Talk:Vacuum tube/Archive 2

Archive 1

Archive 2

Archive 3

Archive 4

Resources

I've come across a gold mine of resources for electron tubes that all will be interested in:

General electron tube books (mostly, some others mixed in): http://www.pmillett.com/tecnical_books_online.htm Electron tube data scans: http://www.pmillett.com/tube_data.htm

All the copyrights have expired on all these scans and links are provided to show this. The most useful book is one by RCA in 1962, "Electron Tube Design", which has an unbelievable amount of information and diagrams on 951 pages - which ought to be transfered here! I've gone ahead and changed the format to DJVU for upload to wikimedia commons:

http://commons.wikimedia.org/wiki/File:RCA_Electron_Tube_Design_1962.djvu

Could this be linked to the main page? Tevonic (talk) 14:44, 2 June 2009 (UTC)

It can be referenced but I did notice that the first of the sources you quoted is a dead link. If there are any current links which can be used in the article as a source of data relevant to this article than go ahead. Sunshine Warrior04 (talk) 05:31, 2 November 2011 (UTC)

Beam Tetrode

The part of the page explaining the beam tube (on this page and the beam tube page) appears to conflict with this explanation:

http://www.r-type.org/static/btet.htm

which notes that the plates were only necessary due to the rectangular structure of the cathodes of the time. Eventually the plates were dropped as new manufacturing techniques were adopted. It notes that the defining chracteristics were the optically aligned grid and screen, and the creation of a virtual cathode between the screen and the anode. Rmed (talk) 06:57, 19 April 2009 (UTC)

An old discussion now, but the development of beam plates came after the Pentode was introduced. Beam plates were not and still are not peculiar to valves with rectangular cathodes. As the referenced article notes, there are 2 independant developments to the beam tetrode, the optically aligned grids and the 'beam plates' (as noted it isn'r actually the beam plates that form the beams, but that's what they are called. All beam tetrodes (to my knowledge) use optically aligned grids. Some pentode designs have also featured optically aligned first and second grids, but the advantage is spoilt by the supplessor grid being a much coarser pitch not being optically aligned (it would also have pushed the cost up to align 3 grids). 86.183.175.94 (talk) 15:31, 13 April 2011 (UTC)

British vs American usage

It is certainly true that the term 'valve' predominates in the United Kingdom, and 'tube' in the United States, but I do not find support for the statement that 'the rest of the world' use the term valve. Both appear to be in common usage, but not in equal numbers, nor close to even.

A quick and dirty use of Google yields these figures: "Valve Amplifier" yields 4120 hits, "Tube Electronics" yields 45100 hits. "Valve Amplifier" yields 69500 hits, and "tube amplifier" yields 445000 hits. All languages included in the search, and exact term restriction.

This article appears to be blatantly ethnocentric towards a viewpoint that valve is the most common usage world-wide. It obviously not. The article should at the very least remove the 'British whole world' vs 'American english' content out and become more neutral.

If someone can prove the claim it is more common, then use it and cite it. If valve was as common as tube, there would not be an order of magnitude difference in the Google hits. Evidence to the contrary?--70.122.69.185 (talk) 10:04, 28 July 2008 (UTC)

Supporting (and redundant) evidence - as of this writing a Google search of "tube amplifier" yielded about 409,000 results and "valve amplifier" about 63,300; a ratio of ~6.46 to 1.00. "Tube electronics" yielded about 39,900 results and "valve electronics" about 3,890; a ratio of ~10.3 to 1.00. I only pursued this because I thought the "valve electronics" numbers (which 70.122.69.185 quite harmlessly noted as "valve amplifier") looked rather low. Turns out they are that low. (Is Google US-biased? :-)

Cheers, Rico402 (talk) 10:26, 29 July 2008 (UTC)

I challenge the theory that valve for a vacuum tube is only British English. Every electric guitarist in the world knows what valve tone means. Nobody calls it tube tone, even in American magazines. So it's a bit more complicated than that! Andrewa 8 July 2005 20:47 (UTC)

Well... I own an electric guitar, but have never heard the term "valve tone" before. Heh. But anyway, I would think the term is pretty community-specific. - mako 04:28, 10 July 2005 (UTC)

I stand corrected. Every guitarist I've spoken to before this has used the terms interchangeably, eg describing vintage Marshall amplifiers (British) as being all tube and Fenders (US) as having valve tone. But my sample space isn't all that big, and of course is biased towards Australia, which is probably midway between the two linguistically. Which is where the wiki comes in! Thanks for your contribution. Andrewa 00:55, 12 July 2005 (UTC)

