Playing!

I have now recorded a few videos of the piano being played; the process is new to me and I consider it all a work in progress!
However, if you click on the following video you should be able to find my channel to see the other things I have recorded so far.
Thanks for your interest!

acknowledgements

Although I did all the rebuilding myself, I could not have done so without the knowledge and advice of those who have come before, which has been passed down for persons such as myself.
I hereby inscribe an “honour roll” to acknowledge those who helped me along the way; my sincere apologies to anyone I may have inadvertently missed!

The Honour Roll:

Arthur Reblitz: for instructive and expansive volumes on both the piano and the player piano

John Ross: for advice, encouragement, and lending of select equipment and tools

Mark Reinhart and Tim Baxter: for moral support, technical advice, and good counsel

Ralph Nielsen and Bruce Newman: for technical tips and instructional videos

Doug Bullock for technical advice and provision of supplies

Julian Dyer, Don Teach, Andrew Barrett and others: for administrating the Facebook player piano groups

AMICA and MMD admins: for keeping these organizations going, in furnishing access to technical advice and documents (particularly Terry Smythe for scanning and uploading much material online)

John A Tuttle: special thanks to John for answering many many emails, providing a dedicated website and video channel for information, and making supplies and parts available for purchase, to help “keep the music rolling”!

My significant other: extra special thank you for your support, encouragement, patience and tolerance of a hobby which took over a surprisingly large area of our home, for quite some time!

DONE! + Project Summary

Although it has been a long time in coming, at long last the piano is done and playing well. I hope to soon upload some videos of it in action; watch this space!

You may be asking yourself the question: what does it take in terms of time, materials and cost to restore a player piano? Let me provide a breakdown of my own project, for your information.

Materials required: I will group the materials into a couple of different categories.

Category 1: General “soft” materials most often required, including cloth, leather and tubing, to rebuild the action. You will likely need most of the following:

  • heavy bellows cloth (1 Yard)
  • motor cloth, for air motor and other medium pneumatics (1 Yard)
  • thin striker pneumatic cloth (1.5″ x 100′ roll, or about 1Yard)
  • tubing for tracker bar to stack, etc (150′)
  • tubing for expression devices and valve boxes (20-50′)
  • hose (twill covered, 3/4″ ID) for air motor and other supply (15-20′)
  • hose (twill covered, 1-1 ½” ID) for plenary supply (15-20′)
  • Gasket material (3-4 Yards)
  • Leather for flap valves (4 pre cut strips, or 1 Yard)
  • Leather nuts (diameter as needed) (200)

This grouping of materials cost approximately $700 USD at the time of publication. The availability and supply of these materials is diminishing, which means the cost is constantly increasing incrementally.

Category 2: Valve materials. A thorough rebuilding also includes the primary and secondary valves, as well as the other miscellaneous rotary and pallet valves. You may need some or all of the following:

  • Secondary valve stem guides (100)
  • Fibre or metal valve discs (100 or 200)
  • Leather for inside and outside facings of both primary and secondary valves, as well as others such as cutoff and expression valves, etc (4-6 square feet)
  • blotter gaskets for secondary valves or other (1-2 Yard)

This grouping of material cost approximately $500 USD, but note that this figure also includes some labour cost of custom punching blotter gaskets and one set of valve facings

Category 3: Player piano specific tools, such as small pneumatic jig, vacuum gauge, tracker bar cleaner, test roll, and so on. This category cost approximately $200 USD.

So far we have a running minimum total of $1400 USD, for the cost of a first time rebuild of a player piano. However, there are futher costs to consider. What have I left out?

International shipping, taxes, duty, currency conversion: I live in Canada, so the $1400 USD, plus logistical costs ($400 USD) and conversion, actually translates into $2200 of my local currency! If you live outside a country where the materials are readily available, you must factor in this cost!

Non-player specific tools and supplies: there are many tools and supplies which are needed, including woodworking and metal working hand tools and some power tools. A glue pot is very handy as you will be working with hot hide glue a lot. Then there is the cost of perishable supplies like the glue, shellac, lubricants, finishes, and so on. If you don’t have any of this stuff you will have to borrow or buy what you need (estimated cost: a few hundred dollars to get started). You will need a suction generating machine to test your work. A shop vac will do in a pinch, although it is not ideal.

