes to the Honda 249cc or 279cc engines, in view of the fact that some 23x18 rollers are 1.5mm larger than sliders....
you can visually see they're taller at any point along the ramp.

Of course there are oversized rollers that will also increase gearing. But at the expense of the low gear at launch.