Comparing google results in the way some people have done above indicates nothing about world usage. The United States of course have a massive presence on the internet, whereas countries such as India where English is widely spoken have a much smaller presence and yet use predominantly "British" terminology. Yes, google is bound to have a natural American bias. —Preceding unsigned comment added by 88.105.125.254 (talk) 10:11, 18 November 2008 (UTC)

In response to the alleged google bias; maybe it's simply a matter of a greater population making more websites than a smaller population group? —Preceding unsigned comment added by 58.174.216.70 (talk) 15:41, 2 June 2009 (UTC)

The great majority of internet users are American, I daresay this is why tube is more popular online than valve. Large swathes of the world - far more than the US population - have very little online presence. So while clearly tube is more popular online, I don't think this can give a realistic prediction of how things are worldwide. Tabby (talk) 16:27, 18 June 2011 (UTC)

Proposed Infobox

Vacuum tube/Archive 2
RCA 12AX7
Classification	Dual Triode
Service	Class A amplifier
Height	2.2 in (56 mm)
Diameter	0.8 in (20 mm)
Cathode
Cathode type	Indirectly heated
Heater voltage	Series / Parallel 12.6 / 6.3
Heater current	Series / Parallel 150 / 300 mA
Anode
Max dissipation Watts	1.2 per section
Max voltage	330
Socket connections
Typical class-A amplifier operation
Amplification factor	100
Anode voltage	100
Anode current	0.5 mA
Bias voltage	-1
Anode resistance	80000

I have constructed an Infobox for vacuum tubes:

http://en.wikipedia.org/wiki/Template:Infobox_Vacuum_tube

Please review and assist with its development.

Armstrong1113149 (talk) 04:47, 22 January 2009 (UTC)

Comment: vector drawings (.svg) are generally better than bit-mapped drawings (.png). --Gerry Ashton (talk) 05:12, 22 January 2009 (UTC)

Agreed, does anyone have a better drawing? The existing image from 12AX7 was used... Armstrong1113149 (talk) 02:31, 23 January 2009 (UTC)

Continued to work on the vacuum tube Infobox. Added a a "typical operation" section for Class C and Class AB. Hopefully this makes it generic enough for transmitting as well as small signal tubes. Armstrong1113149 (talk) 02:31, 23 January 2009 (UTC)

The infobox has been applied to several pages including the 12AX7 and 6L6. To date I have not received any comments. Armstrong1113149 (talk) 22:00, 8 February 2009 (UTC)

I agree, it'd be a good idea to include an infobox. Nick carson (talk) 07:19, 30 June 2009 (UTC)

I see no units on 'amplification factor'. If its the usual mA/V its exceptionally high. Tabby (talk) 16:31, 18 June 2011 (UTC)

Cathode and anodes

I'd like to think that I know the difference... but not always. [1]--Wtshymanski (talk) 19:29, 28 March 2011 (UTC)

Sure, we know, and it's sort of natural that when you're concentrating on the wording and organization, it's easy to overlook an extreme mistake in the actual content. AND visa-versa. So thanks for improving the wording and flow of the text. Every time I write something technical to correct and supplement the content, it comes out too lengthy (especially for the lede) and can be much improved after a few iterations, as we saw in this case. The nice thing about WP is that silly mistakes like that WILL get promptly corrected by another editor, so long as there is more than one knowledgable person watching the page. (But is there any mechanism to insure that there is?) Interferometrist (talk) 13:46, 29 March 2011 (UTC)

Well, professional writers need to get paid for writing, so that cuts them out (unless they are moonlighting to keep their typing fingers nimble). Its up to us amateurs. Legend has it that there are Wikiprojects intended to attract interested editors with the goal of improving related articles. I'd like to believe in such things, I really would...but it's hard to muster evidence that such projects, aside from banner posting, actually exist. I once proposed a meta-Wiki-project "Banners" to make sure every talk page has a project banner on it. The developers could make this the default on opening a new talk page. This would relieve the pressure of the banner-posters from decorating empty talk pages and let them concentrate on article improvement, instead. Wordyness is one of the banes of Wikipedia and technical writing in general; why is it easier to write 20 words instead of 12? Or 5? A paradox. --Wtshymanski (talk) 18:24, 29 March 2011 (UTC)

Well I found out long ago that adding words is MUCH easier than removing them without losing content (think about it!). I feel silly now having spent a lot of time on this article, and it still has mounds of redundancy. Such as the sections on Heat Transfer and Cooling which I don't even want to touch. And it's worse than if I had written something myself and need to trim it down, because I'm reluctant to just wipe out someone else's work, not just because I might piss them off (if they're still watching the page!) but because I just might not have appreciated the importance of what they had written. So this article is extremely long, and I finally realize I'm not going to be the one to make it only as long as it needs to be :-( As far as how WP assigns tasks, I'm too new to appreciate what you're saying (though I can certainly see the problems you're refering to!).

