Talk:Transistor/Archive 1

Archive 1

Archive 2

Archive 3

Archive 4

Transistors are important

The invention of the transistor is supposed to have been extremely important. This article doesn't explain why! Also, what does "amplifying semiconductor device" mean? Is someone who needs a definition of "transistor" likely to understand that explanation?

Quite correct. I've made a start, but there's obviously much more to be said. Why don't you help? --Robert Merkel

Additions needed

Some terms to add/define here: Base-Emitter voltage (Vbe); Base-Collector voltage (Vbc); Collector-Emitter voltage (Vce); other stuff to distinguish: Vcc, Vco; Waveguy

I think this page really needs a "future of" section. This may even deserve a seperate page. Should mention potential replacements like quantum or DNA based computers, etc Iammaxus 19:48, 9 Nov 2003 (UTC)

Things that are wrong

There are some things just plain wrong in this article.

For example, the MOSFET can not be regarded as constructed from two back-to-back diodes between source and drain. There's no PN-junction between drain and source at all. There is a parasitic diode between the substrate (bulk) and the /drain, however. This is normally reverse-biased, so that it does not conduct.

Moreover, CMOS is not a transistor type, it is a technology to combine both n-channel and p-channel MOSFETs on the same chip.

The description of junction FETs (JFET) is missing entirely.

Stefan Heinzmann 25JAN2004

I agree with all that. Feel free to fix it yourself if you have time. -- Tim Starling 04:12, Jan 26, 2004 (UTC)

1. File:XcutMOSFET.png

Here is a MOSFET showing two back to back diodes, the source and drain n-doped regions and the substrate's p-doped region. Whether it should be described as such, I don't know.

2. BJTs and FETs have their own articles, in which things like JFETs get covered. This is an article about all things transistor, not too in depth about each thing.

3. But yes, CMOS is not a type of transistor. I will fix that bit. - Omegatron 21:36, Apr 6, 2004 (UTC)

I've put some stuff in to cover a few of the above points in 'Types' and 'Miniaturisation'. Wiki seems to be doing some stange things at the moment: edits are being lost and the Tran page keeps reverting to earlier issues or parts of and the History page is not up to date: perhaps the recent power failures have made it unwell. Has anyone else had similar problems with this or other pages? - CPES 19:01, 26 Feb 2005 (UTC)

Seems scrambled

Looking at this article, it seems a bit scrambled at the moment. Anyone working on cleaning this up? Gwimpey 00:41, 5 Jun 2004 (UTC)

Factoid on units of conductance and national rivalry in unit naming

This factoid was recently deleted by 128.12.178.70 on the ground of irrelevancy, and I agree. But it's an interestingly little note so I'm recording it here for the time being... this ought to go into some article somewhere, just not this one.

[Conductance is] (measured in Siemens -- formerly called the Mho in the states later changed because the Germans where getting short changed in the naming of fundemental quantities in electronics. Ohm was a Scott, Volta an Italian, Henry an American, Faraday, Watt (a thermodynamicist) were Englishman)

Should be added to siemens (unit) or Ohm (unit), if you ask me. - Omegatron

I've changed this link to the unit, not the disambig page. --Wtshymanski 03:32, 23 Apr 2005 (UTC)

Generally the edits by 24.12.159.248 are going to need some work to organize the information, move some of it to more directly relevant articles, and remove any more jokes he/she might add. I'm inclined to add this article, as well as capacitor and diode, to cleanup when he's done. Gazpacho aka 128.12.178.70

Yes, that is a good idea. Both articles ramble around quite a bit. - Omegatron

Origin of the word transistor

Did it come from transresistance (contraction of transfer resistance)? or did it come from trans:

trans- 3. Change; transfer: transliterate.

and resistor, since it is basically a current or voltage controlled variable resistor? I have heard both. - Omegatron 14:17, Jun 28, 2004 (UTC)

The OED2 says "Blend of TRANSFER and RESISTOR". -- Heron 18:16, 28 Jun 2004 (UTC)

The name 'Transistor' was selected by a ballot held at Bell Telephone Laboratories Incorporated. Here is a quotation from their Technical Memoranda, dated May 28, 1948 calling for votes:

' Transistor. This is an abbreviated combination of the words "transconductance" or "transfer", and "varistor". The device logically belongs in the varistor family, and has the transconductance or transfer impedance of a device having gain, so that this combination is descriptive.'

Other names suggested in the technical memoranda are, 'Semiconductor Triode', 'Solid Triode', 'Surface States Triode', 'Crystal Triode' and 'Iotatron'. (Is there a typo in the original memo: should 'Surface State Triode' have read 'Solid State Triode' ?)

