Edison Opera Repair Experience

[VintageTechnologies]

A friend recently asked if I would repair his buddy's Edison phonograph. It had a broken mainspring. I have hardly the time to repair my own machines much less anyone else's, so I begged off and forgot about it. A month later, he came by and said, “There is a surprise for you in my car”. No kidding, he had an oak Edison Concert (late Opera) machine in the trunk. Sometimes I can't say no to friends and also felt beholden to restore such a rare machine for posterity, so it stayed.

I never had the occasion to repair an Opera even though I have owned one for nearly 30 years, so I welcomed the chance to learn more about the internals. There are so many ways to design a multi-spring motor that I wanted to prepare myself before taking apart an unfamiliar mechanism. I searched the Internet for pictures or repair procedures specific to Operas and found absolutely nothing. This article is my effort to fill that void.

I convinced the owner that I should replace both springs at the same time and ordered them from Ron Sitko (518) 371-8549. They arrived in a few days and looked well made. They were each packaged inside a heavy steel “C” ring. The only disappointment was that the coiled springs were a larger than the Opera's spring cans, so I could not simply drop them in as I have sometimes done before. Each new spring had to be uncoiled by hand from its captive ring. These springs were very stiff and a portent of things to come.

The Opera is large, heavy, fragile and awkward to take apart and service without risking any cosmetic or mechanical damage unless it is properly positioned and supported. I custom built a simple wood box whose internal dimensions are 15” wide x 9.5” deep x 7.25” high. I used a scrap 1x8 board as lumber and nailed a thin plywood bottom to hold the box square. I draped a few cloth hand towels over the edges for cushioning.

I removed the reproducer and two cast iron covers that protect the governor and flywheel assemblies. I flipped the whole machine upside down and set it on top of the open wood box. The cast motor plate snuggled inside the box and wood motor frame rested on the edges of the box. I placed a stack of books behind the box, even with its height, then opened the case, unlatched the support catch and continued opening the case until it rested on the books. Then I had clear access to unscrew the hinges and separate the case and motor frame. Now I had clear access to the motor from any angle.

Removing the spring barrels from the motor plate couldn't be easier, especially compared to other Edison models. That design is similar to many of the Columbia cylinder phonograph models, such as the BK. Loosen one screw, and the arbor rod that runs through the center of the spring cans can be slid out. Then the spring barrels simply lift out. I love it!

Even with the arbor rod removed, the spring cans are still loosely coupled together by a short hollow shaft or hub that has a hook on each end. Every part this shaft passes through is slotted for the hooks to clear. Notice that the spring cans are connected with their end caps face-to-face and have a spacer washer in between. Before attempting to uncouple the springs, twist a flat blade screwdriver gently around the circumference to gradually pry the end caps loose. By twisting the spring cans backward from normal, you should be able to unhook the shaft from one of the springs and separate the two cans. After that, removing the other parts is simple. Under the caps, you will discover a felt washer, then a thin flat metal shield, then the spring itself. Rather than use pliers, I inserted a nail into a small hole in the shaft and used it to twist the shaft and unhook it from the remaining spring. Clean all parts with your preferred solvent.

Notice that due to the spring cans orientation being face-to-face, the spring coils themselves are both installed wound in the same direction. This is distinctively different from some Victor motors that I have worked on. If you drew a spiral on paper to represent the springs in their cans, start at the center and draw a spiral outward clock-wise.

Now a special warning about the new springs I bought from Ron. They are mirror bright, smooth and beautiful with rounded edges, BUT, they are several thousandths of an inch thicker than the originals that I measured. They have less curvature than the originals, particularly toward the outside. I didn't measure, but laying loose on the table, they must have been three feet across. No exaggeration, these were the stiffest and most difficult to feed into a can that I have ever encountered. I have installed lesser springs bare-handed with no other equipment, but not these babies. You will need thin gloves and some kind of vise to firmly hold the can. I recommend you install the springs dry. The first difficulty is fastening the spring to the case hook and then bending the spring enough to follow the curvature of the can. It may be possible to install the spring without fastening to the hook and trust to chance that it will catch when the spring is wound completely. I have done that before, but I prefer fastening the hook if I can. Once you get it started, the first few coils don't seem too bad, but as the can fills, the curve becomes tighter and towards the end it becomes a struggle until the coil finally drops in.