On another tiny matter of content, isn't the heater of a tube ALSO referred to as a filament? I understand the distinction, but had always thought (and you'll find this in casual conversation) that you CAN also call it a filament without anyone caring whether it's also the cathode. Anyway I had put "filament" in parenthesis inside that caption and thought it could only help someone understand it if they hadn't read the fine print about indirect heating (which I think was later in the article). If you agree, why don't you restore it. :-) Interferometrist (talk) 22:41, 29 March 2011 (UTC)

Whoops: looking at the page I noticed that my previous version (including "filament") had been restored, almost as if by magic! I figured out though that it's because I had taken a copy of that section offline before you edited it and pasted my version back in (I hadn't thought anyone else would be editing it!), so it really was NOT intentional. I guess I can ask the opposite question now: do you think that it's alright, or want to again remove that? It's not a big deal, just one word Interferometrist (talk) 23:12, 29 March 2011 (UTC)

The terms 'filament' and 'heater' are not interchangeable. They are different things. The term 'filament' refers to a directly heated cathode (i.e. the bit that does the heating also emits the electrons). The 'heater' is found in indirectly heated devices (i.e. the bit that does the heating heats up another bit that actually emits the electrons. The main advantage of the later arrangement is that it permits the heater to be powered from an AC source. This is not possible with directly heated devices as the AC would superimpose itself on the beam current (but this is obviously not a problem in rectifiers where AC powered filaments have used). 86.181.51.84 (talk) 13:11, 9 April 2011 (UTC)

Anode cooling in a triode: is it important?

In a vacuum diode, we want to keep the anode cool so that it does not start "boiling off" electrons and allow current in the wrong direction. But consider a triode being used for switching or amplifying rather than rectifying. If the application circuit is such that a reverse voltage is never applied across the cathode-anode terminals, does it matter if the anode becomes capable of emitting electrons? Could a tube be designed such that the design doesn't care about the anode becoming hot (only that the tube does not overheat, as a whole?) 192.139.122.42 (talk) 23:28, 31 March 2011 (UTC)

I think you are absolutely right: there is no electrical reason for the anode to be cool, and when I was recently rewriting some of those sections I removed wording to that effect, talking about a "temperature differential" as if it were a thermocouple. Even for a diode, I can't imagine the plate getting so hot that it would seriously go into reverse conduction especially considering that the plate isn't coated the way the cathode is (to lower its work function). So as far as I know, the importance of cooling is purely physical, so that the structure will not be damaged (and also to prevent thermal deformations especially if the parts don't have the same coef. of thermal expansion etc.). One MIGHT conclude that by designing a tube that could take the heat, you could insulate the whole thing and let it get uniformally hot, and thus save on heater power! But it's hard to see how you could use less than 2W to actually heat an entire tube to 800 degrees -- it would require a great degree of thermal insulation. And if you were relying on the whole thing to get warm uniformally, then the tube would REALLY take a long time to warm up (are you old enough to remember "warming up"?!). So what you're saying is correct, and I did remove the misleading wording from the article. :-) Interferometrist (talk) 02:30, 1 April 2011 (UTC)

Also, thinking about it again, as it says in the article there are transmitting tubes whose plates (under NORMAL operation) glow red hot or even (according to the article) white hot! So I'll bet those tubes would easily conduct in the opposite direction after their "anodes" get that warm. And no one cares :-) - Interferometrist (talk) 14:46, 6 April 2011 (UTC)

It is true that the anodes in these high power tubes do glow in use, but the device has been constructed such that the heat can be tolerated without damage to the tube. This is clearly not the case with the smaller devices encountered in day to day equipment. The anode can operate at elevated temperatures (and in the more powerful devices often does) but the necessity of cooling (usually by radiation) is one of preventing damage that would otherwise occur. With a positive anode voltage permanently applied (in amplifying tubes), there is, as noted, no risk of unwanted emission from the anode. I have serviced equipment with tubes that have water cooled anodes (with all the consequent fun of keeping the several thousand volts of anode voltage away from the cooling towers). 86.181.51.84 (talk) 13:02, 9 April 2011 (UTC)

In operation in a high voltage set (big screen=high milliamps) the 1B3 or similar rectifier anode was often a brighter red than the filament.

Shjacks45 (talk) 09:17, 2 August 2011 (UTC)

Vacuum tube mike preamps -- why??

I had {cn} tagged the existing statement "One can also purchase microphone preamplifiers using tubes; these are preferred by professional music recording studios." It's hard for me to imagine that there would be any difference in performance between a good solid state and a good tube preamp so this struck me as odd. I understand that tube POWER amps have different characteristics (thus different sound) when they get near or into clipping. But a preamp should stay linear.

