Modern Vespa

Has anyone tried different rollers in an HPE?

Motovista has asserted that lighter rollers have been 'proven' in an HPE yet I haven't seen a thread here about them?

Curious for example about the malossi 13g rollers (6613697.M0). what rpm do these give when *accelerating* at say 40-60kph? stock is around 6250 at that point. and what rpm at 100kph (speedo), stock is around 7250. taking off at the lights very quickly hits 6250 then has a slight dip to 5900 around 28kph and then rises to 7000 as it heads up to 100. and of course crusing rpm are lower than these numbers and would also be interesting

would love someone with a supertech (or 75th gts or gts with the pmp2 optional module) to use my app and capture some info ... I like hard data, not seat of the pants. (I currently have 17,500km of recorded data for looking at these numbers with stock)

ref: https://modernvespa.com/forum/post2542772#2542772

PM me your address and I'll send you a couple things to play with.

When you get that data, I hope that you will share it with all of us.

Of course.

Do you have cruising rpms for different speeds? Using them and physics I could tell you what I think.

If you have a spare variator lying around you could measure the max radius of the grooves the weights travel in. That would improve accuracy but even so I have a good starting point.

Given we already have 6250/7000 for rpm during accel.

I ordered the AF1 recommended Malossi lighter rollers for my Liberty. Really woke it up.

an update: thanks to Motovista I'll be testing the malossi 12g and 13g rollers soon in my supertech. just trying to make up mind about replacing the belt which is currently 8000km at the same time which would distract my testing as it would be stiffer when put in and need 'breaking in' ...

You're overthinking it. The new belt will be at factory specs. If your old belt has worn and is narrower than it should be, it will affect the performance of the scooter, but not to a degree you can probably notice.

per moto. the new belt won't be any variance.

knock those rollers in and report back.

I fully expect a detailed report!

I always overthink everything in life.

I was really just taking the opportunity to think as I wish to minimise variables when testing and also I had to break in a new set of tires (changed 1 week ago) and also wait for good weather to get a benchmark before and after

the reports will be detailed and have pretty graphs - ya might have realised by now that I like concrete numbers not somebody's feelings or half arsed measurements with a stopwatch which are all too often bollocks. so today i'm thinking about revisions to my graphing code (filters and smoothing etc) to allow clearer comparisons

12g malossi rollers installed. And also new oem belt (it was cracking and had done 8400km) and the sliders cause they had done 31,000km.

dont ask for thoughts yet as I've only done 20km of errands so far. of course I'll report back when I have solid numbers and thoughts.

8400 klm for a belt doesn't sound a lot. I thought you were meant to get at least 10000 klm before they needed changing?

Would it have survived to 10,000? Sure. But since it definitely wasn't fresh looking and i had a new one handy, thought i might as well

early observations with 12g:
(not going to give ya graphs or numbers yet as I want to do at least one day trip in the country first).

stable speed: not noticing a difference, but haven't really studied this yet. have been paying more attention to rpm during acceleration

low-mid acceleration: not radically different (seems to be a small increase in rpm)

mid-high speed acceleration: it revs plenty higher. eg when cruising a bit below the speed limit on the freeway and then give it a kick like when changing lanes. have yet to start crunching these numbers further

point at which the vario maxs out and the speed v rpm increase linearly together: this is at a distinctly higher speed and rpm.

near speed/rpm limits: no idea yet.


that change to revs during acceleration at mid-high speed is *very* evident on my graphs. stay tuned (pun intended )

640km with the 12g -> not good. will give detailed write up later (might actually wait until I've also got decent data for the 13g)

just installed the 13g today -> feeling better than the 12g already but not enough data to really say anything concrete let alone compare to the 14g

No tacho fitted yet then? Please define plenty.

yes I have tacho in my vespa snoop app.

plenty is 200rpm at low speed and 600rpm at high speed. as I said, data will be presented later (it is recorded)

(and ya can really hear that 600 or so change at the higher speeds)


Edit: the aforementioned 640km is recorded at a rate of 10 times a second resulting in 502,434 db rows of time,speedo,rpm,temp,volts,etc and matching gps time,lat,lon,alt data (less frequent) allowing detecting of incline which is used in filter / colouring the graphs that will be presented. So yes, I have adequate data.

I've seen the info from your app. Looks like a sky at night.
200 or 600rpm greater than what?

I suggest doing your own testing and analysis

Done it on my scooter using a tachometer. Between 500 and 1700 rpm increase during acceleration depending on weights. 5500 to 6500/7200. Rider weight, ie with a pillion, and hilly terrain also increases the acceleration rpm as one would expect from the physics of a CVT

But I had a target acceleration rpm based on rpm at max torque. Which was 1000 rpm higher than stock. So I am almost spot on with that.

200 to 450 less rpm at cruising speed (50-65)mph) as I use an aftermarket variator. But at anything above about 45mph I find the wind noise attenuates any increase in noise from the engine. Which is about +300 rpm at 45 mph.