Piano parts and tools: I’ve only mentioned the player parts so far, but obviously the piano itself must be restored to at least the same degree as the player action, to ensure a good result. Piano parts are not cheap, if you need to get into replacing dampers, hammers, strings, etc. This can quickly tally up to a few thousand dollars for these parts, if they all need replacing. If the piano needs refinishing, a new decal, etc this will also be a not insignificant cost to you, in time or money. Apart from the rebuilding, the piano will need regulation, voicing and tuning, all of which require specialty tools used by a piano technician. If you do not wish to have the additional learning curve, you can hire out this work, recognizing that there will be a cost of a couple days’ worth of professional labour as well.

Piano moving: there is no such thing as a “free” piano! A player piano is a large heavy object and should be treated with respect, when moving it. It is a job best left to professional piano movers, for the safety of the piano, the home and the people involved. Budget $300-500 dollars for a one-way, local piano move. It is money well spent.

Speaking of time: I mentioned previously that all work was done by yours truly. While there was no monetary cost, you are no doubt familiar with the saying “time is money”. In my workbook I estimate that I logged the better part of 700 hours to complete this project (this is certainly at least double what a professional rebuilder would have spent).
Why so long? Simply put: I have never done this work before, so it took a long time to get into my head. In addition the sporadic nature of my schedule — doing this work on evenings and weekends across several years — was quite inefficient. I made mistakes, which I knew I had to correct in order to make the project worth while in the end. To be frank, if the global pandemic had not arrived to impose a quarantine period, I would probably still be working on the piano yet! I would add that the hours listed do not account for reading, researching, writing emails to ask advice, staring blankly and scratching my head trying to understand how this stuff works!

I did not set all these things down to discourage the would-be rebuilder; rather, my goal is to provide a reality check so that you know what to expect, when you want to seriously undertake restoring a player piano. If I and others can do it, you can too!

Thank you for following my blog, and I wish you good luck in your own project!
Stay tuned for further adventures in the pianola world!

DONE?

Is the piano done at long last?
Ask yourself the following questions:

Does the tempo of the roll function independently of the volume?

Is the piano “easy” to play?
By easy, it is meant that the treadles offer a proper amount of resistance, and one can feel the air being pulled out of the chest with each push of feet.
It is NOT meant that there is no resistance at all, and one must pump like a maniac to play the music!

Do the control levers, buttons and expression devices all act as they should?

Does the paper of an older roll stay true on the tracker bar, even if there are some occasional rips or tearouts along the way?

Is the piano able to play large successive chords and/or arpeggios, with many notes at the same time?

In other words, does the piano feel, sound and look as it should?

If the answers are all “yes”, congratulations — you’re all done!

Together again, rebuilding complete, ready to play! Although the photo shows the piano with the case parts removed, it is finally time to reassemble the cabinet as well.

Stack Installation and Final Steps

Now that your stack is completely assembled and thoroughly tested, let’s put get it back in the piano (do read the previous post for prior steps, if you have not already).

Carefully put the stack back in its home, without touching the hangers against the piano wippens. Slide the stack back into place and secure it with the large screws.
Install the tubes for suction supply to the motor, and the damper pedal feed, as well as any other expression devices or the shifter if necessary. Ensure that this tubing does not kink or interfere with the piano action.

Connect the mechanical linkages for the tempo indicator rod and the transmission; these need to be regulated mechanically, so that the range of motion of the control levers is optimized.

Tempo and Rewind/Play levers

For the play/reroll lever, regulate the linkage so that when the lever is in play position, the pinion engages with the take up spool gear; it should engage positively, but not so much that it bears strongly onto the gear. When the lever is moved to the reroll position, the pinion hub should slide over enough so that a protruding pin (or other arrangement) engages the reroll sprocket so that now the motor is powering the reroll clutch shaft above.
In conjunction with the transmission, the play/reroll lever must also cleanly activate the stack cutout valve, so that when in reroll mode no suction is supplied to the stack, in order that you don’t hear the music playing backwards at high speed!
Conventionally, a player action features a governor override valve, so that when in reroll mode, the air motor receives unregulated (maximum) suction to reroll as quickly as possible.
When in play mode, this valve should be firmly closed to ensure accurate functionality of the governor, so check that this adjustment is regulated properly.

