Back Checks

2016-04-02 – This is a modification I no longer perform. Although the back checks do work as advertised, the improvement in playability is minimal and they require such precise adjustment that I’m afraid they’ll turn into a liability over time.

The action on a Rhodes piano is a dramatically simplified version of the same mechanism that converts a key press into a hammer strike on a traditional acoustic piano. Although Rhodes was able to remove all but a few essential pieces of the more complex action while still maintaining an acceptable feel, one part many wish they hadn’t omitted is the hammer’s back check. Without a back check, hammers tend to bounce after hitting a tine and returning to the key pedestal. When a hammer bounces, it causes two problems. Since the hammer is tied directly to the damper, a bounce pulls the damper away from the tine momentarily allowing the note to ring a bit longer than it’s supposed to. Additionally, while it’s bouncing, the hammer is not properly positioned for a subsequent key stroke and a great deal of efficiency is lost. This can make playing fast passages a hit-or-miss proposition.

Manufacturing your own simple back check system would not be too difficult but for those not interested in spending time reinventing them, Vintage Vibe offers their own solution for sale. The kit consists of a set of metal tabs, pre-cut self-adhesive felt strips and a set of the same type of screws that are used to mount the piano’s pickups. The idea is that the tabs are screwed to the tops of the keys so that the attached felt strips provide a landing pad for the hammers. In their instructional video, they indicate that the tabs are pre-bent to fit an early model piano with “hybrid” half-wood and half-plastic hammers but that for other models, the bend angles will need to be adjusted. For both of the late-model pianos in which I’ve installed these, I’ve had to significantly change the shape of the tabs so that they can reach the hammers.

To start off an installation job, single back checks are installed on the highest and lowest keys. These two are then used as references for a pencil line to mark the locations of the remaining pieces. Unfortunately, for the very first back check, I slightly missed the mark and it would not reach its hammer. Since moving the screw such a small amount was not an option, I used a file to enlarge the hole in the metal tab allowing me to scoot the tab forward enough to fix the problem.

After the reference line is drawn, the rest is a matter of bending, affixing felt and installing 71 more times. Although the back checks can be installed without removing the keys from the piano, I opted to perform the task in good light on my bench. In their video, Vintage Vibe shows the mounting screw being driven with no pilot hole while the narrator declares that the soft wood will accept the screw with no prior introduction. When installing these on my wooden-keyed Fifty Four, I found this not to be the case and quickly resorted to drilling pilot holes to avoid splitting any more wood. I drilled pilot holes in all of the plastic keys of this Seventy Three.

As they’re being attached to the keys, each back check must be fine tuned to provide the correct degree of support for the hammer. The allowable tolerance is almost impractically tight as the felt must prevent the hammer from bouncing while also allowing it to fall freely to its proper resting spot. Finding the perfect angle to accomplish this can be a bit frustrating but is do-able. This adjustment is much easier to perform individually as each back check is installed rather than after they’re all in and access is blocked by neighboring keys on both sides.

After installing the last back check, I was left with a handful of extra metal tabs and even fewer screws and felts. The extra tabs are important because the process of developing and practicing the perfect bend pattern incurs some casualties.

Tines and Tone Bars

“Asymmetrical tuning fork” is how Harold Rhodes described the assembly that provides the vibrations detected by the pickups in a Rhodes. One leg of the tuning fork is the tone bar and the other is the “tone generator”. The tone generator consists of the tine itself and the metal block into which the tine is permanently mounted. The tone generator is connected to the tone bar by a screw to form the complete fork.

Once all of the tone bars had been removed from the rail, they were separated from their respective tines. This was most easily accomplished by mounting the tine block in a vise before attempting to loosen the hex head screw. The connection between the tine block and the tone bar is critical to the proper resonance of the entire assembly and these screws are typically torqued quite tightly.