I got this from a scanned image of the original memoranda at: 'Bob's Virtual Transistor Museum & History' at: http://users.arczip.com/rmcgarra2/index.html --- CPES 22:55, 14 Feb 2005 (UTC)

We should clarify that it is transconductance varistor and not transresistor or transfer resistor or whatever. The book I read it in claimed "transresistor" and said it was an ironic mistake on the part of the namers since BJTs are transconductors. I guess the book's editors (and the OED) made the mistake. We can get it right, though! - Omegatron 00:10, Feb 15, 2005 (UTC)

I have put in some words covering the above. Also moved the 'Origin of name' from the middle of 'History' to follow 'Intro', where it seems to fit better, because names are already mentioned there. It also seems more like a more general intro type topic. -- CPES 21:06, 17 Feb 2005 (UTC)

Big removed chunk

Someone removed this whole section:

The diode's job in AM detection is the rectification stage, which is followed by the integration stage in the archetypal scheme. Full-wave rectification is akin to the absolute value function in mathematics, but a single diode can do only half-wave rectification, where the signal is passed only if it is above (or below, depending on diode orientation) zero. Not any old crystal would do. It had to be 'naturally doped,' bearing impurities which would color the normally clear crystal.

The typical diode is called a PN junction diode, as it is an epitaxial arrangement of a P-carrier region and an N-carrier region. Biasing the diode in one direction tends to separate the charge carriers, turning it off. With the bias in the other direction, the carriers come together, allowing significant current to flow, provided a threshold potential requirement is met.

This potential, or 'bandgap,' is dependent on what type of material the diode is constructed of. Typically, approx 0.6 V will make a silicon diode seem like a resistance of approximately 1 ohm. When the electrons participate in this 'recombination current,' a fixed frequency of light is emitted, proportional to the energy (electrical potential) required to turn on the diode. This is called 'discrete emission,' like in a neon tube (which is also monochromatic). In silicon this frequency is in the infrared, which is too low to see. GaAs has a 1.2 volt drop, and can be used as a red LED. Germanium, which is fast but has lower reverse breakdown potential, has a typical drop of 0.3 V, even lower in the IR part of the spectrum.

A silicon Schottky diode also has about a 0.3 volt drop, half a typical silicon drop, plus a (much smaller) thermocouple-scaled drop, which is ignored. These devices are also fast, but similarly feature a lower reverse threshold. They are popular as protection circuits on the input of ICs as they turn on at a potential lower than ones that would damage the more expensive IC. They are also popular in mixer circuits, which are used in 'superheterodyne' type radios (by far the most popular topology). This is because of their low turn-on threshold, excellent matching, and low noise. Here they are used in a sub-threshold mode to achieve a logarithmic (multiplying) transfer function.

Should this be moved to the diode page? - Omegatron 13:30, Jul 6, 2004 (UTC)

FETs

I have added a considerable amount to the FET page, which is linked from MOSFET. My feeling is that it would be a good idea to try to move the detail on FET and BJT operation out of transistor and to FET and bipolar junction transistor. I think that rather than explain the detailed FET and BJT operation here, it would be better to conceptually explain transistors as switches and voltage-controlled current sources (e.g. for amplifiers). I think particularly that the FET stuff here is largely a scatter-brained assortment of thoughts on subjects of varying importance (some of which I know nothing about) that perhaps belong in other articles.

Some paragraphs seem randomly thrown in. Dubious accuracy, dubious relevance. Rmalloy 15:52, 15 Jul 2004 (UTC)

Voltage-controlled current sources are very confusing for some beginners. They are allowing a specific amount of current to pass through them from the voltage supplies; not creating current themselves. controlled resistors or even controlled current regulators would be a better description. - Omegatron 15:50, Jan 19, 2005 (UTC)

Relative speeds of p vs n transistors

An anonymous user changed this:

In silicon devices, the ones that use majority carriers that are electrons, rather than holes, are slightly faster and can carry more current than their P-type counterparts. The opposite is true in GaAs devices.

To this:

In silicon devices, the ones that use majority carriers that are electrons, rather than holes, are slightly faster and can carry more current than their P-type counterparts. The same is true in GaAs devices.

(emphasis added). The electron and hole mobilities in each material seem to agree with the latter: Si 1600 vs 430, GaAs 9200 vs 400 [1]. Does anyone have better information? -- Tim Starling 14:33, Oct 26, 2004 (UTC)

80.177.169.33

Thank you for your extensive writing on the transistor page. Please consider registering a user ID, since it helps us differentiate useful editors from useless ones. Also, please consider consolidating multiple edits into a smaller number. Hu 21:00, 2004 Nov 19 (UTC)

Hi Hu

Sorry if I caused a problem with so many small edits. I'm new to Wikipedia and just finding my way around.

Isn't it better, though, to have many small edits rather than one big one? The way I see it, this makes the edits more modular and simpler to control, especially reverts.

