Transistor Symbols

[G6Tanuki]

Spinning of from this thread about diode symbols https://www.vintage-radio.net/forum/...d.php?t=198464

I remember that some early published circuits used 'unconventional' representations of transistors.

One I recall was a horizontal filled rectangle [representing the base] with an arrow pointing to its middle for the emitter and a plain lead from its middle for the collector - see hand-drawn example below.

I guess this was intended as a logical representation of the construction used in some of the first junction transistors.

IBM used a somewhat similar three--boxes [though this time vertically] to represent junction transistors, see second example.

There were some equally strange representations used in a number of Eastern European publications, though I can't find any renditions of these onlune.

[wave solder]

These are some another early transistor symbols that I remember seeing

[TonyDuell]

I have read that originally the symbol we all now use for a transistor meant a point contact transistor and the symbol that G6Tanuki drew in his first picture was a junction transistor.

[G.Castle]

Marcus Scroggie was a champion of that alternative method of drafting the transistor symbol, and gave an explanation why in editions of the 'Foundations of Wireless', though mimicking only early alloy junction transistor construction, (think OC71 and similar).

All later plannar, etc., types are better served by the traditional symbol we see most often.

Turning components into blocks was just a method of simplified drafting IMHO.

[lesmw0sec]

Anyone recall that attempt to re-symbol TTL into 'intentional' logic instead of the usual symbols? It led to the ridiculous situation where a 7400 nand gate would be redrawn as a 'nand' or 'or' gate depending upon which function the circuit was achieving? (positive or negative logic).

[kalee20]

Originally Posted by TonyDuell

I have read that originally the symbol we all now use for a transistor meant a point contact transistor and the symbol that G6Tanuki drew in his first picture was a junction transistor.

It looks very much like a point-contact transistor - a base of material, with two whiskers pressed against it separated by a minute gap! (I've never tried to use a point-contact transistor, though I understand they have rather different characteristics to a junction transistor, in which case different symbols would be hugely helpful).

The first picture in wavesolder's post is to me, ideal for a junction transistor. Though I'd have difficulty changing from what I use now, even for myself, and I'd be out of sync with the rest of the world!

[G6Tanuki]

Yes the design in Wavesolder's post is very reminiscent of the construction of a bipolar junction transistor, the symbol we are all familar with these days with the emitter and collector on the same side of the base-bar brings to mind the construction of the old point-contact transistors which - as Kalee20 points out, had characteristics rather dissimilar to the modern bipolars [see attached transmitter circuit from a 1950s ARRL manual; I can't figure out how the xtal oscillator manages to work!] In many cases it seems that there was sufficient E/C leakage for them to work without the intervention of any specific base bias.

Considering drawings of the transistors in books/manuals/magazines, I remember that some of the early US renditions of the 'classic' bipolar transistor were reproduced in a sort-of reverse-image style, where the circle was filled with black or grey and the detail of the semiconductor was in white. 1930s valves [tubes] were sometimes shown like this too.

Alas I can't find any examples of this style of transistor-drawing at present....

I also remember some early transistors being drawn with, in addition to the usual E/B/C leads, a "substrate" connection, whose exact use I never fully understood.

[SiriusHardware]

...the ridiculous situation where a 7400 nand gate would be redrawn as a 'nand' or 'or' gate depending upon which function the circuit was achieving? (positive or negative logic).

Yes, this came up in a Commodore PET repair thread: Three of the gates in a four-gate IC were drawn conventionally, the fourth was drawn in a way which indicated how it was being used.

[emeritus]

In the almost 14 years I spent as a patent examiner in the UK Patent Office I came across many different transistor symbols. Before I left I started to make a collection of the many different types for the benefit of my successor, but never managed a complete set. I am away at present but can post a scan of what I collected next week. The symbol of #1 was still being used extensively by IBM at the time I left the PO in the late 1980's.

I was at Plessey in the 1970's when BS 3939 introduced the new rectangular block logic symbols. We engineers always used to draw our logic diagrams on paper and pencil using the old symbols, and eventually passed them to the drawing office when the designs had been finalised, who produced formal drawings using the new symbols according to BS 3939, thereby making them pretty incomprehensible to both us and the customer. So we supplied the customer with both the incomprehensible formal drawings as specified in the contract, and under-the-counter unofficial copies of our original drawings that used the old symbols that the customer could understand.

