Piano World Home Page Forums Our Most Popular Forums Piano Tuner-Technicians Forum Repetition lever problem

James, try photobucket.com

Hello James.
Interesting to meet you here. We had some email exchange a few years ago.

A few comments:
This configuration is not commonly found, but there were quite a few pianos with this kind of set up. Never seen one? Expand your horizon, look here: player wippen

When played in this manner, the hammer should drop about 2 mm from let off. If it drops further, either there is a problem with the drop screw regulation or the repetition spring is so weak that it cannot hold the hammer up.

First test: regulate the the rep spring so that the hammer moves up smartly out of check when the key is released very slowly after the hammer is in check.

This being a Feurich, it is pretty safe to assume that the piano played properly when new. So what has changed?

- the wippens are original (no new ones will fit your piano), and the shanks/flanges are too. Unless the action rails have been moved, the action geometry can not have changed, so don't worry about taking all those measurements you have been asked for.

What is with the wippen drop pads (leather or cloth) Were they replaced? If so, they may be too thick. Try some thinner material, as has been suggested. If they are original, they are not the problem.

If the knuckles are original, they are probably flattened. It would have been a perfect time to have them replaced while the hammers were refelted.

I would not reccomend trying to re-leather them. Two reasons - the cloth underneath is old, compressed and flattened and has lost most of its resilience. As well, it is very difficult to releather 88 knuckles with the consistancy required. The leather has to be very snug etc. Much easier to replace the knuckles. As a temporary fix, you could use a darning needle to pull some knitting wool through the knuckles just beneath the buckskin. This can help to bring a worn knuckle back into shape. It was one of the "old" tricks. Again, it can be inconsistant etc. New knuckles are almost always required on a piano of that age when the hammers are worn down, especially on a player.

One other thing I would try as a test is to raise a hammer flange using a cardstock shim, say 1.5 mm. Regulate that note - how does it play now? I have seen a piano very similar to yours where the solution was to raise the hammer flange rail by 2 mm, followed by full regulation. This helped to reduce friction by optimizing the magic lines and raised the hammer center pins to a better distance to the string plane.

Hope these suggestions help.
regards,

Hello Jurgen, thanks for the pics. Very informative.
Your advice are good indeed but saying that one may refer to rail distance on very old action like being original is pushing the confidence a bit far.
Beside with the minimal measures I ask it takes less than on hour to have a proper sketch of the convergence lines and to assess if the position of things is correct.
for the spread, I'd agree up to some point : if the original screw/washer marks that are on the whippen and hammer rail are one and unique , the position did not change. But that does not mean the action stack is at its original height, the keys are on the good height, and so on.

Simples test like sharps burying space under the key cover, level of the whippen, position of the jack under the roller tell much on action geometry.
I am not the kind of guy that pretend that originally the factory worker made lots of mistakes, but they of course sometime did, and between the department happens that the one that worked on dampers had to compromize because the one that worked on keyboard where slightly off standard measure, then the ones that drill the hammers have also to compromize.

This kind of thing lend to unusual spread sometime, it helps to understand why the springs are weak for instance.

I aways have find interesting to take a few measurements, because I have seen to much good technicians , that are so sure they will be able to have the darn thing work, that they forget to check if there is not a glitch somewhere. Not always possible to correct it but it may be interesting not to accentuate it. Changing the vertical dimension within the action stack is a very strong measure, it can apply if the hammers are too long and if there is a little space under the pinblock, but it is better to check why there is too much friction at the roller/jack contatc before and what will be the convergence at the key when you will be done. The vertical dimension within an action is a fixed important parameter, in fact almost the only parameter that deos not have to vary IMHO. I'd better raise the stack than doing so.

I appreciate you enlighted comments and ideas, for what I have read on this forum yet.

I believe that being an engineer, the owner of the piano will appreciate a little jump in the piano action standards. and so will I !

Best regards.

"This being a Feurich, it is pretty safe to assume that the piano played properly when new. So what has changed?"

Well this I totally agree with !

And I understand your point not to overcomplicate things, weak springs is not uncommon in old pianos due to age indeed, one may need to strenghten them without putting kings and kill their progressivness, but it is not so easy nor possible if you don't have the correct tool and method.

Regulating screw for the spring is not always present.

Senior say he learned to regulate, and he was really trained at it. Don't know what it mean exactly , but I suppose he knows how to test springs.

Your step by step explanations are useful anyway.

I meant to write: regulate the rep spring (not "screw" as I wrote). I have edited my message. Thanks for catching that.

