<html> Hi David,                 Are you noticing this on new parts?  Once static friction is broken, I have noted that this happens more frequently with worn or loose keybushings. e.g.  Changing a prehung set of Yamaha shanks, with out rebushing keys. At one time I used to change just the hammer sets at the University.  Now I insist on rebushing keys along with hammer replacement.  It makes a big difference. If the friction is high on the knuckle and centre pinning friction is very low it can give similar results. To trouble shoot.  Remove a hammer and wippen assembly,  and measure the friction of the flange C/P's, Holding the flange horizontal, if it drops under it's own weight, will give you a quick check, but start measuring and repinning if there seems to be problems. This eliminates these as the cause. Next examine the capstan, and wippen cloth,  polish the capstan, and if it's old capstans steam out the cup, dry iron smooth, and burnish in some teflon powder. FWIW. I have repinned more than one set of flanges on new parts.  With friction happening with the right amounts in the right places, the key should go down nice and and steady. Many new Korean pianos have way too much friction on the knuckles, shrinking the C/P bushings has been a common practice to over come friction.  It will mask the problem,  and give acceptable DW readings. Change the knuckles, and WOW, watch the the friction figures change. To to look at geometry issues you have to start with the magic line, key ratio, etc. The old adage, "divide and conquer", applies here.  Just eliminate one thing at a time. Regards Roger At 09:00 AM 8/24/01 -0700, you wrote: <blockquote type=cite cite>I notice, when I am doing a weigh-off on a key board, that on some actions as the hammer rises the acceleration is relatively slow and steady.  Other times, once the hammer starts to rise, the acceleration is quite rapid.   What is responsible for that difference?  Friction would seem to be the logical answer, but friction where?  I heavily lubricate actions before I do a weigh-off.  There seems to be something else at work in the geometry/leverage.  Among other things, with the rapidly accelerating hammer, it is more difficult to determine the measured downweight.  I there any way around that problem?  David Love </blockquote> </html>