One of the tone bars on this piano was a real curiosity. In all but the highest registers, the tone bars have a 90 degree twist so that they can vibrate without risking interfering with their neighbors. On one bar, this twist was in a different location than on all of the others. Additionally, its stamped number duplicated that of its lower neighbor – I had two #22 bars and no #23. My first guess was that at some point, #23 had been lost or damaged and the repair person only had access to another #22 and, in an attempt to make the #22 vibrate at a slightly higher frequency, it was straightened then re-bent at a different location. That theory was challenged by the fact that there was no evidence of a previous bend. Every bend that was made left tool marks behind and there were none where a previous bend would have been. Also, the metal would have shown signs of stress from being twisted, untwisted and re-twisted. I’m left to conclude that the bar was incorrectly twisted at the factory, found to be suitable for a different pitch and put to use in that capacity.

The tone bars for this piano were sent to the plater as-is, without any preliminary cleaning. The tone generator blocks were almost completely covered in oxidation which came off easily using the wire wheel. The tines themselves are not plated and many showed evidence of corrosion even after being cleaned. There are no numbers stamped on the tone generators so to make it easier to reassemble everything later, they were each labeled before being put in a box for storage.

Each tone bar and tine is mounted to the plywood tone bar rail by two long screws. Each screw also passes through a spring which suspends the bar just under a half inch from the rail. Rubber grommets regulate the screws’ clamping power leaving the bars the freedom to vibrate and even wiggle around a little. As I noted in an earlier post, at some point someone modified the highest ten tone bars by removing all springs and replacing them with single rubber blocks installed under only the rear screws. I believe this is to promote a more tonally dynamic response by allowing the very rigid small bars to move around more than they otherwise would. Although the blocks appear to be designed specifically for this purpose, I’ve never read or seen any indication that this was a factory-authorized modification. When I reassembled, I replaced the blocks with standard spring supports.

  • ¬†Update 12/1/11

A couple of the experts at the Electric Piano Forum informed me that the rubber “stand-offs” were actually factory parts.

Different grades of springs were used across the scale of the Rhodes pianos. The lower registers received the stiffest springs in order to manage the increased mass of the larger tone bars. In the middle, a softer spring was used. For the upper third of the piano, the rear spring remains the same as the middle but the front spring is even softer. I’m not certain this pattern holds for every Rhodes, but it has for the three I’ve had apart so far.

There are also three (sometimes four) sizes of tuning springs that follow a similar pattern up the scale. To help keep all of these similar springs organized, they were typically color-coded by the application of some sort of very utilitarian paint or dye. I don’t like the sloppy appearance of the dye and even though it represents a divergence from my otherwise fairly rigid ethos of maintaining the original appearance, I scrubbed all of the color off along with the crud before reinstalling the springs. To clean the tiny tuning springs, I first threaded them onto a length of some fence wire then just pulled the wire back and forth across the wheel. Without the color-coding, it can be difficult to differentiate between some of the tone bar springs but they could always be accurately distinguished by squeezing them together to compare their relative strength.

It was very satisfying to finally be able to reassemble the harp. The piano suddenly went from a pile of parts to a fully-functional Rhodes. Not much progress has been made since as I’ve been having too much fun playing it.

Plating

I spent too long debating how best to deal with the harp frame and tone bars. These internal steel parts were originally zinc plated and most have not held up well to the passage of time. Zinc protects the steel by being a ‘sacrificial metal’. Instead of preventing corrosion, the zinc simply corrodes preferentially to the underlying steel. When the zinc corrodes, a white powdery substance is left behind in its place. Although even this powdery oxidation still provides some protection, once the zinc has been converted, the steel is largely left to the mercy of the elements and soon begins to rust and pit.

My debate over these parts centered on whether or not I was going to outsource their restoration. There are a few sources for home electroplating kits and I gave very serious consideration to the product offered by Caswell Inc. which promised everything I’d need to go from bare metal to a professional-quality finish. This appealed to me for a couple of reasons. Ideally, I’d like to perform as much of the pianos’ restoration work as possible myself. Also, the plating kit would likely pay for itself after just a few pianos by saving the cost of using a third party for this service.

Ultimately, I was scared off from the home plating option by the apparent complexity and scope of the job. Even as a kit designed for amateur use, this is no simple operation. It involves multiple stages each requiring its own combination of potentially dangerous chemicals, tricky operating procedures and some degree of luck. The steep learning curve aside, a plating setup – especially one that could accommodate the large harp frames – would occupy a significant amount of real estate in my small shop.