Hi guys, what a lot of response!
I really should have said, re-felted hammers, original shanks etc etc etc. I tried to save time by saying the rest was perfect (springs etc) to save everyone time, but thanks anyway. :-)
Interestingly, today I did some measurements.
The centre pins geometry (some say magic lines) are almost perfect. The shanks have new knuckles, and if these were 0.5/1mm larger, the centre lines would be perfect and my repetition height would be solved. This is what I need to do to solve the problem.
WHAT IS INTERESTING, is that when I measure the string height, and then the pin height and hammer length (sorry don't know terminology), when the hammers are 90 degrees to the strings, they are 8mm below the strings. In other words, on contact, the hammers do not strike at 90 degrees. 8mm is a lot, however the instrument still sounds absolutely great, and I assume that this must have been a compromise in designing an action for a Welte?
I'll upload some photos tomorrow. I'll sign up to Flickr or something to put them on. I also have a thread on IBach I want to start!
Many thanks Kamin and Jurgen!
James

What you state is called overcentering.

its a way to lighten the friction at letoff time.
On old instrumenst the string's plane was often very high. actually range 190 - 196 mm)
To keep the hammer striking the strings at 90 ) (which is usually what is wanted) a 91 to 92 ° of the hammer on the shank is usual on many european brand and overcentering (renner standard action) is 3 mm.

I noticed overcentering as high as 6 mm (1920 Bechstein mod A with Schwander action) but then the hammer "rake" angle was 94 or even more.

never seen 8 mm particularely if the letoff is above the spread line which is all but usual (it is on or slightly under)

I believe, as you said, the 5% you are looking for are ther (it should be possibly more than 5% BTW)

It is possible that because of the special whippen it may not be easy to open the angle of the hammer that much. What is most important is to respect the good strike line .

What is you theory for the "magic lines" being "perfect" ?

Nowadays with a so high overcentering, the action stack is may be very low, or may be it was made necessary for the installation of the Welte for a reason that excapes me. It is not possible to get the normal drop with those out of standard parameters. (something probably changed)
We use measures and dimensions in that trade to understand a tad what is there. For instance we say hammers are glued 130 mm on the shank with a rake angle of 92° , a 10 mm knukle (roller) at 17 mm from the center pin. This is the way information is useful given.

I could help if you like to try to understand but you talk like if you know yet what you are doing. and provide not enough information.
so it is not possible to me, not to give you advice to do this or that, because I will not , but to say what I understand in your action and why if feel so light. (with a 8 mm overcentering you have not any friction and tactile feedback at letoff, visibly you understood that the roller was 1 mm too high from the jack at letoff moment. That mean the action is too low by that amount.) changing for a thicker roller (is it 9 mm ? 10mm ?) is rarely used, 11 mm rollers are only seen on some rare grands. like Fôrster ,it is possible you need it, but I doubt of it (Jûrgen possibly knows better on that aspect)


Checking how is the key vs the whippen will clear the point where somthing have been changed.

as small change as 1 - 2 mm in parameters can modify the whole behaviour of a grand actioon, as you will certainly discover.

Look at the strings print on the heads

If the print of the strings are off the center of the head (they are on the back of the hammer if the hammers are glued 91° but overcentering is 4 mm more) that is the result of the too large overcentering un corrected with the rake angle.
You could check the strike line with along wooden stick under the shanks raising the hammers on the action so they are at the strin's plane level. You will then see if they are yet in line or not.

Some tweaking may have been used because of the Welte, I only can speak of normal action parameters, but indeed the best strike line is all is wanted, (with hammers hitting at 90° the strings) on an accoustic instrument, and I was said that this aspect can be left unatained when a player system is used, because the instrument then have less nuances abilities, and it should be very light to help the player system to function well

Senior, Unless you have something against Google, I'd suggest using Picasa for your pics albums their software to deal with your pictures locally is very good and fast, the search engine of course is good, it have a nice feature that recoginize any new picture or graphic element you have on your PC -if you use one - and the search for doublons is amazing.

[img]http://www.flickr.com/photos/24794856@N03/2344227530/[/img]

hello!
I've finally uploaded some pictures. the first two show the action type and the third, the non-adjustable damper rail stop to which I will refer.
[img]http://farm3.static.flickr.com/2003/2344227530_e64a51d728.jpg?v=0[/img]
[img]http://farm4.static.flickr.com/3241/2343398283_08a453dd58.jpg?v=0[/img]
[img]http://farm3.static.flickr.com/2383/2343399451_9bec568171.jpg?v=0[/img]

With regards to Kamin's point on the flexibility of long keys, I noticed that the keys are made from a 2 part laminate, with the top cm being made from Beech - presumably to increase the stiffness.

