I installed the 20g sliders in the IceBurg this weekend with expected results.
I did weigh both the DPS and OEM weights. The OEM's were, in fact, 19g and the DPS were, oddly enough, 20g.
Indicated mph @ RPM
MPH --- rpm before --- rpm after
40 ~5K ~4.5K
50 5.2K 4.5K
60 5.8K 5K
70 6.8K 6K
80 7.8K 6.5K
So I am seeing the rpm reduction that is common to this modification. I also noticed a definite improvement in mpg but couldn't really quantify until this morning's commute. My typical mpg for my 33 mile morning ride is right around 62.2 (extrema +0.5, -1.2 mpg). This morning it was 65.25. My average morning ride varies between 1-2 mile stretches of pretty constant drafting @ 50-60 mph mixed with about 20 lights (I generally have to stop for 1/3 to 1/2 of these). If I leave early enough to catch all (or most of) the lights green then there is less traffic so I get less benefit from drafting so it tends to balance out.
As for performance, I admit to buyer's remorse after I purchased the 20g sliders. I wanted 19g but they would have been special order from my local shop (I have, recently, become slightly anal about supporting local businesses) and I didn't want to wait. I selected against the 18g product since my overall goal was rpm reduction at 75-80 mph and I am pretty satisfied with where I ended up in that regard. I probably would have been satisfied with the 18g product as well. I do not notice any difference between the OEM rollers and the 20g sliders. Accleration is probably indistiguishable unless you instrumented for it. My bike has always had a tendency to shudder for the first 5 minutes from cold. I have to be a little light on acceleration to minimize it. After 5 minutes, I can lean on it from a standing start and never experience any shudder at all. This is slightly better with the DPS but I really believe it was the cleaning of the variator and associated components that made the difference. There was a lot of dust mucking up the variator assembly.
In peforming the actual installation, I went with a variation of the tool mike1nw posted recently. Mine was made completely of 9/16 inch plywood cut in the shape of his (I didn't have any metal barstock around and didn't think it warranted purchasing). I did go with the 6, equally spaced, bolts. I covered the bolt threads with vinyl tubing (3/8 ID) and it was a simple slip fit. In regard to the plywood outline, you must have the offset arm (my tool was 8 inches in diameter with a 3 inch wide arm) so you clear the clutch bell with the tool in place. I started with a piece of scrap plywood I had that was 12 inches x 24 inches. I drew the 8 inch circle (7 inch would have been better for clearance) in the bottom left corner of the plywood. Then I drew a line 3 inches from the bottom and a 45 degree gusset tangent to the circle. Cut this out with a handheld saber saw. I drilled the center hole with a 1" spade bit (start from the top, finish from the bottom) and quickly discovered this was too small so I enlarged it with the saw. You could get by with a 1 1/2 inch center hole. Mark the bolt locations as described in mike1nw's thread. 6 bolts is REALLY overkill. I was rather surprised at how easily the shaft nut came off. The vinyl tubing on the bolts takes care of a certain amount of imprecision in the bolt location and makes the tool somewhat sticky in that it tends to stay in position, once placed. Took about an hour to cut and assemble the tool (I work slow and SWMBO is a stickler for cleanup). 3 bolts is really enough to distribute the force and, after having done this once, 1/4 inch plywood would provide ample strength. The bolts I used were 3/8 x 2" with appropriate nuts and 1.25 inch washers on both sides. I'm going to drill acouple of 5/16 holes in the arm to hold the 8x1.25 50mm bolts so they don't get lost.
On a side note, a 15/16ths socket is virtually identical to a 24mm socket. I could've saved $6 using the one I had but did not want to make a trip in the middle so I went ahead and purchased the 24mm.
The entire job, including building the tool, took right around 4 hours from center stand down to center stand up.
What I built:
I did weigh both the DPS and OEM weights. The OEM's were, in fact, 19g and the DPS were, oddly enough, 20g.
Indicated mph @ RPM
MPH --- rpm before --- rpm after
40 ~5K ~4.5K
50 5.2K 4.5K
60 5.8K 5K
70 6.8K 6K
80 7.8K 6.5K
So I am seeing the rpm reduction that is common to this modification. I also noticed a definite improvement in mpg but couldn't really quantify until this morning's commute. My typical mpg for my 33 mile morning ride is right around 62.2 (extrema +0.5, -1.2 mpg). This morning it was 65.25. My average morning ride varies between 1-2 mile stretches of pretty constant drafting @ 50-60 mph mixed with about 20 lights (I generally have to stop for 1/3 to 1/2 of these). If I leave early enough to catch all (or most of) the lights green then there is less traffic so I get less benefit from drafting so it tends to balance out.
As for performance, I admit to buyer's remorse after I purchased the 20g sliders. I wanted 19g but they would have been special order from my local shop (I have, recently, become slightly anal about supporting local businesses) and I didn't want to wait. I selected against the 18g product since my overall goal was rpm reduction at 75-80 mph and I am pretty satisfied with where I ended up in that regard. I probably would have been satisfied with the 18g product as well. I do not notice any difference between the OEM rollers and the 20g sliders. Accleration is probably indistiguishable unless you instrumented for it. My bike has always had a tendency to shudder for the first 5 minutes from cold. I have to be a little light on acceleration to minimize it. After 5 minutes, I can lean on it from a standing start and never experience any shudder at all. This is slightly better with the DPS but I really believe it was the cleaning of the variator and associated components that made the difference. There was a lot of dust mucking up the variator assembly.
In peforming the actual installation, I went with a variation of the tool mike1nw posted recently. Mine was made completely of 9/16 inch plywood cut in the shape of his (I didn't have any metal barstock around and didn't think it warranted purchasing). I did go with the 6, equally spaced, bolts. I covered the bolt threads with vinyl tubing (3/8 ID) and it was a simple slip fit. In regard to the plywood outline, you must have the offset arm (my tool was 8 inches in diameter with a 3 inch wide arm) so you clear the clutch bell with the tool in place. I started with a piece of scrap plywood I had that was 12 inches x 24 inches. I drew the 8 inch circle (7 inch would have been better for clearance) in the bottom left corner of the plywood. Then I drew a line 3 inches from the bottom and a 45 degree gusset tangent to the circle. Cut this out with a handheld saber saw. I drilled the center hole with a 1" spade bit (start from the top, finish from the bottom) and quickly discovered this was too small so I enlarged it with the saw. You could get by with a 1 1/2 inch center hole. Mark the bolt locations as described in mike1nw's thread. 6 bolts is REALLY overkill. I was rather surprised at how easily the shaft nut came off. The vinyl tubing on the bolts takes care of a certain amount of imprecision in the bolt location and makes the tool somewhat sticky in that it tends to stay in position, once placed. Took about an hour to cut and assemble the tool (I work slow and SWMBO is a stickler for cleanup). 3 bolts is really enough to distribute the force and, after having done this once, 1/4 inch plywood would provide ample strength. The bolts I used were 3/8 x 2" with appropriate nuts and 1.25 inch washers on both sides. I'm going to drill acouple of 5/16 holes in the arm to hold the 8x1.25 50mm bolts so they don't get lost.
On a side note, a 15/16ths socket is virtually identical to a 24mm socket. I could've saved $6 using the one I had but did not want to make a trip in the middle so I went ahead and purchased the 24mm.
The entire job, including building the tool, took right around 4 hours from center stand down to center stand up.
What I built: