Trials and Errors:

How much can weighoff help grand action playability? Sometimes, as with poor action geometry, weighoff will not help unless problems are corrected first. 

I recently found myself so challenged by a vintage Baldwin L. The hammers, shanks, and whippens had been replaced and were in good shape. But the action did not play well. My customer wanted to know what could be done for her attractive instrument that played like a truck and no longer held its tuning. I recommended the fairly expensive route of new belly, block, and strings, plus deep regulation, which she agreed to. After Jude Reveley did the belly work, the piano came to me, ready for action fitting and regulation.

When it arrived, two keys were jammed on top of their front key pins, despite the shipping rail being correctly in place. Not enough key pin in the front mortises was the problem, relating to the replaced backrail cloth not being quite thick enough. I solved this by slicing open the rear glue joints, adding medium action cloth, and gluing the backrail cloth back down.

Next, the capstans were too far under the whippen heels – no wonder the action felt so heavy. The cast-iron topstack had detachable hammer and whippen rails, so I decided to move the whippen rail back. When it was out, I discovered that its attachment holes had already been ground quite a long way but in the wrong direction. Perhaps, the previous technician only realized the error when it had already become too costly and returned the rail to its original position, accepting the shallow dip, heavy weight, and compromised playability.

I also noticed that the repetition levers had been cut shorter at the dropscrew end, prompting me to measure spread, which was a little under 109 mm. My memory was that these modern Renner parts with 17 mm shanks and straight-heeled whippens (with the original Baldwin flanges swapped in) should have had a spread of 112 mm. So, I ground the slots in the other direction until I achieved this and reassembled. The weight was much better, but the capstans were now too far forward (toward the player) under the heels.

So, I decided to move the capstans. And knowing that a little goes a long way, I  converted from angled to straight capstans. This meant I could avoid the drill bit being influenced by filler material in the old hole, while only moving the capstan-to-heel connection the little ways needed.

I made two mistakes here, based on assumptions I thought were correct. One, that weight felt at the front of the key was based on where capstan and heel met, as opposed to where capstan entered keystick (and where weight of sub-keystick capstan lay). The other was that 112 mm would be the number to resolve all problems, given these parts.

When the capstans had been removed, holes filled, new holes drilled, capstans reinserted, and regulation roughed back in to real hammerline, letoff, and drop, I discovered the weight improvements I’d achieved with spread were largely undone by the capstan move… Ouch.

Meanwhile, there was another factor applying a constraint. After I had filed the hammers, the top several notes sounded like wood, over-striking to a point on the strings too close to the Capo. One of the odd features of this piano was the use of triangular bars (point side up) instead of keyframe guide pins, with no cheekblock adjustment in or out, and dags at the back that touched the keyframe when in cheekblock-locked position, i.e., there was no moving the keyframe in and out without surgery to one or more elements.

The hammer rail could be ground with much effort to go backward but it already over-encroached on where the whippens wanted to be. Or the whole topstack could be moved with even more effort but that would worsen the capstan/weight issues. I was between a rock and a hard place. Jude suggested raising the hammer centers, thus reducing the over-striking by shortening the centers-to-strike distance.

I tried shimming the hammer rail up but ran into dropscrew-to-stretcher interference. I then tried shimming the front topstack feet, which worked better and did restore tone to the top register.

The job was then interrupted for a month, refreshing my energy for dealing with the situation. Okay. The capstans had to go the other way. After several rounds of sampling different variations, I came up with a solution that lightened things the most, while still appearing to be in reasonable dip territory.

Out came the capstans, in went the filler, holes were drilled (angled this time) to the new line and capstans installed. With everything back together and roughed in, it was clear that I’d been fooled again. The new weight was expected to be little lighter and when things were apart, I had removed one or two leads from each key in anticipation. But the weight was very light, the hammers rose nowhere near enough, and the dip needed to be deeper than it could get.

I was now in the wrong place in the opposite direction. Neglecting to respect my first decision to avoid filled holes, I had overlapped them, and they had straightened up my newly-installed capstans enough to totally throw the geometry off again. There was nothing for it but to move the capstans again.

Chris Robinson had a great teaching technique. Regulate an item too far in one direction and see what happens. Then, regulate it too far in the other direction and see what happens. Now, you better know where it should go and why. This, of course, is just what I had done for myself. The best place turned out to be a very slight move forward instead of backward.

With a more meticulous sampling, and with woodfill now in both directions, the capstan drilling came out right. I did have to contract the spread to 111 mm, but now dip was .400”, hammer blow was 1.75”, and aftertouch was just right. The dropscrews had just enough stretcher clearance. The repetition levers had just enough hammer rail clearance. The angled capstans met the whippen heels just forward of center, where the curve of whippen cloth was like that at the center of angled heels. Note 88 was in the right spot tonally. The weighoff worked with 50g downweight and 20g upweight in the bass. Lead was slightly reduced. Hammers made vertical and traveled to vertical all cleared each other nicely. And with strings fit to hammers, the piano sounded great.

So, it worked. Just. And key to finding that very precise spot were the trials and errors. Perhaps, though, with less assuming and better sampling, I could have found it in one round instead of three.

And then, my customer decided to add case refinishing to the project. What if I had been too timid or too impatient, held my nose, and sent back a slightly rotten action job with the fancy belly work and refinishing? The errors were blessings in their correcting my wrong assumptions and in their calls to get it right. And the trials were gratifying as they found me innocent of failures to try and rewarded me with offering my customer what was deserved from the beginning, the job done.

I had a couple of moments of self-empowerment for my trouble, as well, drilling clearance holes through the cast-iron for easy access to topstack screws and working out accurate bedding-to-bench transference, in spite of the keyframe having no glidebolts: gifts to a successor, should the piano live on to a next deep regulation.

  • Part 1 "Why Weigh Off?" 
  • Part 2 "My Strategy" 
  • Part 3 "Method and Style"
  • Part 3a "Regulation Strategy Review" 
  • Part 4 "Pragmatics"
  • Part 6 "Levers"
  • Part 7 "Playing with Proportions"
  • Part 8 "Weight, Friction, Inertia, and Distance" 
  • Part 9 "Plan by Number?"
  • Part 10 "Best Balance for Best Art"

    Christopher Brown

    It may or may not add more friction, depending on where the capstan-to-heel connection happens relative to the line between wippen center and key fulcrum. I think that difference is fairly small, in any case, because both surfaces are convex (and in both cases one will slightly embed in the other…). The significant difference from my Baldwin L experience is that more of the slanted capstan is closer to the key’s fulcrum making all but the connecting surface effectively a lighter weight traveling a shorter distance.

    Christopher Brown
    David Love

    Why do you angle the capstan? Doesn’t that just cause more scrubbing between the capstan and the wippen heal and thereby add friction? I generally install capstans at 90 degrees to the key.

    David Love

    Leave a comment

    All comments are moderated before being published