Piano World Home Page Forums Our Most Popular Forums Piano Tuner-Technicians Forum Steinway action lowering

Ed, do you always remove the keys before fitting and shimming the stack to the keyframe? [/quote]

I remove the top action and keys, raise all glide buttons, bed the back and front rails, lower the glides until they just barely tap (slightly more so in the bass and none at the very top), and then place the top action on the frame. Any of the feet that don't touch the keyframe get shimmed at this position. With the weight of the keys back on them, the balance rail will drop the few thousandths onto the glides and the front and back rails will be pressed down by the spring while the action is stress-free on the frame.

After that, it is important to install the cheek blocks, press the una corda and recheck all the balance rail contact, and THEN, firmly press both damper pedal and una corda and check again. I do this because it is not uncommon for the use of both pedals together to flex the keybed enough to create a knock under one or more of the glides. If so, I turn the glides down until they don't knock, let go of the pedals, and check again.

I think this to be the most solid and durable approach, and has, in my experience, yielded the best results .

Regards,

I remove the top action and keys, raise all glide buttons, bed the back and front rails, lower the glides until they just barely tap (slightly more so in the bass and none at the very top), and then place the top action on the frame. Any of the feet that don't touch the keyframe get shimmed at this position. With the weight of the keys back on them, the balance rail will drop the few thousandths onto the glides and the front and back rails will be pressed down by the spring while the action is stress-free on the frame.

After that, it is important to install the cheek blocks, press the una corda and recheck all the balance rail contact, and THEN, firmly press both damper pedal and una corda and check again. I do this because it is not uncommon for the use of both pedals together to flex the keybed enough to create a knock under one or more of the glides. If so, I turn the glides down until they don't knock, let go of the pedals, and check again.

I think this to be the most solid and durable approach, and has, in my experience, yielded the best results .

Regards, [/quote]

Very good procedure, nicely described, but don't you screw the action stack before final adjustments (and without the keys) ?

This sound like a good "defensive" approach. anyway.

Then you don't install any "stress" within the action, from there ? (on NY Steinways) ?

>>Very good procedure, nicely described, but don't you screw the action stack before final adjustments (and without the keys)>

Hmmm.

Actually, there is no reason for the screws to be involved when my intention is to have all the feet in contact with the cleats and no stress in either the keyframe or the action rails.
What I am looking for is two assemblies mated with each other without either one putting stress on the other. ( I suppose the same could be said of a marriage).
Regards,

Thanks Ed, much appreciated.

Yes I understand that very well, but now on the Steinways we work on, the keyframe is stressed simply when the blocks are tightened. The stress is light and apply mostly on the front lip, but with a reaction from the balance and back of the stack, it sort of "lighten" the action and can change the tone/touch quite a bit.

On older ones the stress is often gone, and the keyframe apply flat.

I was instructed to tighten the action before bedding the balance rail, and that seem tho be the best way to have it clean and stress free, on other pianos than Steinways.

I believe that the screwed stack imply some invisible stress, something that cannot be detected with shims, so it is better to have it screwed (and more when the front screw are inclined).

But I agree with your checking it also may show if something warped and where.

I would agree here. This is what I was referring to in my earlier post regarding random foot point loads. The unevenness of the bracket feet introduce concentrated and random point loads when screwed down. I have quantified this effect...actually I quantify it on every frame that comes in to my shop for rebuilding.

As I mentioned above, I shoot to reproduce the existing passive foot plane(uneven) when rebuilding a frame unless directed to do otherwise.

The first thing I do when setting up the frame for rebuild, is to mount bare pin flanges at 6 points along the frame (shank&whip). With the frame setting stress free and unscrewed on the jig, I take height readings off all the flanges individually. Then I look for any feet that are making passive secure contact with the mounting surface on the jig (poplar). I do this by tapping the feet. You can feel and hear when a foot is sitting passively secure on the jig base...there might be 4 or so passively secure feet.

If I were to screw those (say 4) secure, passive contact feet down, the flange heights all over the entire frame would change...sometimes by 2mm or so...depending where the initial 4 secure feet were along the frame. At first this didn't make sense to me, until I realized only one point of each of the secure feet, not the entire surface of the foot, but only one point was making passive contact. Screw that one point down, and it crushes the wood at the point load, until the entire foot makes contact. The compression at that point load may be .25-.75mm, but it then creates random levering effects over other parts of the frame, and accentuates the flange height changes, sometimes significantly. This point load compression scenario will be repeated on all 12 feet locations (2 per bracket).