The tempo lever regulation is coordinated with the tempo valve and the indicator / pointer found at the bottom of the spoolbox (normally). When the tempo control lever is all the way to the left, the tempo is theoretically “zero”; the v-slotted tempo valve in the lower action or governor should be closed, which means that the air motor is receiving no suction and does not turn. Adjust the pointer rod so that it accordingly reads zero at the indicator. Likewise, when the lever is all the way to the right, the tempo valve should be wide open now, but just so. This should correspond to a tempo reading of 120 on the indicator. A further check is that the motor should just begin to turn when the pointer indicates a tempo of approximately 10.

The tempo indicator, sitting at the bottom of the spool box. The small tempo pointer is highlighted in black. Note that there is also a larger pointer which can be adjusted to a relative position on the music roll, to follow tempo change suggestions on certain music rolls which are annotated in that way. Normally these indications appear as coloured lines which weave left and right across the roll.

With this done, now the governor itself needs regulating. The above regulation needs to be done well, for the following regulation to be effective. The spring tension of the pneumatic and knife valve need to be calibrated as follows: At a tempo of 70, seven feet of paper need to pass the line of holes in the tracker bar. You will probably need to manually measure and mark this on your test roll, as I did.

Timing lines marked on a test roll, at 1 foot intervals

With the test roll accurately annotated, play the roll from the starting point, marking time on a stopwatch. If the 7′ mark arrives before time, it is running fast which means you need to weaken the retaining spring of the governor pneumatic. Conversely if the roll is running too slow, you need to tighten the spring. Although it may seem tedious after all the other regulation which has already been done, it is important to get this final adjustment just so, to ensure smooth and reliable playing.

A motor governor with typical design features. The spring marked “2” can be turned in the retaining bracket “3” so that the tension level is increased or decreased.

Tubing + Testing (Round 5)

It’s time for tubing — yes!
This is a sure sign that we are approaching the end of the project.
On a double valve stack, the tubing (or retubing) is effected at the end of rebuilding, when we know the valves and pneumatics all work.

This is another task which is pretty straightforward, but still needs a degree of care and planning.
It’s a question of both cosmetics and functionality.
If the tubing is stretched too tight, it may kink near the nipples or come off prematurely.
If it is too loose, it may kink on itself, or get caught in the piano action backchecks.
So, get it in the Goldilocks zone, and make it “just right” in terms of function and appearance.

The smaller size from the tracker bar to the nipple board (piano notes) and to the shifter valve box are usually size 9/64″ inner diameter. The expression and supply lines may be a larger diameter size like 3/16″, 7/32″ and so on. Measure the tubing based on the nipple fitting on which it will need to live. And note that while tracker bar tubing is generally sold in 100′ lengths, you actually need at least 140′ to complete a whole stack – ask me how I know this!

Tubing the stack nipples to the tracker bar. A note that the top row of tracker bar nipples corresponds to the even numbered notes of the stack, helps keep me on track (to avoid mistakenly skipping a nipple)
Stack tubing done!

The tube or hose should be barely snug on its fitting; enough so that it will not fall off accidentally, but not tight to the extent that you have to use a lot of force to push it on. For smaller diameter tubing (e.g. tracker bar), what helps is a bit of lubricant when trying to get it in place. A liquid polymer used for piano work, will serve well for this application, or failing that, good old human spit also does the job! This makes the job easier, and also safer for the piano, as it is not so difficult to inadvertently bend or break the long and delicate nipples on the back of the tracker bar. This preventable scenario is to be avoided!

In the event that your stack still has the original lead tubing, and said tubing is still in good condition (not deteriorated or crumbling), it can continue to serve if you feel it is safe to do so. If there are problems or damage that are evident, it’s really best to replace wholesale with neoprene tubing (again, working safely with the lead), instead of trying to do extensive and uncertain repairs to the old lead tubing.