I packed the cans completely full with black lithium/moly grease that I buy in 1-pound cans at the auto parts shop. I inserted the hollow shaft with hooks into one of the coils and gave it a twist to try and hook it into place. It had a tendency to slip out of place, so I decided to also insert the arbor rod through the first spring can and hollow shaft to keep the wobbly parts in better alignment. After stringing everything back together, I carried the assemblage back to the motor plate and carefully held the two springs tightly together with one hand as I removed the arbor rod. I set the springs in place and re-inserted the arbor rod through the arbor supports and spring assemblage. At this point, I had no idea if the shaft hooks were engaged with the springs. As I turned the crank to wind, I heard a click, click, click. Darn, they weren't catching! Something was not aligning. I inserted a flat screwdriver between the spring cans to separate them as much as possible and turn the crank again. Success, they caught! I held my breathe as I slowly wound up a bit, let it run down some, wound up a bit tighter, and so on, until the springs were nearly fully wound. That distributed the grease between the dry coils and boosted my confidence nothing would cut loose. I have played a number of records now and never heard a single bump. That motor has torque you wouldn't believe.

With the motor running, I found two problems topside that required attention. There is a thin, straight wire spring that couples the flywheel and mandrel shaft as some kind of shock buffer when starting and stopping the machine – it was broken in three places. It measured 0.022 inch thick. I found a close replacement measuring 0.205 inch at a RC model airplane hobby shop. It comes in 3 foot lengths, used for control lines from servos to flaps. I had to crimp the wire at each end to prevent it falling from the flywheel. I probably could have bent the wire sightly in the center where it passed through the shaft to hold it tight.

The governor was also unstable and pulsated with loud noise very dramatically under florescent lights. There was gummy oil and grime caught between the governor flange and the shaft it spins on. I also polished all gunk from the flange where the yoke pads ride, and trimmed a bit of excess from the edges of the pads that had mushroomed until possibly rubbing against the flat springs. I also put lithium grease on the worm gear. That cured everything to run steady and quiet.

I won't say there aren't other methods of repair, but everything described worked nicely for me.

[alang]

Thanks for the detailed description. That is a very nice machine.
Andreas

[Kiwi]

I also appreciate the time and the pictures that you have put into your article.
Well done.
Tony

[Retrograde]

...The governor was also unstable and pulsated with loud noise very dramatically under florescent lights.

Rick Perry? Oops, my bad, wrong forum.

BTW, great job on the rebuild, and the notes.

[Valecnik]

I agree. Nice work and very nice narrative. Some more pics would be great if you have time.

[VintageTechnologies]

Thanks to all for the positive comments. I will post a few more pictures soon and clean up a few boo-boos in my narative. I particularly wanted to illustrate what an unstable governor looks like, so I just uploaded my first YouTube video. The old dog learns a new trick!

http://www.youtube.com/watch?v=2aMZ1GB_95E

[AZ*]

...The governor was also unstable and pulsated with loud noise very dramatically under florescent lights.

Rick Perry? Oops, my bad, wrong forum.

Or maybe Jan Brewer? Jerry Brown? Chris Christie? Quite a few to choose from.

[Dave D]

How much thicker are the springs? Mainsprings increase in strength exponentially as the thickness increases. I saw a formula once on a clock repair forum. The difference of .002 can cause a huge increase in wear to a clock. Granted, they have brass plates that are a lot thinner than the motor frames on phonographs, but it is something to think about.

Standard size strength of a clock mainspring on American style clocks was .018. Many repairers use a .0165 mainspring if the clock is being completely overhauled. The wear decreases quite a bit.
Dave D

[VintageTechnologies]

How much thicker are the springs? Mainsprings increase in strength exponentially as the thickness increases. I saw a formula once on a clock repair forum. The difference of .002 can cause a huge increase in wear to a clock. Granted, they have brass plates that are a lot thinner than the motor frames on phonographs, but it is something to think about.