Among the plating businesses I looked at, it seemed standard to divide plating jobs into two general groups: drum and rack. Smaller parts (up to maybe twelve inches long) can be plated using a drum that rotates like a rock polisher. The parts tumble around inside until eventually all surfaces are consistently finished. Larger pieces must be hung from a rack and dipped individually in the plating bath. For my parts, I needed to find a plater that could provide both drum and rack services.

Prior to sending the parts away, they needed to be cleaned and polished. Zinc plating is extremely thin and does nothing to compensate for imperfections in the surface of the base metal. I tried using a blast cabinet to prepare the parts but even using a fine blasting medium, the result was a satin finish that was too far from the original look. Instead, the bulk of the cleaning was done with wire wheels, both on the bench grinder and chucked into a hand drill. After achieving an acceptable finish with a wheel, a final polishing was done with some 400 grit polishing papers.

In the old days, after applying zinc plating, parts would often go through a ‘chromate conversion‘ process that added a protective layer to the zinc. This is what gave the Rhodes parts their iridescent yellow coloring. Although you can still have zinc chromate plating done, it’s more recently been heavily regulated due to reasons discussed in the movie Erin Brokovich. Replacements for the chromate process have been developed but they offer little or none of the protection provided by the old, more toxic process. Although the plater I used offered the old stuff, I chose to go with his replacement option. The resulting color looks pretty good but, at least on the tone bars, could not stand up to even the slightest abrasion.

Pickups

With their delicate, hair-thin copper wire, Rhodes pickups are particularly vulnerable to the dampness that only caused cosmetic damage to the rest of the piano’s components. After a pickup was wound at the factory, it was generally wrapped in plastic tape to protect the wire. Earlier pianos that I’ve seen were wrapped with clear tape while in later years, opaque white tape was used. The consensus seems to be that the pickups wrapped in white tape are more prone to broken internal connections rendering them completely dead. My guess is, it has nothing to do with the tape itself but probably the grade or type of wire used to wrap them. Even for a white-tape unit, this piano had more than its share of deads.

As pickups were removed, they were each tested. The good ones measured consistently in the low 190s which is right about where they should be. Others didn’t report any continuity at all and by the time I had removed the first twenty pickups, I knew by the size of the reject pile, I was going to have my work cut out for me. By the end, I had 46 working pickups and 27 out of commission.

I’ve wound Rhodes pickups before. My Fifty Four also features white-tape pickups and so far, six have expired on me. When I did those, I judged the amount of wire to use by measuring an original pickup’s diameter and comparing the new one. I have no idea how close I came to the correct number of wraps because it turns out that the 42 AWG wire I used was much thinner than the original and I ended up with a pickup that read over five times hotter than it should have. This time I was able to get some proper 38 gauge wire but I still wanted a better way to calculate the wrap count.

While looking for counting mechanisms on-line I saw some mention of using a computer mouse. The center scroll wheel on most mice uses an optical sensor to determine when the wheel has been moved. This can easily be re-purposed to count revolutions of a pickup winding spindle.

To replace the mouse wheel’s little pinwheel that serves to alert the optical sensor to movement, I went with a CD-ROM. I cut a small notch in the edge and mounted the disc to a hanger bolt with some washers and nuts. Next, a mounting block was made to hold the pickup as close to the center of the rotational axis as possible. The mounting block was screwed onto the wood-threaded end of the hanger bolt and the machine-threaded end was chucked into a hand drill which was clamped into a vise. The mouse sensor was just stuck to a two-by-four and the height of the drill was adjusted until the CD-ROM passed between its eyes. Now, for every rotation of the pickup, the sensor would see one brief glimpse of its sending unit and send a single movement signal to a computer.

To use the output of the mouse sensor, I wrote a small Visual Basic program. In addition to showing the number of rotations, a target can be specified and it will also show a count-down and a progress percentage (hmm.. maybe I’ll add a progress bar). An audible alarm would be very helpful so that I don’t have to take my eye off the pickup. I started to add a beep that increased in frequency as the target approached but ran into a little issue and set that aside.

The counter works amazingly well. I first tested it a few times by making around fifty to seventy wraps with string then manually auditing the count. When I started wrapping pickups, it took a few tries to arrive at a good number but it seems like 3,100 puts me in the best range. The first was a little low, the next a little high and the third was just right.

