first production transistor

Huh. I was going to post a Wikipedia link and sneer at you for not looking -- but they just have the patent dates, not hard production figures.

Looking at several pages, and doing some inference, it looks like the point-contact transistors went into limited production in 1948, and junction transistors sometime before 1954. I know that practical FETs came later -- but I have no clue whether it was the 1950s or the 1960s when they started coming in.

I believe the first JFET prototype was made by William Shockley around

Naturally, I wouldn't expect them to have production info.

Oh, right, point contact and not BJT was in 1948.

Are you sure? I checked your link and others and couldn't find anything to that end. If he built them, surely there'd be photos of them, and some core performance data.

Look at the Lilienfeld patentS..he BUILT them.

Yah...his descriptions look like they were copied (stolen?) from the inventor Julius Edgar Lilienfeld. And J.E.F. made them as well..

" Tom Del Rosso "

A US patent was granted to him in 1961.

Early devices, sold under the trade name " Technitron", used Germanium and were intended for RF.

.... Phil

"Phil Allison"

Pun intended.....

..... Phil

The earliest JEDEC jfet registration is for 2n2385, in a TO5 can. This is a sign of rather late (1965?) commercialization.

It's possible that the usefulness of the insulated gate, developed at about this time, was more easily demonstrated and exploited.

RL

Which patents are probably invalid... there's that little phrase "reduced to practice". ...Jim Thompson

Wikipedia and other sources say it can be done just by filing an application.

Ok, then we have different definitions of "made". Usually when someone actually made a device in the last 100 years or so there is photographic evidence, plus witnesses. In the case of Shockley there is no lack:

Hey, they even donned ties for the occasion :-)

Now I am not saying that Lilienfald might not have done the same, I just haven't come across any such evidence.

Probably. Our patent examiners are hired from the job pool of those who TSA rejected :-(

But the law _requires_ reduction to practice. ...Jim Thompson

-- | James E.Thompson, CTO | mens | | Analog Innovations, Inc. | et | | Analog/Mixed-Signal ASIC's and Discrete Systems | manus | | Phoenix, Arizona 85048 Skype: Contacts Only | | | Voice:(480)460-2350 Fax: Available upon request | Brass Rat | | E-mail Icon at

Julius Edgar Lilienfeld's *patents* show various construction methods. Naturally, patents do not have photos and almost never anything close to performance data. Check all of the patents by Shockley and find the same "lack".

First? The Lilienfeld patents start in the mid '20s..

Filing a patent application is a "constructive" reduction to practice. The Manual of Patent Examination Practice details the legal support for this:

But on the theme of patent validity: Has anyone tried to build a Lilienthal transistor from his patent? Patents have to be "enabling;" they have to contain sufficient information for an ordinarily skilled person to make one.

I'm getting good at finding the answers to my own questions: One Bret Crawford published his replication of the Lilienfeld transistor in his Masters thesis, in 1995.

Wasn't it customary in the really old days that the inventor had to actually bring his invention into the US Patent Office and demonstrate it?

I think so. I can recall seeing pictures of Ford sitting in the Patent Office with several of his engineers, with parts on their knees or on the floor in front of them ;-) ...Jim Thompson