[Electronpusher0]

Originally Posted by SiriusHardware

Yes, this came up in a Commodore PET repair thread: Three of the gates in a four-gate IC were drawn conventionally, the fourth was drawn in a way which indicated how it was being used.

Using negative logic can be confusing either way, I am working on a very slow timebase I designed for an X-Y plotter, the pen lift function uses a nand as a negative logic "or". How does one indicate this on the schematic so that others (or my future self) will understand how it works?

Peter

Edit somewhat OT, perhaps discussion of logic symbols should be spun off as another thread?

[emeritus]

As long as the type of gate is specified, it should be possible to work out what is going on by drawing up a truth table.

[David Simpson]

A few years back I had thread out about my wee collection of rare Crystal Triodes.
This is what "Transfer Resistors"(Transistors) were called in those early post Bell Lab. years into the early 50's. They were point contact, and had a "T" profile for their connections, or other profiles which differ from the later common OC71's or AF117's & so on.
I recall posting an extract from my AVO Transistor Data Manual(1968) which shows the likes of the 267 LAN which is listed as no 81(T profile) on the listing sheet.
Incidentally, these & the likes of the GET1 need to be only tested in the common base mode.
If anyone would like me to scan the complete diagram pages, I would. I could add them to this thread, or send as an attachment to an email. Luckily(or un-luckily) the glue on my old TDM(over an inch thick) has loosened over many years & diagrams have detached & can easily be scanned.

Regards, David

[emeritus]

Here's the far from complete collection of transistor symbols I made in 1988, reflecting those in use at the time.. Mostly FETs as that was where most of the activity was at the time.

n.b. TDB =IBM Technical Disclosure Bulletin, a monthly publication about the size of a telephone directory that IBM used to publish, containing ideas that IBM didn't want to patent themselves, and so published them to prevent others from doing so.

[G6Tanuki]

I remember seeing the "Schottky Transistor" with the sort-of-Z extensions on the base.

And one of those shown for a FET reminds me of the things from times-past which had a 'substrate' connection.

[emeritus]

The substrate connection was sometimes referred to as a "back-gate", as by applying a "back-gate bias" voltage, you could adjust the threshold voltage.

[DrStrangelove]

Originally Posted by emeritus

The substrate connection was sometimes referred to as a "back-gate", as by applying a "back-gate bias" voltage, you could adjust the threshold voltage.

Well there's a thing, I'd thought the monolithic dual Jfets all had substrate connections and when I looked at the databook it turns out that they don't.

Though some have connections to the case.

My delusions are shattered.

The mosfets had disappeared from the databook by the time I joined in 1979, those do have substrate connections.

There's stuff in the 76 databook that I've never even heard of, much less seen & tested. Sic Transit & all that.

[Chris55000]

Can anyone suggest why these unusual transistor circuit symbols were employed in "Principles Of Electronics" by M.R. Gavin and J.E. Houldin (C.U.P., 1965!) – this is a British publication and it's thermionic–valve and other symbols are totally standard ones!

. . .They do mention the standard symbol as "point contact transistors", but by the time this book was published the standard symbols had been in use for over fifteen years, and it seems odd to introduce an unusual one that no other publication has ever used before or since!

. . .(I do remember there was some discussion, (some of it quite argumentative!) in the Hallowed Letters Pages of "Wireless World" Magazine in the late 1950s and early 1960s regarding transistor circuit symbols!)

Chris Williams

[G6Tanuki]

That's a new one on me! I wonder if it's meant to be stylistically like a little valve...? The arrow being a simulacrum of a cathode emitting electrons [or holes?] and the plate at the top being an anode?

[Vintage_RC]

Despite the odd symbols for junction transistors the symbols for point contact ones are fairly conventional.

[paulsherwin]

It took quite a while for modern standard transistor symbols to become universal, and early designers were very influenced by valve technology and the 'crystal triode' concept.