Before winding each pickup, the previous wraps were removed with a box knife. This actually turned out to be the most tedious part of the job. It was difficult to avoid damaging the plastic bobbin with the knife.

I don’t do much de-soldering and don’t have the proper tools for it. Besides that, I like to keep the number of times I heat up the lugs to a minimum as the plastic they’re mounted in tends to soften easily. So, instead of melting off the old solder, I used a small nipper to trim the lugs. This was important as any little snags can easily catch the wire as it’s being wrapped.

Even though pickup wire appears to be uninsulated, it actually is covered in any one of several types of coatings though I think generally either polyurethane or enamel is used. I believe I’ve heard that Rhodes pickups were wrapped with enamel-coated wire but I think my wire is poly. Either way, the insulation was first scraped off the end of the wire before it was wrapped around the lug to ensure good conductivity. The beginning of the wire is wrapped around the first lug – the one closest to the pickup’s mounting tab. Wrapping it clockwise around the lug helps the wire find the groove in the bobbin allowing it to reach the hub without obstructing the subsequent wraps. To wrap the wire around the bobbin in the correct direction, the drill was run in reverse.

I wanted to be able to control the speed of the winder and using the drill’s trigger for this control was not a practical option. Since I also wanted to be able to start and stop it remotely, a dimmer switch was in order. I basically wired a dimmer switch into a short length of extension cord. This provided a pretty good range of speed control.

Note: this is not the proper method for controlling the speed of an electric motor. A common dimmer switch will eventually cause the drill motor to self-destruct.

I found that to get the correct number of wraps onto the bobbin, I had to make them pretty tight by pinching the wire between my fingers as it ran off the spool. The spool was mounted a few feet away, oriented so that the wire fell off the end rather than unreeling by spinning the spool. With this heavier wire it probably would have worked either way but lighter gauges may not be strong enough to handle the load of trying to turn the spool.

Using this setup, I’m able to wind a pickup from start to finish in about ten minutes. Again, the most time-consuming part is probably removing the old wire. Otherwise, it’s a pretty easy and fun job.

There are a couple of videos on YouTube. One of the counter being tested and another clip of an actual winding job.

The counter program can be downloaded here. It’s just an executable and doesn’t use any special libraries so it shouldn’t need an installer. Just download and run.

The source code for this program (much of which was copped from somewhere on the web) can be downloaded here. It’s written in Visual Basic 6.

Hammers and Dampers

I don’t think this piano was ever played much. The hammers showed only a little wear and all of the felt seemed to be in decent condition. I suppose it could have been serviced prior to being stored in whatever damp basement was its home for years but I think a disinterested owner is more likely. I’m going to stick with the existing felt everywhere except the bottom of the name rail but I replaced the hammer tips to give the piano the greatest chance of sounding and looking brand new when finished.

With the action completely disassembled from the piano, it was the most convenient time to work on the hammers. The first step was to remove the old tips. This was done by peeling the tips sideways until they broke free of the hammer end. The neoprene hammer tips on a Rhodes are grouped into five sets, each set is formulated to a different hardness. The softest material is used on the bass notes with the tips getting progressively harder as you move up the scale. The last and hardest set of tips have a wooden core to make them particularly hard. While the softer tips tended to leave remnants of rubber as they were pried off, by the time I reached the highest notes, the hard rubber was snapping off quite cleanly.

Even though the treble tips were breaking off cleanly, they still left an uneven surface of residual glue behind so every hammer was carefully cleaned using a small sanding drum on a Dremel tool. I first wore the coarse sanding drum down on some cast iron to take some of the bite out of it but it still required a steady hand to keep from cutting and burning the plastic hammer surface. As each hammer was cleaned, I snapped it back into the hammer comb already mounted to the action rail.

Since I was waiting on delivery of the new tips, I proceeded with the dampers first. The damper strips had some of the most tenacious corrosion I had yet found. I ended up resorting to the same Dremel sanding drum to remove most of the staining. The thin metal damper arms act as their own springs and as such are subject to metal fatigue over time and I’m worried that my aggressive cleaning tactics may have prematurely aged them. They’re relatively easy to replace so I’ll wait until the piano’s playable before deciding if they’ve still got enough spring in their step.