Once everything is back together, it’s time for the final bench test of the stack. Hook up your suction supply to the stack and shifter device, with a vacuum gauge for calibration. Put a music roll in the spool box and wind it down onto the takeup spool. If you have a diagnostic test roll, this is an ideal choice. Testing the notes sequentially will help identify problems quickly and accurately. Fire up the suction and manually advance the transmission gear so that the roll plays through its notes. Make sure to check which note is the first on the roll — not all test rolls play 88 notes!

As before, if all notes work well, with good actuation and repeating, at various “tempi” (feed rate of the music roll), and if the shifter device keeps the paper tracking as it is supposed to, and you can do all this at 7″ of IWC suction, you are in good shape!

Here once again is Bruce Newman demonstrating what a properly restored stack looks like on the bench.

If you haven’t done so, finish assembling the spool box and any other components, to get the stack ready for its new home in the piano.

Before putting stack back in, this is the time to address any outstanding issues with the piano. Give the piano a good tuning, and voicing, if necessary. Double check the regulation and performance of the piano action. You can also check the functionality and regulation of the pedal pneumatic at this time, since it is easier to make those adjustments now.

Stack Reassembly + Head Shelf Installation + Regulation

Once the stack has been thoroughly tested and you are satisfied, it’s time to install the top shelf and the other peripherals as well.

The items that live on the top shelf include the spoolbox (which is commonly integrated with the shelf), the shifter (tracking) device, the air motor, and sometimes some other small valve boxes which serve as signal boosters (as is the case with mine).

The “head” consists of a shelf (horizontal piece in line with the yellow arrow above word “transmission”) and the various components attached thereupon

Assuming that your shifter and other valve boxes have been tested, and that your motor has been tested and regulated already, these things can be mounted on the shelf.
If you haven’t already done this, you can check out my earlier post on the shifter here, and my post about the air motor here

This is also a good time to tidy up any outstanding cosmetic issues, should there be any. I am thinking specifically of hardware and switches around the spool box area, as once the shelf is fully loaded these are harder to get at. I noticed that I had neglected to replace an old cloth washer on the spring loaded idler chuck for the music roll, so I went ahead and disassembled that completely and gave it a bit of spit and polish. I did not replate any of my hardware, but for the round and oval head screw heads, I did clean and polish them, to match the transmission which had received the same treatment. It was a more uniform look!

final polish and replacement of cloth washer in the roll idler chuck

We’ll look at the transmission in more detail shortly, but for now, reinstall it in its home, make sure everything is lubricated and free, in terms of the bearings and so forth. Different rebuilders use different products, including motor oil, sewing machine oil, 3-in-1 oil. I chose Mystery Marvel Oil. Whatever you use, only put it where you need it, and remember: “less is more”! Paradoxically, too much lubricant can spread to the upper brake drum; making the drum sticky over time and causing excess drag on the music roll!

Regulating the position of the take up spool and the clutch, as well as other transmission adjustments

Now, the spool box itself has to be regulated too, oh yes. First we align the take up spool to the tracker bar, so that it is centered in position (use the shifter/tracker holes as a reference). On my piano there is a lock nut on the right side of the gear shaft, in the transmission, to fix the location. With that done, now we set up the music roll idler (left) and clutch (right) shafts.
What is required here is that when a music roll is placed, and the paper drawn down and wound onto the take up spool, three things need to be in alignment:

  • The shifter (tracking) device, in neutral (center) position
  • The clutch (reroll) shaft centered on the tracking cam
  • The paper roll centered above the tracker bar, just as the take up spool is centered below


I have noticed that although the width of the paper on most piano rolls is 11 1/4″, there is variation in the thickness of the spool ends, which changes the precise location of where the roll sits over the tracker bar. This in turn will change the position to which the shifter defaults, when the piano begins to play. There is a bit of fiddling involved to achieve a happy medium between rolls of different brands.

Also note that all bearing points of the shifter linkage must be free, to respond smartly to the slightest signal of misalignment from the designated tracker bar holes. This is important!

Brakes: most transmissions have two brakes; one brake for play mode (which bears on the upper clutch/reroll shaft), and another for reroll mode (which bears on the lower gear/drive shaft).
Current convention dictates that the play brake should be regulated so that it presses just lightly enough to keep the paper from flying around loosely.
Most contemporary rebuilders are of the position that the reroll brake should be disengaged completely, so that older or original rolls have less of a tendency to rip or break when being rewound at high speed. Use your own judgement!

