The guys at Scooterworks.com had an NCY drivetrain upgrade kit for the PCX which allowed them to determine that the contra spring has the same dimensions as all of the standard GY6 125cc springs. They stock the "1000 rpm", 1500, and 2000 rpm springs. I tried the 1500 for $18 shipped and installed it with the stock weights which will take about 20 minutes next time. It is really stiff! I'm glad I made a rough measurement of the comparative spring rates before putting it in. You can measure it with a digital scale. Choose a socket with an extension in it that is about 70% of the free length of the spring and place it inside the spring on a scale. Tare the scale to zero and then press down with your hand until it is just skimming the top of the extension which is being used as a travel gauge. I found 6.5kg for the stock spring vs 10 kg for the new one. The aftermarket spring is a whopping 35% stiffer. Takegawa lists their spring at 11%. I wonder what it really measures. I put the NCY in anyway to see what it would do. My PCX is now much more responsive. It now immediately down shifts, or revs up, whichever way you want to look at it, when increasing the throttle. All of the rpm ranges were shifted upwards. Even mild coast down. Light cruise at 30 mph is running 5,000 rpm now and it immediately jumps to 6,400 on accel. Maybe this is the 1,500 rpm the rating is referring to. Once past 30, the pulling rpm continues to rise up to 7,400 whereas it used to be 6,600. I may get a chance to do some timed runs later today but it feels much quicker. There is still some action at 50 mph. The stock drive train has the weights locked all the way out by 50 mph with almost no kick down available. The new spring still allows some down shift from 7,000 to 7,400 at that speed. Pulling speeds of 8,000 are said to be at the power peak so I can still go with slightly lower mass in the weights but the 12gm sliders I have are probably too low now with this spring. There is a big loss of top speed though so I am going to try the next lighter spring and then play with the weights and may even spend some more money to see if there is any difference with a kevlar belt. I'm not sure if the loss comes from the rear torque pulley being still too active over 60 and pulling the weights off of the stops, or if the added spring force is squashing and stretching the belt slightly causing it to ride lower in the front pulley even though the weights are all the way out. Top speed now varies from 62-65 mph at the rev limiter depending on the wind. I used to get 67 every day easily. The good news is that there is quite a bit of untouched metal on the face of both pulleys. Once I settle on how much spring and mass to use, it occurs to me that there may be another big advantage to the sliders. It might be possible to tune a set for any given combination of parts on a bike by looking at the remaining virgin metal. We might be able to remove some of the material on the part of the slider that is in contact with the variator at the outer stop little by little until full belt travel is achieved.
.
.
.
.
.
.
The last image shows the famous torque grooves. Honda is using a progressive profile. Kick down should become less active at high speeds.
.
.
.