I haven't registered yet because I wanted to wait till I got a better understanding for the procedures on Wikipedia first.

What caused the recent tirade by 24.91.84.72 on the tran history page? Was it something I said? As far as I could see I only made some minor editorial changes and explained the reasons in the summary box. These changes were similar to those of other contributors. Is he a regular contibutor or just an interloper? Does that sort of thing happen often?

Thanks to PJacobi for removing the vandalism from 24.91.84.72.

Regards

80.177.169.33, 21:00 UK, 2004 Dec 23

remember that edits by anons are scrutinised FAR more closely than those by registerd users (because most vandalism is done by anons). Also ips are harder to remember than usernames and can change. Until you register you will always be somewhat of a visitor with no real reputation in the community. Plugwash 21:16, 23 Dec 2004 (UTC)

(now reg as CPES)

Still not getting it - how do they work?

In a bipolar junction transistor (BJT), an electrical current is fed into the base (B) and modulates the current flow between the other two terminals known as the emitter (E) and collector (C).

In field-effect transistors (FET)s, the three terminals are called gate (G), source (S) and drain (D) respectively, and it is the voltage applied to the gate terminal that modulates the current between source and drain.

Ok, I must be missing something here. How do you feed a current into the base? Where's the circut?

Likewise, how to you apply a voltage to the gate? Isn't voltage only meaningful across two points?

This is why I've never understood transistors. If someone can answer these questions, I think we can improve the article significantly from a "transistor-newbie's" point of view. --Spikey 19:38, Nov 26, 2004 (UTC)

I have added to the section on BJTs, which I hope will now make more sense to you. However, after doing that, I discovered that we have a detailed article on BJTs at bipolar junction transistor. D'oh! I think I am now going to have to chop out the entire BJT section and move it to the dedicated article, and do the necessary merging. While I'm doing that, I will probably also move the entire FET section to the field-effect transistor article. This will help to cut down the unwieldy length of this article. Any objections? --Heron 21:54, 26 Nov 2004 (UTC)

No objections, please do so! --Pjacobi 21:59, 26 Nov 2004 (UTC)

OK. I just did the BJT section. The FET section is next. --Heron 13:38, 27 Nov 2004 (UTC)

current into or out of the base is caused by the voltage across the base and the base's internal resistance. the voltage across the base is actually from the emitter to the base. yes voltage has to be between two points. so a voltage from the emitter to the base causes a small current to flow from the base to the emitter. this "opens up" the channel from emitter to collector and allows current from the supply to flow through it.

i wish there were a good mechanical analogy for this. there was a device in my car that worked similarly. it has a diaphragm that controls a plunger that stops up a large pipe. so a small vacuum at the diaphragm lifts up the diaphragm, which pulls the stopper out and allows a large flow through the other pipe. this is like an FET. the BJT version would have small holes in the diaphragm, so you need to constantly pull a small flow to keep the valve open. - Omegatron 15:56, Jan 19, 2005 (UTC)

This is an EGR valve. The vacuum line controls the flow of exhaust gas through another section of tube. The difference between BJTs and FETs can be thought of as the difference between an EGR valve with a complete diaphram and one with small holes in the diapgrahm. The one with small holes is like a BJT, because it needs a constant small flow of current to keep it open. The difference between N-type and P-type can be compared to the difference between negative and positive backpressure EGR valves. One requires pressure and one requires vacuum. http://shbox.com/1/EGR_valve.jpg - Omegatron 04:12, Mar 2, 2005 (UTC)

Critical history missing.

What would be the best way of incorporating the following critical historical perspective into the Wikipedia writeup on the transistor?

According to witnesses at a lecture given by John Bardeen at the University of Illinois, Altgeld Hall circa 1978, when Bardeen thought he was in failing health, Bardeen stated that the trio of inventors were ordered to stop work on the transistor by Bell Labs management and had to ecome insubordinate to continue work. Quoting Sherwin Gooch, one of the attendees to this lecture at Altgeld Hall:

John Bardeen himself gave a talk one evening at Altgeld Hall on the University of Illinois campus, circa 1978, in which he related various experiences surrounding his inventing the transistor. At the time, people suspected that the scheduling of this presentation may have been related to Bardeen's health.

Professor Bardeen showed us the B&W 16mm film BB&S had made at Bell Labs immediately after they got the first transistor to work (and, presumably, before Bardeen's boss got to work the next morning...) I have seen individual frames and out-takes of this film since, but I don't know if the entire film still exists. The "rolly-cart" with their experimental set-up is plainly in evidence on the film.