Standard size strength of a clock mainspring on American style clocks was .018. Many repairers use a .0165 mainspring if the clock is being completely overhauled. The wear decreases quite a bit.
Dave D

That is an interesting point you make about accelerated wear in the clocks due to stiff springs. The Edison motors are robust and have steel bushings, so I'm not nearly as concerned. Just keep the gear bushings greased or oiled and they should be okay.

I used a dial caliper to measure, and the new spring was .002 thicker than the (assumed) original. I don't trust my memory now, but I think the new spring measured .033 thick.

The other spring that I replaced was different than the first and I think it was a later replacement. The hole in the outer end was punched differently and the spring must be thinner because it was not nearly as stiff; it was also the spring that broke.

I suspect it is important to match up springs of equal strength, otherwise the weaker spring is going to be wound more fully and therefore do more work and be more stressed than the stronger spring. The weaker spring might be fully wound at some point when the other is only wound part way.

[VintageTechnologies]

In a previous post to this thread, I reported that after reassembly of the Opera, I discovered that the governor was very unstable and would go into a horrible wobble. After some adjustments, I thought I had solved the problem, but then the instability returned. I posted a video to YouTube to demonstrate that instability and wobble.

Looking for subtle clues, I noticed some barely perceptible patches of discoloration on the brass governor flange that I thought might be remnants of corrosion that had been partly polished out. I decided to to disassemble the governor and polish the flange to remove any imperfections, a practice I have used with success before. My attempt at disassembly was stymied by little screws that refused to budge. I posted a plea for suggestions on another thread and eventually was able to loosen the screws after a month-long soak in penetrating oil. While waiting for the oil to work, I had considered what was at stake if I should break off screw heads of an Opera governor. I formed more opinions after I succeeded in the disassembly. I want to share those insights.

If a screw head breaks off, it is even less likely that the remaining stub can be removed without buggering the part, so the hunt begins for a replacement flange, but where? Does anyone make perfect replica governor parts for an Opera? I am sure originals would be hard to find. A few measurements confirmed that Home or Triumph flanges won't substitute; they are made for thinner shafts. The Opera shaft is exactly 0.250 inch in diameter.

I noticed some other differences that make Opera governors unique as well: 1) The flange has three shallow channels milled lengthwise into it so that the weighted springs lie on a flat surface, rather than on a rounded surface as is the usual case with other governors. Something must have required the expense of those extra manufacturing steps. 2) When removed from the assembly, the weighted flat springs are found to be bowed. 3) There are thin shim springs that lay on top of the weighted flat springs, but only from the flange end to the weight. The tiny governor screws fasten both spring and shim at the flange; the other end of the shims is tucked under the weights, causing the weights to tilt slightly, rather than than rest flat atop the springs. Again, there must be some reason for such an asymmetrical arrangement. I suspect all of these features were adapted by trial and error to suppress some type of instability unique to the Opera; otherwise why have them? Therefore, I wonder if regular weighted springs would ever suffice as a replacement and yet be stable?

Before polishing the flange, I examined it at 20x power under a stereo microscope. It is interesting to see how porous metal can look under magnification. Sure enough, I saw three shallow depressions in the surface. Could something that shallow cause oscillations?

I inserted the flange partway into a drill press chuck and ran the drill press at a low RPM. I wrapped a sliver of #400 grit sandpaper lengthwise around a popsickle stick and lightly pressed the stick on the rotating flange to evenly polish out the shallow depressions. I followed up by polishing with some #0000 steel wool. The resulting finish had a beautiful mirror-like appearance.

At last, the reassembled governor no longer breaks into an uncontrollable wobble, no matter how I try to provoke it. However, a slight fast flutter has taken its place, no matter what I try to do. It is not too bad, but I am a perfectionist. I tried several times loosening all the screws to relieve tension, running up the governor to speed and down again, and then tightening the screws. The governor seems to run as quiet as could be expected, and there is no visual wobble or pulsing audible whir to indicate a problem. Flutter isn't noticed until a record is played. I tried a number of familiar records to eliminate out-of-round records being suspect. Clearly something of the delicate balance has been disturbed. Perhaps in time with enough use the parts will settle in and smooth out the flutter.