Kamin: thank you very much for telling me about the geometry and terminology - I find this hugely interesting. Are there published articles about this? If not could I please encourage you to write a post with some diagrams? Sadly, my short stay at home has ended and I'm unable to take any measurements for you at this time, although your offer was very kind.

Jurgen: I experimented with raising the hammer flanges on a few, however the problem is that the wippen then also needs raising to relevel the hammers correctly and this in turn raises the part of the mechanism which lifts the damper! As can be seen in the third photograph, I can't raise the damper rail level without altering the permanent wooden stop.

The knuckles are 9mm, whereas on the same action in my Roenisch, they are 10mm. Changing them to 10mm would allow me to lower the wippen by 1mm, thus increasing the travel between rest and the repetition lever screw - thus, hopefully, solving the problem.

Action parts for amateurs seem horribly expensive in th UK. I shall wait until the summer when I go to germany to pick up a new set. If it works, and I remember, I'll add a final reply to this post so the outcome is recorded!
Thank you to everyone for their comments.
best wishes,
James

Hello James ,

Very funny action and very nice pictures.

Indeed at first sight the whippen rail semm secure, but it have probably be unscrewed then put back, tyou may check precisely if the rail did not move even 0.5 mm to the right, as it would explin a few things.

I see the damper upstop rail as being normal, with screws and a possibility to put the stop rail higher or lower. (possibly it is a tad low)

I dont understand well what is the green square wooden stick under the keys. it does not limit the key dip ?

Looking at the way the treble hammer is glued on the shank with a 90 ° angle or little less, the action is not intended for the huge overcentering you noticed. BTW the 9 mm roller from Abel is yet being displaced iby the jack because of the bad ivercentering and it is yet not in line 90° from the shank. I doubt the rollers have to be 10 mm but you may ask your technician.

Looking at the setup of the action on the table, the stak seem to be a little to low, as the hammers shank is too high - the jack is not well in line with the roller probably the hammer have been raise to have a norma travel (45 to 47 mm from the strings. (the shank rest rail is also a little high.) This was may be due because of not enough felt on the heads. Difficult to see if the stack is too low but the springs are p^robably weak, as the hammer shank are not supposed to stay at rest on the rest rail, but a few mm above.

You may want to check if the whippen is level when the key is at rest (the front of the key are may be a tad high, how much of the key pin is inside the mortise ?)
etc ect.
i'll be pleased to check the numbers or exchange more, but you visibly don't have much time for that.

You coud buy new rollers, but it is not an easy task to change the rollers another time, particularely if initially the job was done by ABel, and it was probably the case. They certainly have not mounted smaller rollers. bigger ones will add friction and can cause problems if the action is not intended for. (you can make some testing indeed). Numbers will say the rest, they are not that difficult to obtain.
I appreciated the pictures and the concerstaion. Will send you my mail by PM if you like to exchange about.
Unfortunately tehre are no books on those kind of things , it is particularely long to have some clue about and you can hear more than one bell ringing. !

Best regards good luck for you and that nice instrument

BTW, the way you placed the parts on the picture is exactly showing what is possibly there :
the distance between the hammer center and the whippen center is too large, the jack is on the weong side of the roller, and the whippen is not level but too high.

Just an option but it make sense.

The huge bending of the backchecks is also strange, unusual, but may be due to the separate whippen or inappropriate installation at first, or the thick leather gluied on the tail of the hammer (usually there is nothing at that place the hammer's tails are only wood (?)

Oh yes James ! your lessons on grand regulation may date a little you regulated the jack the wrong side of the roller (I should have seen it from the first time).
Try the same setup with the jack in its normal position it is not normal that the jack regulating button is screwed totally as it is. The edge of the jack which may align with the wooden core in the roller is the other edge , not the one which is seen on the action part picture.

Most players have a mechanism that keeps the keys from moving when the player is operating. I suspect that is what it is.

The damper stop rail has slots so that it can be adjusted. Many of them are nailed as well as screwed, or perhaps the felt on the back of the rail has become embedded into the finish by the pressure of the screws and being held against there for so long. If there are no nails, you might try prying it loose.

I think the hammers need to be longer not just change the bore distance. More felt. The original may have been very large and filed a lot. Player pianos would tend to groove the hammers more rapidly than just normal playing. Since you refelted, the mouldings could not have been adjusted to make the overall length proper. A dense modern day hammer of the same length will weigh much more than the parts of yesteryear.
A 10 mm knucke would have to be installed to see the difference but when regulated, would the core line up better instead of worse??

Thanks BDB, yes the damper upstop rail may move.

