Thanks for all further responses.

Daniel:

Yes, I used a paddle, but one that is wide enough to file up to three hammers simultaneously. When doing the rough shaping, I often worked several hammers together, but when I did the finer work, I mostly worked individually - although sometimes I fine-shaped two or three hammers together to get the curves right, and to check that I have the strike point at right angles. One thing that I wondered about, in terms of gang filing, is this: the wear on the hammers was very uneven, e.g. C4, D4, G4 and A4 had much deeper string cuts than e.g. C#4, D#4 or G#4. I didn't want to remove too much unnecessary felt from the lesser worn hammers. Also, the hammers weren't glued perfectly in the factory. Some heads in the treble pointed slightly down or up, so if I had gang-filed those, I would have removed more felt from the one shoulder than the other. I thought it more important to preserve the symmetry of each individual hammer. But I don't wish to appear as though I'm ignoring your advice.

And thanks for the hint on the rougher grit, I'll certainly give this a try.

rXd:

I'm sorry for having appeared to ignore your previous posting and "groping and stumbling around in the dark". I did take note of the importance of hammer-to-string mating before doing voicing. Some strike points were not quite at 90°, others were slightly rounded, i.e. not striking all strings equally. I corrected this as best I could before attempting any voicing. My apologies for not mentioning this explicitly.

However, I want to add that the mating only improved the tone for soft playing. At medium to hard levels, I couldn't really hear any difference after having improved the mating, and that's why I went on to voice a few sample notes.

In answer to your question about the sound: with one exception (see below), the tone of the filed (unvoiced) hammers on the problem notes was as follows:
pp to p: clear and sustaining
mp to mf: bright and sustaining
f: metallic and a somewhat percussive attack
ff: earpiercing and percussive.
Is it "full and solid"? I wouldn't call it full, since the high partials are so dominant, but yes, I'd call it solid. Is it "thin or sizzling"? "Thin" perhaps in the sense that the overtones dominate, but certainly not in the sense of lacking volume/loudness. And no, it is not sizzling.

The one exception (F#4) sounds similar, but somewhat thinner and sizzly. I checked its string mating, which appeared fine. However, the shank has abnormal side play, so the flange bushing is evidently worn or damaged, and the hammer bounced off the strings at an angle, so it appears it's not hitting the strings in the same place it used to. I'll have to repair this first.

986:

Thanks for elaborating. Compared to what you describe, I've been much more conservative up to this point. The shoulders, especially the high shoulders, were so hard that I could hardly get three #6 needles in more than 1 to 2 mm without their bending and breaking. Rock-hard, really. Even a single needle was a challenge. And I only did a few stitches - certainly I stopped long before the shoulder became anywhere near "easy" to needle. This has taken some of the metallic percussiveness away, but only slightly.

Mark, Thanks for your responce. I was trying to imply that it was a great learning experience for you to shape them individually. But now that you have done that next time you can improve your speed by doing more at once. I have never used a paddle for three hammers but that sounds like a good way. Are you referring to the hammers that were coming loose from the moulding? Or are you saying the hammers were not glued properly on their shanks? Either way I think I agree with your action there too.

About uneven wear: Not everyday does one see a piano like that! But it does happen. I do sympathize with you about not wanting to remove more material than necassary but...I would be more concerned with even hammer blow distance so a good even regulation can be achieved. But you do have to consider waht you are working with and do what is best in your situation. I think you are doing a good job and you have done good research and have good common sense. The latter being very important! I looked at your pictures and your hammers did not look too uneven so don't think I am telling you your job was not sufficient.
Make certain you don't have more than one loose center pin. that will do some wierd things to a piano's sound.

I will be waiting to see what you do and how it turns out!

-Daniel

Yes. I have used a paddle that way but to shape only 2 hammers at a time. I see no reason not to do 3 I'll give it s try. Great time saver and keeps the noses square to the sides.

When old hammer sizes vary a lot, it's really time for new hammers but if the old hammers must be kept then the taking off more felt from some hammers is unavoidable for any chance of an even result, visually or tonaly.

I have used fabric softeners dissolved in water plus a little alcohol to help it soak in. We used to use different temperature water but I have forgotten which way it worked. Ir was a long time ago. I used it a lot when I had over a hundred rental pianos to deal with that were rotating in and out all the time. The problem with using water was the drying time.

We tried acetone as a carrier but the results weren't the same although I read of others having success with it. I always quickly shaped the hammers first. liquids will soak into the loose incomplete layers of felt first and might not get to the hardest felt of the complete layers.