Stall rpm increased from 3150 to 5300.

So if you know your acceleration rpm stock you can ascertain the weights needed to get, in your case, 5000-ish rpm during acceleration. If you already get that then you don't really need to do anything.

had a proper long ride with the 13g a couple of days ago and well, way better than the 12g but not convinced that is better than the 14g. early analysis says small bit slower acceleration than 14 but faster than 12.

my hunch as to why these are giving me worse than 14g is that although the higher revs are nearer to the peak power (listed in the manual as 8250) it's now further from the peak torque (5250) and the loss of this torque is hurting more than the increase in power is helping. Or it's belt slipping or inappropriate variator spring or something like that

note: I do want to try more careful throttle control insterad of wide open just to see what happens. eg try around 80%

please remember this is on a near stock HPE (as far as matters) not a comment about any other engine.

during the week I switch to new oem rollers. have done 100km so far (shopping including freeway to the suburbs for other bike parts).

The early numbers are showing a measurable difference* versus the oem 14g that I had on before this sequence of changes. Don't forget that I changed the belt at the start as well - seems to have measurable impact. Will have to check some older data from previous belt change to see how much that changed. Hmm.

Mind you, already getting 0-40 and 0-60 times that match or better the 13g and 12g


(Yes those promised graphs and final numbers are still coming)

* more data might reduce this difference

incase ya wondering where my promised graphs are:

I changed my graphing to not use the speedo (the one shown on the supertech dash) but instead to use the rear wheel value from the abs which very closely matches the gps (doesn't have the approx 8% exageration). Why is this a good change? cause there is less lag in the number. the speedo value lags slightly compared to the value from the abs plus the speedo value is rounded to a whole integer and has a bit of smoothing to it.

also want to do a little more testing and country riding now that I'm back on the stock 14g but the weather has been crap.

Pretty sure The scooter guru Robot said Primavera and Sprint 150 lighter rollers make a big difference but they tried different rollers on the newer GTS and concluded it is best to stick with stock rollers.

Part 1 of 2

(expect me to edit and edit and edit this post)

some graphs (I'll try to get off my arse and write explanations of what each is showing but in the mean time you can try and guess

what is Rear Wheel? It's the kph value from the ABS unit, not the ecu or dash. It correlates very closely with GPS. (unlike the dash which has approx 9% exaggeration)

Data capture and graphs: My Vespa Snoop android app

14 6months ago: 30/dec/21-13/jan/22 800km summertime. Newish tires and belt. Blue line.

14 pre test: 19/may/22-21/jun/22 1031km winter. Oldish belt, New tires. Blue line.

12g: 22/jun/22-3/jul/22 642km winter. New belt. Green line.

13g: 4/jul/22-19/jul/22 812km winter. Red line.

14g post test: 20/jul/22-14/aug/22 1027km winter. Blue line.


My scoot (some of the following wont matter but listed anyway):
Oct/2019 Vespa Supertech EURO3/APAC (yes euro3, if you need to argue then pls start a new thread)
33,500km
Arrow Exhaust with cat
diablo rosso tires at approx 27/29 psi
small oem fly screen (gotta affect areo)
oem top box (gotta affect areo)
sip suspension (the rear will affect how much the bike hunches down when accelerating from zero)
malossi clutch + bell
malossi white spring and that yellow cap (forget the names for these)
I'm 67kg / 175cm.
lots and lots of thrashing and country touring


I had a dysfunctional throttle body all through autumn. it died directly after 14 6months period and I installed the replacement directly before 14 pretest

Part 2 of 2

(expect me to edit and edit and edit this post)

What the graphs show
lighter weights are notably higher rpm during acceleration and minimal change during stable. (12g highest, 13g only sligher higher than 14g)

speed v rpm is the same from approx 100kph upwards.

max speed / rpm same.


Accel Times (not yet shown)
not better, basically within margin of error / possibly slightly worse with lighter weights.



Suspicions/Thoughts

I like large sample sets as there are so many variables

maybe the more power (peak kw/hp is documented as being at 8250) didn't outweigh the affect of loosing torque (peak is documented as being 5250)

the vario (front pulley on the half with the rollers) did blacken up some what when moving to 12g. Was this slippage or grime (not unusual due to my love of unsealed roads) or some residue from the new belt?

I've never come up with a good explanation of why the stable speeds have that break in the line.

Robot said in a video that there's no point in changing the rollers. I agree.

Maybe if I also changed the big spring in the rear pulley (forget the name) might help balance the lighter weights?

That's to be totally expected, as you're now in top gear with rollers fully extended. As long as the rollers are the same dimensions (apart from weight) there should be no change to the top speed. That's where the Dr Pulley sliders score, giving good acceleration and a slightly increased top speed. Only 1-2mph (2-3kph) though.