It was John Bardeen himself, at Altgeld Hall, who related that his boss had said that the "solid-state amplifying device" which they wanted to develop was "not feasible," and that, "even if it were possible, it would have no practical application." Dr. Bardeen related that sometimes, when his boss stayed at work past 5 p.m., the three of them would become very impatient waiting for him to leave so they could roll their setup out of the coat-closet, and get busy on what they, apparently, thought was the greatest "cool hack" of the day.

I wonder who Bardeen's boss was. His boss should be immortalized in history next to the NASA manager who advised the last engineer withholding approval of the Challenger launch to "put on your management hat!"

Jim Bowery 20:25, 28 Dec 2004 (UTC)

Wikipedia does not publish original research, which includes first-person accounts and novel claims published only on the web site of a particular individual. This is not a reflection on the merit of that material — Wikipedia is simply not the proper venue. We have to stick to material already well-established in traditionally reviewed publications, so the best thing for the author of such material is to publish it in a respected print journal and then we can reference it. —Steven G. Johnson 01:04, Dec 29, 2004 (UTC)

Fair enough. This would indeed require obtaining primary documents and creating a credible report based thereon. Jim Bowery 16:55, 29 Dec 2004 (UTC)

billion transistors

Anyone got a reference to billion transistor chips? Pending that, I have "downsized" that paragraph to quarter-billion. I'm not sure on exact numbers. More importantly, I've removed the DRAM reference from here, because DRAM's have nowhere near as many transistors as that. DRAMs use capacitors to store charge (approx 1 billion per chip these days) and only use transistors to boost this (and I would guess with only a few hundred thousand transistors). Mat-C 21:28, 5 Jan 2005 (UTC)

Here's one:

In fact, 1 billion transistors has already been reached in high-capacity DRAM memories. One gigabit is conventionally 1,073,741,824 bits, and each bit is stored using a transistor and a capacitor. However, the more demanding and prestigious milestone will come when a billion-transistor digital logic chip enters volume production. -- http://nwc.serverpipeline.com/technology/23900954

If you need convincing that a DRAM cell consists of a capacitor and a MOSFET, see http://www.ece.umd.edu/courses/enee759h.S2003/lectures/Lecture2.pdf (sorry, it's a PDF, but it has lots of nice pictures). Or you could read our article on DRAM. --Heron 09:01, 6 Jan 2005 (UTC)

Transistor operation

I think it's useful to think of a transistor as a voltage or current-controlled resistor. It's not that inaccurate. I didn't understand them until I thought of it this way. "Amplifier" or "voltage-controlled current source" are quite confusing for a newcomer. - Omegatron 01:48, Feb 7, 2005 (UTC)

Hi Omegatron. Thanks for comment. I take your point. Stangely, though, when I was learning about transistors I found the concept of thinking of a transitor as transferring a resistance misleading. It wasn't till I realised that the current in the collector is controlled by the current in the base/emmitter that things became clearer. The worring point to me about tranferring resistance is which resistance is transferred to where? The very nature of a transistor is that the collector current, to a first approximation, is constant, and is effectively independent of collector voltage, unlike a resistor. CPES 11:52, 7 Feb 2005 (UTC)

Yes, I know it's not accurate for calculation purposes, but it's a lot better for complete beginners to think of it as a controlled switch first, and then as a controlled variable resistor, and then they can learn the exact behavior. Telling them its an amplifier or current source will confuse them a lot, trying to figure out how a transistor generates current. I think the term "voltage- or current-controlled current regulator" might be the best? We should be discussing this on the transistor talk page, probably. - Omegatron 15:35, Feb 7, 2005 (UTC)

You're right: a complete beginner or even someone just getting an overview of transistors may have a problem with current transfers and how that would work. Also, later the article does expand on transistor operation so anyone wanting to dig a bit deeper can get a more detailed explanation. I felt so strongly about 'transferring resistance' that I originally put in a statement that the concept of transferring resistance would not be helpful in understanding how a transistor works. This was later removed. When I was learning about these odd new devices, way back when tubes (valves) were king, we went through 'h' parameters and all that stuff, but like some of the above contributers I didn't really understand the basic operation and the DC conditions, especially as we normally focused on the grounded base configuration. Your term 'current controlled current regulator' sounds good (no mention of resistance). I hope there is no mixing this up with PSU regulators though. Would you like to update the Tran section? If you would like me to do it let me know. Regards --CPES 17:10, 7 Feb 2005 (UTC)

well, is there a better word than "regulator"? we need to convey that it allows a certain amount of current (without "causing" that current. it's just allowing the power supply to cause the current). regulator might get confused with a filtered power supply or something. - Omegatron 18:26, Feb 7, 2005 (UTC)

current "gate" "valve"?