Acetone based hardners and softners can be deceptive. They may appear dry after a few minutes but I like to leave it to dry 4-6 hours if I can. Read the sample pages of "grand obsession" to get an idea of the results of piling on liquid upon liquid without adequate drying time.

I haven't used chemicals for a long time except on odd recalcitrant hammers. I can't use them when I am aiming for specific tonal levels.

The reason voicing comes last in the order of piano maintenance is so that the other aspects that can affect the tone have alredy been eliminated. Uneven hammer hanging must be dealt with first to the extent of rehanging the ones most out of line, using heat on the shanks of those that are not too bad. The object is to keep the strike point even whether gang filing or not.

The state of the flange pinning makes a huge difference to the tone so if some pinning does not allow the hammer to make a beeline for the string, that certainly has to be dealt with first.

Very often hard tone is the result of too loose pinning. Tone regulation doesn't seem to last as long if the flange pinning is loose. Simply tightening the screws that hold the flanges to the rails will improve tone, as will tightening the coils around the tuning pins whether they look as if they need it or not.

These stringing and action things will not make a hard hammer soft, of course, but will give more support, adding depth of tone and balance the tonal spectrum often reducing the amount of needling later.

Needles not going into the felt easily is a sign of an over hardened hammer, either in manufacture or later. I have observed some of the techs that made it through WW2 wipe their toning needles on a felt pad with lanolin on it.

Some of them also used needling technique which would be shocking to those easily shocked (when anybody says they are shocked it always puts me in mind of Victorian old dowagers) but the results were lasting. They were techniques used in the superior factories a hundred years ago. We've lost so much over the years. Those wars sure did kill off a lot of good people along with their old techniques.

I'd suggest we persuade rxd to write a book one day, to fill in the the gaps in the manuals. Meanwhile should we have a thread on forgotten techniques?

One snippet I picked up from Clavio is that at least 50% of the improvement you can achieve in the sound of an older upright has nothing to do with tuning and hammers. Tightening coils must be part of that.


Over hardened hammers remind me of concrete or hard baked earth; pneumatic hammers break concrete up into lumps, and you'd better wait for some rain before going out to dig that earth.

By chance, I discovered one way to soften hammer felt. I wanted to to clean off a dirty mark on a hammer and, as she usually does, Mary came up with a range of methods. One was a dry cleaner in the form of a white paste. I put on a small amount and the effect was dramatic; the felt became malleable, a bit like a jelly. I used this to smooth out the ridges at the end of the grooves. In the morning the hammer was fine again.

So I wonder whether you could soften up a hard hammer a little by applying stuff like that to the flat sides of the hammer; then, when it had become malleable, flexing the felt. The aim is loosen up some of the bonds between fibres, permanently. That should both make needling easier and improve its effects.

Will it work? I don't know, but I'll try if someone here in the UK has some suitable discarded hammers available.

Daniel,

What I was referring to was not hammers coming loose from the molding (although one hammer had this problem), but indeed that the hammers aren't glued properly to their shanks. Or perhaps they were incorrectly burnt in the past. Whether it can be corrected by burning the shanks, remains to be seen.

rXd:

I take your points. Thanks for all the extra information, on giving the tone a more solid support base. (Interesting that the coils should have an effect, seeing that they're not even part of the speaking length?)

All:

I'd be very interested to know whether there is any technique to test the flange pinning (friction) with the hammers in the action. The one test I know (besides hammers that have obvious side play), is to dismantle the hammer and let it swing to-and-fro. The other one I try to use is to see how easily the hammer falls away on a very slow keystroke, after let-off. But this depends on the strength of the butt return spring... Any other suggestions?

The coils are springs attached to the strings.

But why does that affect the speaking length?

The tension in the coil affects the tension in the speaking length. When the hammer hits the string the tension in the speaking length increases so the string tugs on the coil. As the coil is a spring, it will then oscillate affecting the tension in the speaking length. The effects can be noticeable if the coil is loose, otherwise rxd would not waste time tightening it. Tuning provides an example of the effects of changing tensions in the coils and strings; if the coils and the speaking lengths were not connected there would be no sound!

You can tap you foot on the quiet pedal and bounce the hammers toward the strings with the hammer rest rail. The hammer return springs will be a factor in this, but it can give you an idea of what is going on. Or if the action is out of the piano you can remove the hammer return spring rail, turn the action up side down and swing all the hammers at once. (Like is done with grand actions.) It's a bit of a pain to get the hammer spring return rail back in, though. Be careful not to bend the springs.