The dampers are connected individually to the hammers by way of a bridle strap so that as a hammer rises to strike the tine, its damper is simultaneously pulled out of the way. Old bridle straps can stretch out making it difficult to maintain the proper relationships between hammer, damper and tine. At first I wasn’t sure how to determine the condition of my bridle straps until I learned that the hammer comes equipped with a built-in length gauge. A small protrusion that I had previously thought was a remnant of the mold used to form the hammer turned out to be a hook placed there to aid the installation of the bridle straps. The distance from the hook to the strap’s anchor point marks the appropriate length for a correctly trimmed bridle. Since all of my existing bridle straps fit perfectly to that hook, I determined that they did not need to be replaced.

With the dampers all installed, I could finally reinstall the damper rail. Through the years, damper rail hinges have been designed in different ways. On this model, one of the rail’s hinge pins protrudes further out the side than the other. The excess length is occupied by a spare tone bar spring so that the opposite pin can be inserted while the spring is being compressed. The whole arrangement is locked into place by what appears to be an extra wide tone bar clip.

When I first installed the damper rail, I was concerned that it was only making partial contact with the lowest note’s damper spring. Since there was plenty of rail at the treble end, I decided to try to shift it a bit by the addition of a washer. The small shift was enough that it prevented the locking clip from being inserted so, rather than make any more modifications, I put it back the way it was.

By this time, the new hammer tips had arrived. These are installed by just gluing them in place with a small amount of medium viscosity super glue. When I replaced the tips on my Fifty Four, a couple of the high ones fell off after a little bit of playing. The hard rubber on the wood-core tips is very smooth and slippery so this time I first scuffed them with a small file before gluing them down.

With the action rail ready to go, I was able to reinstall everything and the project started to once again look like a real piano.

Action Rail

The action rail is an aluminum bar that supports both the hammers and the dampers. It in turn is supported on each end by the harp supports and the whole assembly is bolted to the bottom half of the plywood case. After removing the harp frame, the action rail was completely disassembled. One end of the rail was much more corroded than the other and a few screws needed some convincing before they’d come out.

Most of the action rail was covered in corrosion and some white dust. When I cleaned the guide pin rails, I used Barkeeper’s Friend with decent success. I started cleaning the action rail with Barkeeper’s Friend and a Scotch-Brite pad but it was messy and I soon learned that a dry Scotch-Brite pad by itself was the way to go.

As I disassemble the piano, every screw that appears salvageable has been cleaned up on the bench grinder’s wire wheel. Although it probably also strips plating, the wire wheel is very effective at removing all rust and corrosion from the screws.

The cleaned action rail still shows some pitting where the worst of the corrosion was but, all in all, its condition has greatly improved.

Key Guide Pins

I knew when I first saw them that the key guide pins were fragile. Obviously, they would require careful handling and storage while they were outside the case. It was with this in mind that I began cleaning their aluminum rail and promptly broke two pins. And then I broke one more.

I don’t know where I can get replacements for these. There’s a seller on ebay that has the balance rail pins but not the front guide pins that I’d broken. My only option was to attempt a repair. When an unorthodox repair procedure is called for, one of the first tools I usually reach for is a torch and this job was no exception.

I heated up a picture hanger nail and used it to poke a hole in the plastic base. I also used the hot nail to kind of center-punch the underside of the pin. A drill bit was used to finish the hole in the pin.

My plan was to drive the picture hanger nail up through the base to support the plastic pin. The nail head protruding from the underside would prevent the pin base from sliding into the aluminum channel so I melted a little divot to allow the nail to sit flush.

The hole I’d drilled in the pin was slightly smaller than the diameter of the picture hanger nail. After pushing the nail through the base, I carefully heated its tip and pushed the pin down over it. The plastic melted just enough to provide a surprisingly snug fit. The repaired pins were sturdy and ready for action.

Tolex

Using one of the old pieces of Tolex as a template, I began by cutting the sides from the new material. I decided against trying to use any of the old face pieces as templates as they were too torn and brittle to work with.