I can't think of a good noun phrase to describe the BJT. Also there is the complication that both BJT and FETs are covered on the Tran page. This is part of the problem with the existing introductory definition. The best I can come up with is to have separate descriptions for the two types of tran: BJT = 'device where the output current flow is controlled by the input current' and FET = 'device where the output current flow is controlled by the input voltage'. They are not the most elegant descriptions but at least they get away from any implication that current is generated or caused; it is merely controlled. It could then be explained that in each case the output current comes from the power supply. The existing detailed description about terminals etc could then be moved to the 'How do they work' section'. - CPES 19:44, 7 Feb 2005 (UTC)

Yeah. Hence the "current- or voltage-controlled". The current existing description is accurate, but doesn't convey "current wants to flow because of the pressure/voltage from the power supply, and the transistor acts as a variable valve to allow a precise amount of current to flow, based on its input current (BJT) or voltage (FET)" just saying the current is "controlled" isn't quite enough. It's "allowed"? - Omegatron 20:24, Feb 7, 2005 (UTC)

Could say a simple way to think of it is a switch, which is accurate; they are used as switches all the time, and then move on to a "switch that varies smoothly from on to off"? - Omegatron 20:25, Feb 7, 2005 (UTC)

Your first definition captures the essence of transistors operation well. Unfortunately, 'valve' can mean vacuum tube in the UK, but this shouldn't be a big problem. Using your words, how about: " A transistor acts as a variable valve which, based on its input current (BJT) or input voltage (FET), allows a precise amount of current to flow through it from its voltage supply. - CPES 22:46, 7 Feb 2005 (UTC)

That's very good. I put it in. - Omegatron 14:26, Feb 8, 2005 (UTC)

I still want to add resistance though.  :-)

"It acts as a variable valve which, based on its input current (BJT) or input voltage (FET), allows a precise amount of current to flow through it from the circuit's voltage supply, effectively by varying its output resistance." - Omegatron 14:34, Feb 8, 2005 (UTC)

Your intro reads pretty good to me as it is, but if you think that a mention of resistance is important: go ahead. At least it will say 'effectively'. Regards CPES 15:25, 8 Feb 2005 (UTC)

But but but... BJTs are voltage controlled devices! There is no physical mechanism that lets the Ib directly exert control of Ic. Instead, the Vbe is the main input signal, and the value for Vbe controls both the small base current and the large collector current. Sure, if you want a simplified rule-of-thumb mental model, it's best to imagine that Ib controls Ic directly. But if we want a physics explanation, we must point out that Vbe controls the width of the BE depletion layer... and this depletion layer is what controls Ie. Ib is then determined by the shapes and doping of the B and E regions (i.e. how much of carrier population coming from Emitter will be intercepted by opposite carriers in the Base.) --Wjbeaty 20:32, Mar 20, 2005 (UTC)

your point is an interesting science VS engineering type problem.

the physisist would say: when we vary the base voltage it controls both base and collector currents.

the electronic engineer would say: when i change the base current the collector current changes, the change in the base voltage is negligable.

BOTH are correct they are just approaching from a different POV. Plugwash 03:22, 21 Mar 2005 (UTC)

Both aren't correct. Sure, Ic=Ib*hfe describes how a transistor behaves, but if we wish to explain *how* the base current controls the collector current, we have to come clean and admit that Ib cannot control Ic. Instead, transistors rely on small changes in Vbe. They're not as insignificant as they seem, instead they're the whole basis of transistor operation. Most textbook authors don't seem to know this, and most BJT explanations end up as misguided attempts to show how Ib controls Ic.

(And as a double-E myself, I'm very familiar with the engineer POV which says "I don't want to know how it actually works, just give me the SPICE model.") grin. --Wjbeaty 05:07, Mar 22, 2005 (UTC)

Invention/history

Suggest that the 'Invention' section would fit well in the 'History' section. At the moment it just seems to pop up in the middle of the physical/technical stuff. Or would 'Invention' be better after or even before 'Origin of name' ? -- CPES 01:24, 18 Feb 2005 (UTC)

I have moved title 'Invention' to be below 'Origin of name'. Also title 'How transistor works' to above 'How semiconductors work' where it seems to fit logically. -- CPES 20:46, 18 Feb 2005 (UTC)

Now, I have moved 'Invention' to be above 'Origin of name' It seems to fit better there: first Bell invented tran and then they needed a name for it. (Wiki has sorted itself out now and is firing on all cylinders) - CPES 01:31, 27 Feb 2005 (UTC)

Images

I moved the SMD image to be with 'Types' where it is covered by the text. This complies with Wiki style (image size clash) and also gives a bit more 'colour' to the text lower down.

