Swap Transistors and U-235

Okay, partly in opposition to this thread and partly because I've always wanted to ask anyway, is it feasible for transistors to be invented in 1935 instead of Uranium-235 being discovered (that being put off until sometime post-war), and if so, what would happen?
 
Arguably you could have transistors in the late 20s depending on circumstances, though certainly they are feasible by 1935 or just thereafter. Delaying U-235 somewhat might change the course of WWII and there will definitely be butterflies in equipment, especially as military funding will promote and develop computer technology both for weapons design and codebreaking. I think that Hahn will still discover fission on time for OTL and that the nuclear research program will being in earnest, though not sure how far it will go or what the results will be. But if computers are developed earlier with transistors instead of vacuum tubes it might accelerate the project somewhat. If nothing else ASW, radios, targeting systems, and other electronics will get smaller and more efficient so it would be interesting to see where that leads. Should the Germans get the same technology then all sorts of wild applications are possible.
 
There are many ways to get the timeline you want but the one that almost perfectly fits your description would be "Silicon Jumpstart"

Which goes like this:

1930: Experimenting with a new process to purify silicon for his
rectifier research, Bell Labs scientist Russell Ohl accidentally
cracks a cylinder of the stuff, and is surprised to find that, as
a result, it produces a voltage when exposed to light. Ohl shows
this to his boss, Mervin Kelly. New Bell Labs hire Walter Brittain
is introduced to the silicon mystery by his supervisor Joseph
Becker.

1931: Karl Lark-Horowitz adds the electronic theory of solids to
his research program for the Physics Department at Purdue
University, since it seems a fruitful area of research with little
competition that can be done reasonably cheaply.

1932: Becker and Brittain return to the silicon mystery after Ohl
manages to duplicate the process. They theorize that, somehow,
electron-rich impurities in the silicon can send their electrons
to regions made electron deficient by a different impurity. Mass
spectrometry seems to indicate the presence of phosphorus in the
negative (or 'n') end, which would fit the theory. Ohl tries
deliberately salting part of a run with boron, to produce a positive
(or 'p') end.

1933: Following the work of Peierls, Bloch, and Wilson, Nevill Mott
at Bristol University, James Franck at Johns Hopkins, and Karl
Lark-Horovitz with the visiting Wolfgang Pauli make deep advances
into the theory of conductors and semiconductors; Lark-Horovitz
follows up with experiments on germanium crystals.

1935: Becker, Brattain, and Ohl at Bell Labs, Lark-Horovitz at Purdue,
and Julius Edgar Lilienfeld at Ergon Laboratories in Medford, Mass.
independently invent the point contact transistor. Newspaper stories
publicize the precedence battle, but lab notebooks show BB&O have
clear priority.

Now you just need to kill off Arthur Jeffrey Dempster and you have your timeline.
 
Last edited:
Yes, I've often though killing Dempster would delay the discovery of U-235 and thus the bomb, I've just never figured out whether the Mass spectrometer would be required for the transistor.
 
Yes, I've often though killing Dempster would delay the discovery of U-235 and thus the bomb, I've just never figured out whether the Mass spectrometer would be required for the transistor.

Well, if you're not willing to handwave people somehow not looking for U-235 after Dempsters death things get indeed trickier. I would say, you might be able to modify the timeline above. Replace the "mass spectrometry" part by replacing it with "raman spectroscopy", but this in itself would require some sophisticated tweaks on its own.
 
I think there was a miscommunication, what I was asking is did the mass spectrometer play any real roll in the development of transistors, because if it didn't we could say Dempster (and perhaps also F. W. Aston) in the Spanish Flu and be done, but if the mass spectrometer is important to developing the transistor, then we're in a bit more trouble.
 
I think there was a miscommunication, what I was asking is did the mass spectrometer play any real roll in the development of transistors, because if it didn't we could say Dempster (and perhaps also F. W. Aston) in the Spanish Flu and be done, but if the mass spectrometer is important to developing the transistor, then we're in a bit more trouble.

Sorry I was a bit in a hurry when I wrote that yesterday. But to answer your question, I think you need something to analyze the composition of material in almost anything we know as the historical transistor.
Thus my suggestion to have something else than the mass spectrometer do the job, the raman spectroscope

Now if you just want something that acts like a transistor than that is far easier. I used negative resistance based devices in my own timeline for example. They don't necessary require the invention of the spectrometer to work, although it would help to improve them.
 
Thank you for that. Yes, I'm not too bothered about the actual operating principles, I just wanted something that functions like a transistor, and that is a fraction the size of the then-common valve units.

Would it alter much do you think?
 