The point load deformation creates an example of the classic "action interactiveness" effect, where one change effects all the other parts, and must be accounted for...or at least that how I approach it...

When the finished frame is installed on the keyframe, this "foot point load scenario" will be repeated, introducing stresses to the keyframe and action frame. From the in-situ bedding procedures perspective we are discussing here, this would suggest to me the the inevitable stress created by screwing down the frame should be expected and accounted for in the bedding order of events.

Jim Ialeggio

HI Jim ! I did not notice so high uneveness on the stacks I soldered, (also because the rails are cracked in that case generally, I cannot account the stack for an "original "geometry)

Your experimentation is interesting, for sure there is some amount of pressure between the foot and the wedge, and it may move the centers (I did not make such experiments, but I will)

I sold the stack on a flat bench, but I measure /verify the centers. As I use a "copy" jig I have find some nice uneveness in rails, as a warping in the mediums, that was only due to the soldering in my opinion.

Witch order for the soldering, would be the most secure, in your opinion ? (to avoid the warpage due to the soldering heat ?

I am unsure if it is better to begin in the center, or at the extremes. (my gig allows 3 points secured)

Email me privately if you wish.

The correction of geometry on Steinway stacks is all but easy (but necessary often) I like to modify the old stack, just because of the difference in flange size, to avoid too much shimming (the holes have to be enlarged) but actually I would like to use directly the parts at the original dimensions.

I have seen a set mounted on an old NY Steinway that sounded well adapted, geometry wise. I thought of Tokiwa but cannot be certain. Nice shanks, good cold pressed heads, I suppose it is sort of a standard on the old NY models, but liked to know where are the parts from (I can send pics)

Too bad someone rubbed the knuckles with graphite while the piano could be regulated correctly without that, if only one accept to have 6 leads in the basses )

The action ratio seem to be fantasist on the old models, changing from one to the next....

I teach a class called "Understanding the S&S Tubular Rail Frame"...I will be presenting it at MARC this year. Having examined these frames from the perspective of dead-nuts production accuracy across the frame, and having developed the procedures and equipment to actually accurately measure and target this I can safely say that the level of center inconsistency, at least in vintage and stacks well through the 1980's explains much about the inconsistency techs have complained about in S&S actions for years.

When rebuilding a frame in your shop, you will find inconsistency in the newly soldered frames if the old rails have been indexed to the old. This is because the old rails were inconsistent due to:

-positioning inaccuracies in the original factory fabrication...very common
-old split rails often untwist as the split, moving the flange centers with the twist...sometimes increasing spread, at points in the rail, by 2-4mm
-solder faiure...though I don't find this to be a dimensional issue as much as a flapping- around-in-play issue.

also keep in mind:

-new rails all have twist. They are almost never without twist...from 1 deg to 3 deg, which depending on the flange can make a significant difference in flange center locations, unless the soldering jig is designed to correct for this inaccuracy.
-new rails are bowed. Even if you take the bow out before boring, after they are bored, the rails will take on a new bow as the boring changes the stress pattern in the rail/dowel.


Perhaps some twist is introduced in the soldering process, but with my setup, I have not noticed this changing spread more than .33mm, and usually only when soldering the hammer rail. Its more a matter of having a setup jig capable of correcting for the twist and bow the new rails all exhibit.




An understatement. This is an incredibly challenging frame to control in fabrication, especially in a factory production setting.

You mention different aftermarket parts...Each manufacturer's "S&S" flanges have different center location in relation to where the flange sits on the rail. In a rebuild, if you choose to use the existing frame unaltered, you can sometimes use this difference between aftermarket parts to your advantage, as Tokiwa whip flange will decrease existing spread by almost 2mm depending on the setup, and Renner and WNG will increase the spread a bit as well. By mixing and matching, as long as you knew what geometry to shoot for, you could get pretty close, assuming the existing rails are not too wildly inconsistnet to begin with.

I don't mix and match, but I have the equipment to put the rails/centers where I want them with the chosen parts.

Jim Ialeggio

Thank you Jim, Your curse may be very interesting.

decreasing the spread (even 1 mm) ? I though that the Tokiwa parts where more in the good dimensions (at last for the shanks)

I have some new rail I will analyses with what you wrote in mind.