The 3 current images of transistors (thru hole, smd & 2* power on h/s) look good, but some some humans would be nice: there is a good image of the three people tran developmet team on the net which would fit in well with the 'Invention' or 'History' sections. The standard image of the first tran would be good too, say with 'History'. On the technical side, the tran elec symbols are missing. Suggest a single graphic image of: npn bjt, pnp bjt, p chan igfet, n chan igfet, p chan igfet, n chan igfet (both igfets without parasitic s/d diodes). Also suggest that the symbols be in line, stretching across the width of the page. Where to put it though: perhaps under 'Types' and move smd image to 'Minaturisation'. Although I can generate the tran symbols image, as a Wikirookie I feel a bit nervous about uploading images, especially as Wiki is a bit fractious at the moment. If ok I may have a go when things settle down. - CPES 20:09, 26 Feb 2005 (UTC)

Humans? What are those?  :-)

I got the schematic images uploaded already I think. I'll link 'em. - Omegatron 01:23, Feb 27, 2005 (UTC)

I'm not sure what they are. I think there are two types of humans: P type and N type, but I never worked how you can tell the difference- I'm just a techi. Look forward to seeing the tran symbols though. (the system seems to have caught up with events now) - CPES 01:52, 27 Feb 2005 (UTC)

I didn't; I was thinking of wikibooks, but I made a new one:

I'd like to make a single picture that has EVERY possible symbol so people can figure out what they are looking at easily. Or else make one for each kind and table them together for easier editing? I can arrange them however we want, flip them, etc. These already exist:

File:IconsMOSFET.png File:IconsBJT.png

And these on the french wiki:

fr:Image:Transistor_mosfet.png fr:Image:Transistor npn.png

And probably should include all of these:

http://www.silisyum.net/pic/fet/mosfet_sembolleri.gif http://ourworld.compuserve.com/homepages/g_knott/mosfet2.gif http://ourworld.compuserve.com/homepages/g_knott/elect248.htm

Separate images for each family would be best, I guess. Then we can arrange them like I did on the wikibook: wikibooks:Electronics:Symbols - Omegatron 01:57, Feb 27, 2005 (UTC)

The symbols look good on the tran page. I don't like to be picky, but suggest the sybols for the IGFETs should be the ones now used in the U.S. and U.K. as in the International Rectifier FET data book for example. I can't draw them on Wiki at the moment but they only have three connections (like the physical IGFET) and they have a 'bent gate line' with the connection comming out of the circle adjacent to the source. Also three short lines on the D/S side, the source being internally connected to the two lower short lines and the drain being connected to the upper short line. Be nice if they could be added to the present image on the tran page. I like the way you have kept the symbols organised so that all poitive terminals are at the top. - CPES 02:26, 27 Feb 2005 (UTC)

I meant I was going to include every possible symbol, including more than one symbol for each type if there is more than one, so that we can serve as a reference for all the confused engineers who don't know what all the little broken lines mean. Can you find the image you mean on google images? - Omegatron 02:49, Feb 27, 2005 (UTC)

"I like the way you have kept the symbols organised so that all poitive terminals are at the top."

Did I? So BJTs are flipped to stay the same polarity, but JFETS have the drain and source on the same side regardless of polarity? - Omegatron 17:18, Feb 27, 2005 (UTC)

This is more or less it fr:image:050512.gif except there is a circle around it and the arrow bit internally connects inside the circle to the bottom line on right which is then the source terminal. - CPES 03:18, 27 Feb 2005 (UTC)

agree that we should include ALL symbols for a transistor type. That link on fr is broken. im SURE that it is technically possible to have a seperate bulk connection on an igfet though few practical discreet devices do.

(4am local + pressure from high power n type human = errors) Thanks to Plugwash for sorting out. I'm not with images yet. Agree that substrate can be brought out of case as a separate terminal but, as you say, on nearly all igfets the substrate and drain are internally connected. Therefore, to keep it simple suggest we show just the two symbols with sub/drain connected and perhaps explain in text about independant substrate and use same approach for parasitic diode betweem s/d. There has been a lot of confusion with various igfet symbols in the past and I feel that if we show too many symbols it might cause confusion rather than help. This is particularly true when the symbol itself doesn't show graphically any difference between the drain and source. I will see if I can get the 'good' IGFET symbols to you in some way. - CPES 11:43, 27 Feb 2005 (UTC)

Here they are: http://www.irf.com/product-info/datasheets/data/irf520ns.pdf (N Chan), http://www.irf.com/product-info/datasheets/data/irf5210s.pdf (P Chan) but both without internal diodes. - CPES 15:40, 27 Feb 2005 (UTC)

Well we don't want to get too in-depth explaining details about each type in the main article. We'll leave that for the smaller articles. But I am imagining a section called "schematic symbols" or something, and it will have a table with all of the symbols in it and clickable captions underneath each that go to the appropriate articles. I think showing more symbols is better than less. - Omegatron 17:18, Feb 27, 2005 (UTC)