Ian. What you say makes sense except seating the coils doesn't disturb tuning. Not even the stability. (I must add that I haven't done this with anybody elses tuning so I can't vouch for that except by my own experience of my own work.)

It does, of course, disturb tuning if the coils are separated but I don't know the tonal difference because the tuning is disturbed. Dealng with scruffy stringing is done before tuning.

Simply tuning a unison will liven up that note. Seating at the hitch pin will sometimes liven a note. As Kamin has said, simply stroking the strings of a unison with a feather will liven that note.

rxd. Thank you for the clarification and extra pointers.

I'm afraid I was only trying to make the mundane point that changing the tension in the string anywhere between the tuning pin and the hitch pin will affect the sound. I hope the edited text helps to make this clearer, though I fear not completely, "Tuning provides an example of the effects of changing tensions in the coils and strings."

Duplex scaling might have been a better example of one part of the string affecting another.

There was a report in the Sunday Times yesterday about a camera developed at MIT that works at the speed of light. Even if that is not necessary, it would be interesting to use high speed photography to see how the strings, and the other parts of the piano, actually move. Expect some surprises.

Back to Mark's piano!

The coil is not a spring. Hooke's law does not apply any more than it would on any other length of the wire, and that effect is restricted by the friction between the wire and the tuning pin.

The reason for tightening coils is that if they are loose, there is the possibility that some part of them will slip, affecting the tuning. But if it does, unlike a spring, it never goes back to where it was before, changing the tuning.

It's a torsion spring wrapped round a tuning pin.

There's more spring in the pin than in the coil.

Certainly when the coil is tight enough.

Sorry for responding only now. I was on leave, and though I had sporadic internet access, I actually spent my time with my family, on outings, and playing and working on pianos.

Jeff:

I've never actually seen a hammer return spring rail in the flesh (well, wood). All the pianos that I've seen have the hammer return springs either mounted on the butt and pulling against a silk cord glued to the flange, or mounted on the flange and pushing against a groove in the butt. So, unfortunately, the idea of removing the rail and swinging all the hammers at once would not be practica(b)l(e).

Ian:

Sorry, but the string coils, when mounted under normal tension in the piano, behave nothing at all like a spring. They are tightly bound by friction to the tuning pin. It's exactly the same principle that
... allows one to pull an extremely heavy load by wrapping the rope around one's hand or wrist a few times, while the rope would slip when simply held in one's hand,
... keeps my Roman blinds up if I wrap the cord four times around the cleat hook, while the blind falls down if I wrap the cord only once or twice,
... runs the cable of a car's electrical window winder around the motor's drum a few times, not only once.
This has nothing to do with "springs" of any sort, but everything to do with obtaining sufficient friction by running the wire or rope around its anchor/pin/cleat more than once to prevent slippage.

If indeed the string coils were to act as a spring, there would have to be slippage (back AND forth!) between the coil and the pin. String tension would never allow the coil to move back (unwinding) AND forth (winding) on the pin. Once the coil has been pulled tight on the tuning pin, much like a rope on a winch, there can be no more such slippage - and if there is, it would only be in the unwinding direction. The taut coil is, for mechanical purposes, part of the pin.

And last but not least: to transfer any oscillating tension from the speaking length to the coil and actually get the "spring" to "oscillate", the string would have to slip on the V-bar AND on the pressure bar with every string vibration. To my knowledge, it doesn't.

Sorry, but as OP of this thread, I'll take the liberty of saying that your "spring" theory seems bereft of any mechanical (common) sense. It certainly doesn't answer my wondering why tightening the coils should have any effect on the tone, to such an extent that it should be done before voicing.

Mark

Welcome back!

Yes, I understand the capstan effect, but you originally queried how things beyond the speaking length have an effect; "Interesting that the coils should have an effect, seeing that they're not even part of the speaking length?"

They evidently do, so for an explanation one has to look at the string as a whole from tuning pin to hitch pin.

Won't changes in tension in the speaking length due to hammer blows affect the tension in the string beyond the capo bar/agraffes and the bridge pins?

Weinreich, in his paper discussed in the string coupling thread, treats the bridge between the pins in a unison as a spring. Likewise, if one wanted to do the mathematics, one could treat rxd's slightly loose coil and Jeff's tuning pin as a (non-linear) spring.

When a hammer hits a string with a slightly loose coil, won't the coil will give a bit and then, as the wire is stiff, pull back a bit. Won't these movements affect the speaking length momentarily?

A stronger hammer blow could be more than enough to take up any slack in the coil. Then there would be a sudden shock as the coil became tight.

