How fragile was the Talking Doll?
The doll itself was fairly robust. Its head could be broken if struck hard, because it was made of bisque, an unglazed porcelain. The arms and legs were strong and not likely to fracture.
One problem was loss of the crank. Nothing retained the crank on the shaft protruding from the doll's back. Since the crank simply slipped off the shaft, it was easily separated from the doll and misplaced.
The phonograph, however, suffered a number of flaws and vulnerabilities, as reported below.
What parts broke most often?
The Stylus Assembly
Reporting from Paris, France, on May 27, 1889, Albert Blake Dick informed Thomas Edison, "All but two of my dolls are already out of repair. The reproducing needle point has become loosened from the glass diaphragm."
In a letter dated June 3, 1889, A. B. Dick wrote to Edison, "By the way all of the small reproducing needles (or whatever they are called) have finally become detached from the glass diaphragm, and yesterday I bought some porcelain glue and repaired two of them … "
The Wax Cylinder Record
On June 19, 1889, A. B. Dick complained to Edison about "the solid wax Phonograms I have with me" (for use in his sample talking dolls). "These seem to be very fragile and are, I find easily broken in changing to and from the Phono." Dick continued, "I am confident that the more delicate parts now contained in the Phono. must be removed and more substantial parts substituted, particularly for use in dolls, as they will be handled mostly by children, who are not as a rule very careful."
The wax records could crack and break, from contraction of the wax material at low temperatures, which tightens the shrinking record against the supporting mandrel (the "drum"). On January 19, 1890, Charles Batchelor wrote, "We find that it is necessary to put a piece of canton flannel on the drum & ream out the wax cylinder so that the shrinkage will not crack them."
MBJ004;TAEM 90:457 - Image 58 - Entry 642 and Entry 643
Record Surface Damaged in One Spot
Mechanical shock to the doll could cause the needle to scratch the surface of the wax record or break the record.
MBJ004;TAEM 90:457 - Image 66
In January 1890, Walter S. Mallory (General Manager of the Edison Iron Concentrating Co.) took a passenger train from New Jersey to Chicago, carrying an Edison Talking Doll on the trip. When Mallory reached his home in Chicago, he tried "cranking" the doll, but it "made a cracking noise ... as if the cylinder was cracked." Upon examining the record, he found that a "bad place" had been worn on the playing surface, "as if ... the needle had constantly been moving over the surface of the wax [owing] to the shaking of the train."
Some Breakage; Some Useless for Talking; Others Excellent
Curtis H. Kimball (a businessman in San Francisco), wrote to EPTMCo on April 24, 1890: "Among the samples received two were broken and four utterly useless for talking purposes, but those that were in good order are excellent and will meet with a ready sale."
Loose Works; Dolls Won't Talk; Record Rapidly Worn Out
Horace Partridge & Co. (a Boston store known as "a pioneer in the Christmas toy and the Christmas present trade") complained to EPTMCo on April 25, 1890:
We are having quite a number of your dolls returned to us and should think something was wrong. We have had five or six recently sent back some on account of the works being loose inside, and others won't talk and one party from Salem sent one back stating that after using it for an hour it kept growing fainter until finally it could not be understood. We should like to see somebody at once regarding this matter. Let Mr. Briggs come up and see us.
We shipped one doll out to St. Joseph, Mo. The party returned it saying it was not in order and it cost us somewhere about $1.60 to get it back. Of course all these expenses we are charging to your account as we do not feel that we are to blame in the matter as we shipped the dolls as we received them, consequently we suppose you will bear us out in this.
Hoping somebody will come up to enlighten us on this subject as to what to do as we dislike to send out the dolls when there are going to be so many returned.
How did the wax record hold up, with repeated plays?
In one case, not very well: "... after using it for an hour it kept growing fainter until finally it could not be understood."
Decreasing sound volume and increasing noise are telltale symptoms of progressive groove wear.
Why didn't Edison fix the problems and try again with an improved design?
Actually, it seems that Edison did fix the problems.
During May 1890, Edison conducted "Experiment[s] on Dolls to produce better talking." He also worked on a "New Model Toy Phonograph 100 Threads," from May through October 1890. The U.S. Patent Applications that Edison filed on July 30, 1890 stem from this period of intense activity.
On July 18, 1890, Daniel Weld (Secretary and Treasurer of EPTMCo) informed Edison:
At a meeting of the Directors of this Company, held yesterday, at which the models of the new Doll were examined, the following vote was passed: Whereas the models submitted by Mr. Edison appear to be satisfactory it is: Voted, that Mr. Edison be requested to make by hand as soon as possible, (50) fifty machines of materials to be agreed upon between Mr. Edison and Mr. E. E. Magovern.
Unfortunately for Edison and the many other EPTMCo stockholders, by the end of October 1890, EPTMCo was more than $50,000 in debt, could not get a loan on the company's merchandise, and had failed to capitalize their foreign rights. There simply was no money left to fund manufacturing of the improved talking doll.