Cumulative testing, round 4

With the partial stack back on the bench, let’s review where we are, and confirm where we should be. At this point, the pneumatics were all tested on their decks, so we know they should be tight. The secondary valve chest has also already been tested, so that we know the chest is tight both when the valves are on, and when they are off.

While regulating the stack to the piano, the activation of notes may have shown any weaknesses in the secondary valves or pneumatics. If yes, fix them first. In other words, it is a good idea to continue the idea of “cumulative” testing by ensuring that the secondary chest and decks work well in concert, together, with no leaks heard either when the valves are on, or off.

Now let’s add the primary chest into the mix (assuming you have a double valve style stack), and run the tests again. The primary chest has the bleeds inside, so if there was an issue with sluggishness of returning secondary valves when testing repetition, this should be corrected and repetition should now be lightning fast. That is to say, the valves and pneumatics need to repeat as it is humanly possible, and then some; 10 reps per second is a good starting point. Any other problems relating to the interaction of the primary and secondary chest should manifest by now. For example, I discovered at this stage that two of my valves were weak (one was completely stopped, in fact), by troubleshooting the source of the problem I discovered that I had put too much sealant in two channels of the L board, so I had to rectify that problem before continuing.

With the suction attached, actuate each note in turn again, watching for rapidity and evenness of response. Listen for any leaks, and get visual confirmation if possible by calibrating suction loss with a vacuum gauge.

If the three main elements of the assembled stack (primaries, secondaries and decks) are performing well together: congratulations! It’s time to proceed to the last stages of the rebuilding.

Regulation: Player Stack to Piano

Regulation is technically not “restoration” work, yet it is just as exacting, and an important step in the process of rebuilding the piano.

There are a few operations which need doing at this time. I don’t really have many pictures to offer of this process, but some thoughts on the important things to watch for.
John Tuttle has a visual offering here

Although I sorta glossed over it in the previous post, we want to ensure that the positions of the finger pushrods are regulated properly with the pneumatic hangers.
The pneumatics should be at rest (open) position, and the two leather nuts should be just snug around the hanger, but not tight.
In other words, the nuts should be holding the rods tight enough that it does not slip up or down, but not with such a grip that it actually impedes the pneumatic from closing very quickly.
The fingers should be resting firmly on the rest rail; if the nuts are set too high you will see a gap between the rest rail and bottom of the finger, which is not ideal.
If the nuts are set too low, the pneumatic may be pulled slightly up (closed) at rest, which is also not ideal. The other thing to check is the vertical alignment of the pushrods; we want them to be as plumb as possible, so that the travel is true. Although some degree of care was taken at the factory, if you see room for improvement now is a good time to do so.

With that done, it’s time to put the deck assembly into the piano, and see how it looks.
Exact positioning is pretty important here, so I reattached my secondary valve board which allows me to align to my two manifold supply blocks precisely. This way I know that my (partial) stack is sitting exactly where it will live when it is fully assembled and installed.
Why is this important?

Valve board reattached to the deck assembly, to regulate the player fingers to the piano. I have noted with blue tape where some of the fingers need adjustments to spacing.

The importance stems from the fact that we are going to do some regulation and we want it as precise as possible. The other issue to watch out for is the side-to-side alignment of the assembly. We also want the fingers of the player action to line up very well with the piano wippens, and this is the time to check and correct as necessary.

The striker fingers (shown on top of the deck assembly) need to align as best as possible under the wippens in the piano

Once the fingers all line up in the horizontal plane, now we turn to the vertical plane.
The idea is that there must be a gap between the player fingers and the piano wippens, but that the gap must be quite small. The reason for this is that we want close to zero lost motion and efficiency when the striker is activated. We do need to make allowance for seasonal changes in humidity, recognizing that the gap may close when relative humidity is high. What we do not want above all is the wippens resting directly on the fingers, as this will keep the jacks from resetting and prove fatal to performance of repetitive notes.

Exactly how the fingers need to be regulated depends on the model of player action; in mine there are small capstans on each finger, which can be turned up or down. The size of the capstan is unique to this piano, I have never seen another exactly like this model, so I had to custom make my own tool for regulation! It is based on the tool for a similar larger capstan common in Asian pianos.