Maybe rxd or others could say from experience whether tightening coils affects pp and ff equally or not.

No tuning pin coil (or anything else for that matter) is infinitely tight. Sure it acts like a spring, but insignificantly as far as the pitch of a string is concerned. I mean how could this change the speaking length which is the distance between the bearing points???

But it can be significant as far as tuning stability if it causes trouble during the tuning process. This same sort of trouble occurs with some combinations of light rendering, tight pin blocks and spring pins. Messy coils can add to this sort of trouble.

I usually deal with the piano as it is. I don't go through a piano trying to tighten sloppy coils. If they are obviously sloppy, I haven't had much luck anyway. If the original coil was catywompus, then the kinks are there to stay.

Jeff

Wasn't rxd talking about the effects of flange pinning and coil tightness on tone, i.e. overtones and noise, rather than pitch?

The BBC screened a documentary about the Higgs Boson last night. Its existence, postulated by theoretical physicists is still an open question. That question may be resolved by experimental physicists this year.

Whether slightly loose coils behave like springs or not, is there any doubt they can affect tone?

Ian:

I try not to think, let alone talk, about tone more than need be. It is such a subjective quality. Before you know it the color of the drapes in the room affect the tone.

I am not saying what others should talk about, just why I shy away from the subject.

Now as far as string tension is concern, until the string actually slips along a rendering point (which is caused by the difference in tension between the speaking part and a non-speaking part being more than the rendering friction can withstand) then I don't see whatever is going on in the non-speaking matters.

But the point I am making is: while a string is being tuned, all these things DO matter. Some combinations make it very easy to "set the pin" and have stability and other combinations make it very difficult.

Mark, Jeff

The theoretical effects or non-effects of the ends of the string on the speaking length appear to be rather complex and, as you seem to imply Jeff, perhaps best ignored.

In his master's thesis Fredrik Öberg describes how the front duplex in a Steinway affects the speaking length, "Comparing piano tones with the front duplex segments vibrating and dampened, respectively, it is clearly heard that a dampening of the duplex segment influence the main string vibrations as well."

See section 3.4.1 Main tone and front duplex, http://www.speech.kth.se/prod/publications/files/3435.pdf

Ian:

Well, there is obviously leakage from the speaking length to the front duplex or else the it wouldn't sound like a bunch of snakes hissing. But there is another bearing point before the coil and pin.

On reflection, Mark:

The reason for "tightening the coils around the tuning pins, whether they look as if they need it or not," is to prevent the coil behaving like a spring.

When the coil is not tight enough tone must be suffering because string tension is less than it would have been had the coil been tight, and behaved as you describe.

From a mechanical point of view, the initial impulse resulting from a hammer strike will increase the tension at the coil. Then, if the coil then moves at all, tension will fall along the string and the effects of the impulse will be damped.

One possibility is that the increase in tension at the coil causes the wire to flex at the becket. As string tension returns to normal the coil will go back to its original position ready for the next hammer blow. Precisely what oscillation occurs while the note is sounding is anyone's guess.

PS Roman blinds analogy: "the blinds stay up if I wrap the cord four times around the cleat hook, while the blind falls down if I wrap the cord only once or twice". When you wrap the cord (say) three times the blinds will stay up until you tug upwards on the cord. In this situation attach a suitable tethered spring to the end of the cord to represent the coil at the becket; the blind will move up up a bit after a little tug on the cord and then back down when you let go.

No, that's the whole point. At three wraps around the cleat, the roman blind can fall, because between the two of us, the blind and I don't apply quite enough friction to the cleat. But I can NOT pull it back up, because when I pull on the cord, then between the two of us, the blind and I DO apply enough friction to lock the cord to the cleat. The cord will snap before the blind moves upwards.

It can therefore not "oscillate". It can only slip down - but never up.

Have you actually ever tried this? If not, you should. It might prove enlightening.

I think this thread has just about lived its useful life...

Well, of course Roman blind cords wrapped round cleats and piano strings wrapped round tuning pins are not quite the same. The nature of the load is different too; the analogy breaks down. Tension is the issue.

For tightening of coils to affect tone, as rxd describes, there must be some sort of equilibrium point or region between the strings tending to slip, on the one hand, and being tight enough, on the other.

If there is no such possibility, any improvement in tone from tightening the coils must be a figment of rxd's imagination. I doubt this; partly because I am prepared to take his word for it, and partly because people say the becket should take some tension.

However, I heartily agree there is no further value pursuing the discussion.