"Did I? So BJTs are flipped to stay the same polarity, but JFETS have the drain and source on the same side regardless of polarity?" NO, I just noticed: the P chan JFET symbol is the 'right' way up but the source and drain idents on image are wrong: they are swapped. Arraagh, also noticed another point: suggest that P chan JFET should be at top of image to be opposite PNP BJT. This would be more consistent and less confusing; the P chan JFET being the same 'sex' as the PNP BJT. - CPES 00:38, 2 Mar 2005 (UTC)

Excellent! That is the info I want. Yes, I am supposed to know this stuff, but I forget a lot.  :-) To be fixed in the next few hours... - Omegatron 01:37, Mar 2, 2005 (UTC)

Can you think of any reason why one should be on the top and one on the bottom?

generally i would put the pnp/p channel at the top. This is what i have seen in books and tends to correspond with most switch and amplifier (but not follower) usage where a npn/n channel would connect to the negative rail and a pnp to the positive rail.Plugwash 01:59, 2 Mar 2005 (UTC)

Ah! A perfectly good reason! Arrange it the way it is arranged in a CMOS inverter.

I'd like you all to look at my modular electronics diagrams proposal. - Omegatron 02:09, Mar 2, 2005 (UTC)

(This comment is out of sequence; It was blocked by the entry of above) Agree that having a separate page with a load of symbols would be the best so the main page is not too busy. That way we could put in symbols for darlington, IGBT, arrays, etc. But to me, also keeping a simple image showing the basic 6 symbols on the tran page would still be best. But, I really would like to see the more modern IGFET sybols, as on IR data sheets, put on the main page. We could show the other IGFET symbols on the new page (and perhaps warn that if you use them you are liable to cause confusion on your circuit diagrams. We could also say that if you are reading an existing circuit diagram with these symbols, you will just have to work out for yourself which end is the sorce and which end the drain). - CPES 02:28, 2 Mar 2005 (UTC)

Re: P channel being reversed. Are you sure? http://www.onr.navy.mil/sci_tech/information/312_electronics/ncsr/images/d_jfet_p.gif - Omegatron 02:34, Mar 2, 2005 (UTC)

refresh cache - Omegatron 02:41, Mar 2, 2005 (UTC)

It's ok now, but it was wrong on my screen- honest - CPES 13:17, 2 Mar 2005 (UTC)

I just meant "i uploaded a new version. don't forget to refresh your cache or you won't see it." - Omegatron 15:39, Mar 2, 2005 (UTC)

Now it's reversed again. I tried 'refreshing' (MS Internet Explorer) N times. Also re boot. ? - CPES 00:59, 3 Mar 2005 (UTC)

Hmm? I see PNP/P-chan. on top and EBC CBE | SGD DGS from top to bottom. You don't? - Omegatron 03:59, Mar 3, 2005 (UTC)

Still reversed (my machine should be ok = Dell Dimension 4400, 20gb disk free, 750 mbyte memory. Netgear DG814 DSL Modem Internet Gateway. Windows. XP & IE are fully up to date. Havn't had problems on other sites). The tran page does seem to do some odd things: often old parts of tran page will appear in place of current, especially when edits have recently been implemented. Also a block of recent edit records will disappear from history page (I put this down to Wiki data base upating). ? - CPES 11:57, 3 Mar 2005 (UTC)

...... and now they are correct! (look good too) - CPES 12:03, 3 Mar 2005 (UTC)

I replaced it with 4 images and a table, in keeping with my new philosophy of modularity and uploading things to the commons:

	PNP		P-channel
	NPN		N-channel
BJT		JFET

Now each individual image can be used in other articles in the same way, like this, for instance:

	PNP
	NPN
BJTs

MOSFET images

I need to replace this image:

File:IconsMOSFET.png

I intend to make several small images and do the table thing like above. But I don't know what these images are.  :-) There are depletion mode and enhancement mode, correct? And there are also these kind:

File:Emfetn3.png

Please help! Show me examples of each and I will make them. - Omegatron 17:45, Mar 12, 2005 (UTC)

Will look into it and get back to you - CPES 22:54, 18 Mar 2005 (UTC)

File:Transistor symbols (02).png

Transistor symbols

Complete set above - CPES 15:38, 21 Mar 2005 (UTC)

Excellent. What about the fourth (substrate) connections? - Omegatron 17:21, Mar 21, 2005 (UTC)

Perhaps I will show the images above, but have one example of, say, a p channel IGFET depletion mode with and without the substrate connection, and leave the rest with three pins... - Omegatron 20:42, Mar 21, 2005 (UTC)

Suggest that we keep it simple and should just show the types in the above image as these are the types normally used. It makes a consistent set. Otherwise we may cause confusion, especially for those finding their way around. We could explain in main text that in a few cases the substrate is brought out as a separate connection. This applies to all the types shown. Many MSI/LSI chips also haver a separate substrate connection. We could also say in text that circle around symbols is optional (at one time no circle indicated that the tran was embodied included in a chip and a half arrow showed that the tran was Si rather than Ge, but that has all gone now) - CPES 00:19, 22 Mar 2005 (UTC)

Alright. I'll start drawing. Definitely add the info about the circles and arrows. I only knew that they were optional. - Omegatron 01:36, Mar 22, 2005 (UTC)

Please double-check the polarity of the JFETs.

