Yesterday I took my test gun apart again and removed another O-ring from the buffer, so it is now just two #113s of 70 Durometer.... In order to allow the hammer to engage the buffer, I simply wound the setscrew I am using for a striker out 3 turns (0.108") to allow for the missing O-ring.... Note that as you back the striker out CCW below flush, you increase the hammer throw (to where it first contacts the stem) beyond stock but reduce the preload by the same amount.... At 3 turns below flush, the hammer stroke is now 0.69" and the preload is effectively at 3 turns out, even though the spring remains at coil bind when cocked.... After a few shots to determine where the buffer started to engage the hammer face (2 turns below flush) I shot some strings at maxmum preload.... and then at reduced preload on the last string to move the curve back below 2000 psi....
If you look at the top three curves, you get a perfect picture of what happens as the buffer progressively engages the face of the hammer.... The blue line, with the striker recessed 2 turns into the hammer face, shows the buffer not engaged at all.... The red line, at 2.5 turns recessed, show the buffer starting to engage on the 4th shot in the string, and surpressing the velocity in the second half, shortening the string by 2 shots.... The solid green line, at 3 turns below flush, shows the buffer with increasing engagement, and the entire first half of the shot string is missing, it would require filling to over 2000 psi to see it.... or reducing the hammer spring preload to move the curve to the right.... That is what is taking place with the dotted green line, which is still 3 turns below flush on the striker, but has the preload backed out 5 turns to create an entire shot string without filling past 2000 psi....
As it turns out, I now have three shot strings all at around 750 fps with the preload optimized, but with buffers of 2, 3, and 4 O-rings of 70D.... When I look at the data, I find the following:
Four 70D: 16 shots @ 758 fps (18.2 FPE) @ 1.24 FPE/CI (equals 32 shots in full size Disco tube of 135 cc)
Three 70D: 17 shots @ 748 fps (17.8 FPE) @ 1.29 FPE/CI (equals 34 shots)
Two 70D: 15 shots @ 754 fps (18.1 FPE) @ 1.40 FPE/CI (equals 30 shots)
It is dangerous to drawn any firm conclusions from such a statistically small sample, particularly when the average fps and FPE are not identical.... For example, the small decrease in velocity from 758 fps down to 748 fps could account for part of the increase in efficiency from 1.24 to 1.29 FPE/CI, and maybe for the extra shot as well.... However, there seems to be a slight trend towards increased efficiency with the shorter, and therefore stiffer buffer, but maybe at the cost of a lost shot or two.... Statistically speaking, they are pretty much a wash, so you would be hard pressed to make a choice between them.... When you compare the above with what I got by detuning a stock Disco with just an RVA to about the same power, there is again nothing to choose from.... I got 33 shots at 18.6 FPE within 4% ES with an efficiency of 1.36 FPE/CI from 1800 psi down to 1000....
Bob