I worked on the bottom half first. With no leg compartment and only one face surface to take care of, it would be a simpler job to start out on. I thought it might be easier to spread the glue with the case hanging from a roof truss. This turned out not to be the case and when I got to the top half, I didn’t bother with that apparatus.

I applied the Tolex glue liberally to both the case and the fabric using a small 4″ paint roller and a foam brush. After waiting about 15 minutes for the glue to become tacky, I applied the first end piece. Although it went on well enough, I was a little concerned that the glue wasn’t holding as well as it should. The glue is touted as having an extraordinarily long working time (like days or even a week) but as the job progressed to the face piece, its poor short-term performance was making things a little frustrating.

I found the screen repair tool I had to be very useful in creating tight inside edges.

I spent a while trying to smooth the flat surfaces and keep the edges tight but eventually decided it was an act of futility and left it for the night. By the next morning things were looking up. The glue had finally agreed to display its adhesive characteristics and I was able to touch up a few trouble spots from the night before. As further time passes, the glue seems to be improving its bonding strength and I’m more confident now that it will mature to an appropriately semi-permanent finished product.

All in all, I’m pretty happy with the way things turned out. There were a few victories and a couple of lessons learned. I started with four yards of Tolex and had only barely enough to cover all surfaces.

Case Preparation


After removing all of the external case hardware and everything from the inside, it was time to peel the Tolex. In some places, the glue was still holding strong and the vinyl took some wood with it. Other areas had clearly seen more moisture and gave up their covering without a fight.

The Tolex repair video produced by Vintage Vibe says to use the old pieces as templates for cutting the new ones. Some pieces were better suited than others for this. Luckily the more intricate end pieces survived largely intact.

Eventually, both halves of the case were freed from all remnants of Tolex. About three hours and a couple of 80 grit sanding discs later, they were taken down to bare wood. Removing the old glue from inside the leg box was exceptionally tedious and eventually I had to quit and call it good enough.

Despite having gotten wet, the plywood seems entirely structurally sound. The only places that were a little soft was where screws had transmitted moisture into the wood. Several of the old screw holes needed to be drilled out and plugged with new wood. I ended up deciding to just refill all of the spots where wood screws would need to grab hold.

Before putting on the new Tolex, all of the nicks and dents needed to be filled with Bondo. Finally, after another pass with the orbital sander, the case was ready to be covered.

Waterlogged


eBay is chock-full of Rhodes pianos and I immediately found a good prospect. It was definitely in poor condition but still had all of its components (although no legs or sustain pedal).

The seller posited that the piano had gotten wet at some point in its life and, after examining it in person, I’d have to agree. Much of the Tolex was damaged and peeling and some of the metal inside was a bit corroded. The threads of many of the screws that secured the case hardware showed that they had spent some time steeped in damp wood.

When I hooked it up to an amplifier, I was pleasantly surprised to hear sound coming out of most of the pickups. Only one tine was actually broken, although more will likely need to be replaced. Almost every tonebar/tine was completely out-of-whack, some pointing nowhere near their respective pickups while others pushed directly up against them. Apparently one tonebar had gone missing at some point and was replaced by a mismatched unit. The repairman re-bent the incorrect tonebar to make it resonate at a frequency closer to what was intended. Also, someone removed half of the upper-register tonebar screws. I’ve seen this modification before and believe it to be an attempt to allow the tines to travel more to better produce the trademark growl on hard key strikes.

The piano was not supplied with either its legs or sustain pedal hardware. Unfortunately, those items are expensive to purchase – either originals or aftermarket reproductions. I’m not concerned with maintaining only original components so I’ll be keeping my eyes open for options.

Being a relative novice to the world of Rhodes pianos, I was a little surprised to see that the keys were completely plastic. Even my late-model 54 has wooden keys. This doesn’t concern me too much but I’m sure these plastic keys are considered less desirable than the wooden ones. This also means that the guide pins are plastic and the keys do not employ felt bushings where they contact the pins. This is bound to change the feel of the instrument and I’ll be interested to compare it to the wooden keys when it’s playable. Given the piano’s history, the plastic keys are probably a blessing. The soft wood of the older keys would certainly have soaked up any moisture that had gotten near them.