It is ok to stop it with little nails but more to treat the source of the problkem and installa a screw or a capstan to limit the travel of the sustain pedal mechanism (generally easy to do)

Keith the hammers are yet very tall, may be some felt is missing.
The treble hammers where undercentering at first sight, asi it is often seen on those era and kind of piano. But then, with the Welte, perfect installation is not possible.
8mm ovecenter is way too much, but also they need to have little friction for the Welte to function well so the misplacement have benn done purposely may be.

not realist to find higher hammers. I have a set cold pressed that could be used (I checked it on a 1917 Bosendôrfer) but definitively you'll never have that thin treble with modern hammers. and the weight ... on those kind of instrument there is a loss in value with new hammers unless you get a special order with hand made heads.


May be 2 mm felt

Hi guys,
I'm afraid I may be causing confusion using wrong terms!?
The point of the third picture was not to show the rail which stops the dampers bouncing too high (which is clearly adjustable) but the vertical stop which prevents the damper rail which lifts the dampers from going too high,
Clearly such things could be modified - I merely wanted to demonstrate that the dampers are currently in the same region as when they left the factory. Raising the whole action 8mm for example would mean that the dampter stop would need cutting shorter.
Next, when I rested the action on the cloth I didn't line things up properly. Clearly I would and have not set the jacks up on the wrong side :-)
A hammer that overcentres by 8mm strikes the strings 2-3mm further forward on the string than had the hammer hit at 90 degrees. As I've said before, the hammers/shanks/pin blocks etc are as original - Another point which makes me think this was an original compromise. I can not imagine the hammers having had more felt on them - have you ever seen a top hammer with 8mm of felt on it?

If I re-jig the whole action to reduce the overcentering, I may have to reset the treble hammers further down the shank as the action itself can not come further forward to adjust the strike point.
I'm startingto agree with Kamin that the rollers are proably the same size as original (Abel wouldn't make that mistake)

So the big question is, who thinks I should live with it, and whole thinks I should re-gig the action? This second option is something I've never done before.
I may return home this weekend and take some of the meaurements Kamin has suggested...
cheers,
James

James: I think you are onto what I alluded to earlier:
these player grands with the pneumatic stack above the keys are configured quite differently (geometrically) than normal grands.
Many compromises had to made to accomodate the player action.
For example, the balance rail is in a completely different position, the key ratio is closer to 1:1 than the more normal 1:1.9. As well, the piano action is situated quite high above the key bed, because the string height had to be pushed up so high. This changes a whole lot of things!

There is little question for me that the "normal" action geometry can not be attained with this type of piano. While 8 mm hammer overcentering really raises my eyebrows and deserves looking at, it may be worth the time to carefully look at what is there now and exactly how it is working. This cannot be done any other way that being at the piano, really. Is it playing as it was designed? If so, it may be quite difficult to improve on that - the designers that Welte and Feurich had were no slouches.

For example the overcentering: Your hammer 88 is surely the same length (+/- 1 mm) as the original. So if you were to raise the action 7 mm, what would happen?
- what about the capstan/wippenheel interface?
- what about the interface with the player action?
- would you still be able to accomodate the damper lift?
- where would the strike line end up?
- when you shift the action to adjust the strike line, how would that effect everything else?
etc, etc.

If you have a lot of time on your hands, you can play around with all of these things and more. However if you simply want to enjoy your piano, look into the most simple and effective way to correct the current problems - I suggest the knuckle size needs re-visiting.

I would be very surprised if the magic lines on your action are anywhere near ideal, and it may be possible to improve that aspect a bit. But straying too far from the original set up may not lead to a workable solution. In all, the geometry of these instruments is a compromise, and I suspect the compromise chosen by the original designers was not the worst one.

These piano may not play as nicely by hand as their non-mechanical sisters. But that is not to say that they have to play poorly.

Hi Jurgen,
that's probably the reply I've wanted to hear all along :-)
It does bug me that the repetition screws are as high as they can go, but at the end of the day, it actually plays extremly well.
I have made a few small adjustments: for example if I set let off 2.5mm below the strings, then I find that the jacks do not always clear the knuckles as well as I'd like. As the instrument is in a small room and tends on the loud side, I've raised the hammers in their rest position from the standard position.
For now I'm going to leave it. After the work I did last week, the action is a lot more consistant than it was before I did the work. I now feel it's good enough to dedicate all the time required to adjusting bleeds/bellows etc for the welte part of the instrument.
As a final point, I'll just say that the same style action in my Roenisch does regulate to all the standard dimensions. So not all of these mechanisms required compromises.
I think that concludes this post for now.
Many thanks to everyone who contributed!
James