The bowed rails I have seen where misplaced and the bow was apparently due to the thin plank in the mediums, to sold the rail straight, the hole should be enlarged (due to improper positioning somewhere.

Too time consuming probably.

The angles, they are very difficult to manage precisely. I believe we may find similar angles at the wedges and for the rails

I for sure see the challenge it is to have those stacks normalized, (just the feet bending is something surprising !) but still they have many advantages.

I suggest that there are also inconsistencies. in balance rail position, the spread is sensitive if we are in the 111 mm region, but as soon as we are in more common dimensions there is some margin.

Steinways action make me work the action analysis thru tactile feedback, as measuring and finding the mistake source is rarely easy. I am now sensitive to keyboard acceleration (magic line position) and to the quality of the letof/drop rub. Recognizing those as sensations when moving the notes very slowly is often a big help (but mistakes and confusions happens indeed !).

I plan to add a 4th support to my jig , the 3 points allow to align but that could be more precise (that said I have seen new rails soldered with a very basic setup, shims to lock the tubes and a rough control in 5-6 places, in the end may be not worse than originally...).

Very interesting reading Ed, Jim and Kamin, thanks for the replies!
I took some measurements today.

With the action sitting flat on the keyframe (plus 4mm shim front side and 2mm shim back side in the treble to fill the gap).

HCP height is consistent at 144mm.
Repetition CP is 80mm in the bass and middle, and 79mm in the treble and top treble.

without the shims, the treble end HCP would be pulled down an additional 4mm and RCP down 2mm...

String height is 190/189mm tapering to 187mm in the top treble.
Hammer bore is 45mm.

Another thing i noticed was that the keyframe front rail is 2.5mm thicker in the bass than in the treble.This is another thing i don't think i've come across before, what would be the reason for this?-I thought for a bit that it might be to bed the keyframe or something due to the 'twist' in the action, but that would make it worse, as it would want to be thicker in the treble wouldn't it?

The bedding is best with the shims too..

This steinway is very quirky!

I emailed Kent back today with the serial number that he requested, so i haven't heard his suggestions yet..Should be interesting.

I just soldered a broken repetition rail at C 88 because the hole for that stack bracket was not aligned when a new key set was installed. The bracket screw, when installed, forced the action bracket away from the repetition rail and broke the solder joint. After soldering the joint, I plugged the errant hole, and drilled a new hole that matched the bracket location. Just like new!

That's because its a non-parallel strike line action, that is, the strike line is not parallel to the key fronts.
Whew!
An understatement...it is an arcane design. It has some good points but the rails and the rail profile creates for me, shall we say, job security

Jim Ialeggio

Thanks Jim.
I'm just wondering, how would you say the Hamburg steinway action compares to the NY ones?

Ok it.s 6:50am here, i must be still a bit tired!-Why does a non parallel strike line mean that the front rail must be thicker in the bass than in the treble?

Greetings,
I think what was meant was the balance rail has to be lower under the longer keys to keep a constant relationship between the whippen centers, capstan contact, and balance rail. To effect this, the longer keys must be taller.
Regards,

Ah i see, ok thanks Ed...Good to see you are not being censored on here!

THanks Ed, this is what I was referring to.

However, as I thought about this during the day in relation to Musicbased's comment of a thicker keyframe frontrail in the bass, it really doesn't compute. I can see no reason for the frontrail dimension to change...balance rail yes, but it would decrease...but front rail??

Add to that I haven't seen S&S taper the balance rail for B's & D's. M&H does, at least under Bruce Clark, but have you seen other manufacturers address the tapered balance rail?

Jim Ialeggio

No, Jim, I have not. And to tell the truth, until Bruce pointed this out, I had never considered the concept! Once it was drawn out, it was clear. I personally think the congruence of whip center, capstan contact, and balance point is important, but favor the alignment of the three to occur at let-off instead of at half-stroke.

This comes from my understanding of "gain geometry", reinforced by Del's opinion of sliding motion being minimal (if I understood him correctly). When I get the time, I intend to come up with the numbers,but I do know that the action ratio changes during the keystroke, and I don't want it to be decreasing during the latter half.
Regards,
I will have a post later about the four day WNG class in Haverhill. After formal training and 36 years of this work, I was still treated to at least one epiphany a day!

Looking forward to your report on the WNG class!

A New York B is a parallel strike and all the keys are approximately the same length.


Ed Foote ... Are you in Haverhill now??