Thank you for that. Yes, I'm not too bothered about the actual operating principles, I just wanted something that functions like a transistor, and that is a fraction the size of the then-common valve units.

Would it alter much do you think?

It really depends how things unfold in this hypothetical timeline. The only sure thing is that there will be cheaper, smaller electronics. How and what will specifically change is up to the individual author.
There is really a lot of potential for technology to go in different directions. The biggest problem will obviously be understanding the effects of a delayed mass spectrometer (It is probably impossible not to get one a few years later since the basic technology was invented before 1900).
 
I don't think the mass spectrometer would play a role in transistor development.

"I think you need something to analyze the composition of material in almost anything we know as the historical transistor"

Sorry but I'm confused by this? Why would you need to analyze the elemental composition of a transistor when you already know what that is as you manufactured it in the first place? MS and other methods can be useful for testing of component oxidation or other degradation, but to develop a transistor I don't think they are necessary.

Raman spectroscopy measures molecular vibrations, and can be used for analyzing oxides or other degradation products etc. So can infrared spectroscopy. Metals and alloys won't give you a vibrational spectrum though, so Raman and IR won't tell you much about the transistor itself.
 
I don't think the mass spectrometer would play a role in transistor development.

"I think you need something to analyze the composition of material in almost anything we know as the historical transistor"

Sorry but I'm confused by this? Why would you need to analyze the elemental composition of a transistor when you already know what that is as you manufactured it in the first place?

I got the (probably) wrong impression that, in order to understand how transistors work you need to learn about the doping material etc in the first place. Obviously once you understand what is going on you can control what happens in the manufacturing process. But if this is not the case than the original question is solved even easier. The same goes for Raman spectroscopy, I just encounters it when I looked up information on "doping"
and might have come to some false conclusion. The good news is that the original idea can work without any of my suggested adjustments.
 

Ancientone

Banned
Thank you for that. Yes, I'm not too bothered about the actual operating principles, I just wanted something that functions like a transistor, and that is a fraction the size of the then-common valve units.

Would it alter much do you think?
In the real world valves did not disappear and were not replaced overnight. Although pocket transistor radios appeared in the mid 1950s they were rubbish. Larger equipment, TV, table radios, Hi-Fi sets, computers and radar equipment continued to use some valves on "hybrid" boards, especially for high power requirements until the late 1960s, even early/mid 1970s for some equipment.
 
I got the (probably) wrong impression that, in order to understand how transistors work you need to learn about the doping material etc in the first place. Obviously once you understand what is going on you can control what happens in the manufacturing process. But if this is not the case than the original question is solved even easier. The same goes for Raman spectroscopy, I just encounters it when I looked up information on "doping"
and might have come to some false conclusion. The good news is that the original idea can work without any of my suggested adjustments.

I apologise, I misunderstood what you meant.
 
In the real world valves did not disappear and were not replaced overnight. Although pocket transistor radios appeared in the mid 1950s they were rubbish. Larger equipment, TV, table radios, Hi-Fi sets, computers and radar equipment continued to use some valves on "hybrid" boards, especially for high power requirements until the late 1960s, even early/mid 1970s for some equipment.
Very, very true. Some high power applications took a LONG time to go solid state.
 
No. Semiconductors needs higher tech and takes lots of thinking to do. Wiki says a Canadian tried early, but couldn't make it work at his tech.

And WW2 means that's where priorities had to be IOTL.
 
Very, very true. Some high power applications took a LONG time to go solid state.

Big Tubes are still around for high power applications.

http://www.youtube.com/watch?v=L4BKJFIxOf4

Each one of those tubes, given decent antenna gain, worth ERP of 500,000 watts in the UHF range

from the wiki
---
The inductive output tube (IOT) was invented in 1938 by Andrew V. Haeff. A patent was later issued for the IOT to Andrew V. Haeff and assigned to the Radio Corporation of America (RCA). During the 1939 New York World's Fair the IOT was used in the transmission of the first television images from the Empire State Building to the fair grounds. RCA sold a small IOT commercially for a short time, under the type number 825. It was soon made obsolete by newer developments, and the technology lay more or less dormant for years.
The inductive output tube has re-emerged within the last twenty years after having been discovered to possess particularly suitable characteristics (broadband linearity) for the transmission of digital television and high-definition digital television.
The power output of the modern 21st century IOTs is orders of magnitude higher than the first IOTs produced by the RCA in 1940–1941 but the fundamental principle of operation basically remains the same. IOTs since the 1970s have been designed with electromagnetic modeling computer software that has greatly improved their electrodynamic performance.
___


cavity magnetrons then klystrons then inductive output tubes

each more efficient and more precise in frequency output than the last, and all got started during or just before WWII
 
Top