1. Is the more positive pin at the top?
2. Are the source and drain labeled correctly?
3. Is the "offsetness" in the correct direction?

All 8 of them

	PNP				P-channel
	NPN				N-channel
BJT		JFET	IGFET enh	IGFET dep

Do you like these symbols? They're kind of small on my screen to show this kind of detail...

Please double-check *every* detail. I know something here is wrong... - Omegatron 02:13, Mar 22, 2005 (UTC)

The symbols look good. Just one thing though, on my screen the P JFET idents for S & D are swapped (hope this isn't my old update problem returning).

Yeah. That's the way they were previously and everyone said it was correc. This is why I say to double-check. Are you absolutely positive that only the JFET p-channel should be switched?  :-) I see both directions online, so it's a common error... - Omegatron

Yup On my screen the PJFET has D at top and S at bottom which is wrong. The other P devices have S at top and D at bottom which is ok. Do you see PJFET D at top on your screen? - CPES 17:12, 23 Mar 2005 (UTC)

Yep. That's the way it is on my screen, as well. Is the asymmetry in the correct direction, though? The gate should be closer to the source? Note that I have seen the opposite online. If you press ctrl+F5 or ctrl+refresh or something like that you should now see the fixed version. - Omegatron 17:30, Mar 23, 2005 (UTC)

Thanks that has fixed it. Symbols all correct now. - CPES 17:41, 23 Mar 2005 (UTC)

The transistor symbols are slightly small on my screen (17" 1280 by 1024 pix). Suggest the following minor points: Is it possible to make the vertical line on the BJT slightly thicker than the rest of the lines and pro rata for the other symbols (as my image)?

Yes. - Omegatron

Could you change text 'N-channel' & 'P-channel to 'N Channel' & 'P Channel'?

YOU can change that. That's the whole point of making the images modular... I think N-channel is better english, though. - Omegatron

Could you make the arrows all the same: on my screen the BJT arrows look different to the FET arrows? On BJT the arrows are normally shown at ends of line.

Sure. - Omegatron

Could you change 'enh' & 'dep' to 'En Mode' & 'Dep Mode' (perhaps on second line), just to make it clear what is meant?

Again, you can change the text as easily as I.

One final request, rather than present ident scheme, would you consider showing individual, self-contained text idents for each device (as my image)? That way the symbols could be equally spaced which to my eye looks better. Also it would make each symbol self contained. ........

You can do this, too.  :-) It's just a table with several images inside it. That way you can use each individual image separately, modify the text/format, etc. - Omegatron

Also could you render the symbols on a background of grazing widebeests with an orange setting sun and perhaps a lion stalking them in the brush? - CPES 14:25, 23 Mar 2005 (UTC)

Sorry, but I can't set a background image for the table using a WP image. See the approximation below, though.  :-D - Omegatron

I just noticed, the images above that you say 'need to replace' look about the right size on my screen. - CPES 15:40, 23 Mar 2005 (UTC)

Yeah; on mine too. I was basing these off the klunky schematic editor and my idea for completely modular electronics schematics, which I have now decided is stupid. In the (far) future we will have SVG rendering, making my proposal a waste of time. So I can completely redraw all of these in a larger (antialiased) format if everyone thinks they need it. I agree your uploaded image is a good size. For now, make sure these are factually correct, and I will make them prettier some other day... - Omegatron 16:13, Mar 23, 2005 (UTC)

	PNP	File:SOHO solar flare sun large 20031026 0119 eit 304.png			Watch out for that lion!-channel
	NPN				Wildebeest-channel
BJT		BEESTFET	IGFET enh	IGFET dep

Ho! Ho! Ho! excellent. Glad you have a sense of humour and can put up with my rantings (thanks for above info). But would it be possible to have a male lion facing the other wa ######### - CPES 17:12, 23 Mar 2005 (UTC)

Here is the only wikipedia image I see of a male lion facing the other way: Image:Lion sex.jpg - Omegatron 17:25, Mar 23, 2005 (UTC)

That's the one to use: add a bit of action to the tran page - CPES 17:41, 23 Mar 2005 (UTC)

New "Transistor Talk" topics should be entered at Talk:Transistor - CPES 21:43, 27 Apr 2005 (UTC)