Once the “starting” position of all the fingers has been regulated, now we need the second regulation for the “stopping” position of the pneumatics. If you haven’t already done so, install the finger stop rail on the player action. It is important to note here that at this stage, the piano action needs to have been well regulated so that the key travel and hammer checking are all consistent and accurate. Otherwise the following regulation time is wasted.
Now, add the secondary pouch board to complete the valve chest, and with the suction supply connected to your stack assembly, activate each note in turn, and regulate each stop button so that the checking distance of each piano hammer, when actuated by its player pneumatic, is the same as the distance when hand played on the piano key.
This is one of those operations where it would be handy to have a third arm, but alas we must make due with what we have!

Once you are satisfied that the travel amount of the player action has been regulated, pull the assembly out and take it back to the bench, it’s time for a bit more testing!

Gluing Pneumatics + Testing (Round 3)

Your finished pneumatics have been tested, and are standing by — in order — to be remounted.
You have a large pot of medium thick hot hide glue at the ready, with brushes, wet and dry cloths, etc.
The temperature of the workspace should be as warm as possible; a nice humid summer day with the windows open is ideal, to give the glue as much “open” time as you can.
Give a final inspection to your decks, confirming that damage has been repaired, and that all your scribe lines are clear for easy location of the pneumatics. The stage is set!
Gluing down pneumatics doesn’t need to be stressful, since hide glue is water soluble and therefore reversible.
Having said that, your first attempt should be your best, since it does involve some work to get back to ideal gluing conditions.
Once again, Bruce Newman to the rescue!

I would simply add that it may be better to leave a slight ring of bare wood around the hole of the stationary board, to avoid getting too much glue into the supply port. This tip came to me from John Tuttle.

If you have good glue, and you move quickly to place the pneumatic exactly where it needs to go, you should have success. Some rebuilders advocate for using clamps, but really hot hide glue will set up to be self-clamping for this application, so in my opinion it’s overkill. If you see a nice little bead of squeeze out around the perimeter of the board, that is a good visual confirmation that you have likely been successful.

As Bruce demonstrates in his video, you can test as you go (or after) by blocking each port completely then trying to open the pneumatic to check the seal. If you had previously resealed all the deck channels, and you also tested your pneumatics, then everything should be well.
On the off chance that there is a small leak, visually inspect to see if there is an interruption in the glue bead. Take a listen for any slight hissing indicating a leak, and if so carefully apply a bit more glue or sealer to take care of it, without touching the cloth!

There are some odds and ends to finish up the decks. If there is a rest rail for the fingers, reinstall it now (after testing), and put down some new cloth which matches the old. After your fingers have been repaired or rebuilt as necessary, and the pushrods are clean and polished, it’s time to put them back. First install the upper leather nuts on the rods. I used a slotted piece of plastic cutting board as a kind of “stop guide” to rough in the position of these nuts. Attach the rods to the fingers (if necessary), insert a cloth washer (if needed) then carefully reintroduce them onto the metal hangers or fingers of the pneumatics.

Returning finger rest rail to the pneumatic deck
a piece of plastic cutting board makes an ad hoc stop jig for approximating the position of the upper leather nut

Careful is very much the watchword here, as it is imperative to not damage your pneumatic cloth at this point – doing so will throw cold water on the great progress that has been made!

Using your fingers (the extremely cautions approach) or the leather nut driving tool (the somewhat less cautious approach), thread each leather nut (with cloth washer) onto its rod.
If ever you slip and scratch the cloth, breaking the air seal, it may be possible to repair a pinhole such as this with diluted PCVE or similar flexible sealant. With diligent work it should not be necessary.

replacing the push rods onto the pneumatic hangers

The last details, if not yet attended to, are the edge gaskets for the decks, and possibly for the mounting blocks for the primary chest (if present).
In fact the measuring and cutting for these gaskets is more easily done before the pneumatics have been remounted, but the application of the gaskets should come last, since they are rendered effectively useless if there is glue or sealant leaked onto the working face.

punching gaskets for the edge of the pneumatic decks

With this done, we now see the light at the end of the tunnel.
Let’s do some regulation!

A 1919 Willis player piano