So one editor tried to help by qualifying the statement: "..... these are SOMETIMES preferred by professional music recording studios FOR CERTAIN APPLICATIONS." Unfortunately that doesn't really take care of the problem, now someone needs to explain what applications (I would have said that amplifying the output of a mike IS ONLY ONE application!) are involved, and why. If the editor is just saying that there are SOME sound engineers who do things a silly way, then that isn't surprising or notable and the whole statement should just be deleted. I know that they do sell tube preamps (often for lots of money) but that doesn't mean they have any useful advantage. Or more to the point, that they serve any particular applications. Does anyone have any specific info that would help either strengthen and qualify this statement? Or should the statement read "Due to personal preferences some sound engineers prefer to use tube mike preamps even though there is not any evidence that they perform differently from solid state preamps." Or just cut it completely. Interferometrist (talk) 23:28, 5 April 2011 (UTC)

It can only be some engineers who prefer tube microphone preamps, for certain apps. They are not quite as rugged on the road, even the most roadworthy ones, and they are not especially repeatable or reliable. The preference for tubes is one I could never figure out, so I can only say look at the literature about the subject. I think it is a combination of superstition, half-baked science concepts, and gear envy. The shiny "warm" tube is attractive to some. Binksternet (talk) 01:51, 6 April 2011 (UTC)

Thanks for your professional opinion. If no one else weighs in (I'll wait 24 hours) I'm going to downgrade the claims to what you just wrote (but I'll find a NPOV way of saying it ;-) And since we had been talking on the other page (Talk:Yagi-Uda antenna) about wireless mikes, I can hardly help but note that anyone using a wireless mike has obviously chosen to accept using a solid state preamp (unless there is some huge wireless mike with tubes and a big battery inside to power them!) - Interferometrist (talk) 14:42, 6 April 2011 (UTC)

One of the reasons is that microphone preamplifiers have a relatively high gain (particularly when used with ribbon type microphones). Tubes are inherently less noisy than solid state amplifying devices with triodes in turn being less noisy than pentodes. I note that microphone preamplifiers are indeed still available using tube technology so they clearly are still prefered by some. It is fair to say that use is more likely to be confined to studio use because of the bulk and power supply problems. This may not be the case as very small tubes have been available for decades.

However: the claim in the article (that they are preferred) is missing that all important citation so you would be within your rights to delete it. If the original contributor wishes to restore it he needs to provide the appropriate citation in accordance with WP:VERIFY. 86.181.51.84 (talk) 12:54, 9 April 2011 (UTC)

Terminology: Anode vs. Plate

86.181.51.84 wrote:

rv: Anode is more common in the world - Plate is confined to North America.

I was just about to apologize for having learned a term not used in international English, but I did a quick check using the Google advanced search on 4 English speaking countries. I searched under "Triode" including "plate" and not anode, anode and not plate, and both, to see how many hits it reported. Note that these are non-overlapping searches. I don't have time to format this as a table, but:

Country; Anode only; Plate only; Both

US 69700 153000 71100

Australia 2120 3860 834

UK 24800 15500 15900

Nigeria 208 132 87

So the UK and Nigeria do favor "anode" but 86.181.51.84's statement isn't exactly correct. In any of these countries at least 40% of the articles mentioned "anode" and at least 50% mentioned "plate." So BOTH terms are widely used (in all 4 places). I will try to go through the article and give equal weight to both, and to make it very clear that they refer to the same thing. How's that? Interferometrist (talk) 19:27, 8 April 2011 (UTC)

The terms are still mixed up. My problem was that there are references to 'anode' in the text and in other places to 'plate' without any qualification. In other places the use is qualified, but still mixed up as the text includes "... anode (plate)" and others it refers to "... plate (anode)". The article still contains these anolmolies. The Wikipedia rules on Britsh -v- American English requires that the terminiology used should be that in which the original article was written in (unless the article has an obvious teritorial stake - which this hasn't). The original article used 'anode' throughout. The references to the 'anode' or 'plate' should either be 'anode' throughout (which meets the guidline), but I believe "... anode (plate)" throughout would be both consistent and an acceptable compromise.

I have ammended the article accordingly. 86.181.51.84 (talk) 12:24, 9 April 2011 (UTC)

Listen, you (86.181.51.84) have been very unhelpful: The concern was raised in an edit summary that "plate" is only used in North America and I would have been willing to convert it to international English, but then found out that wasn't true, both terms are used widely in all the English speaking countries I checked (see above). So I reverted that but spent CONSIDERABLE TIME going through the article:

1. I tried to make sure that both terms were identified with each other, repeatedly, so that no one reading it would be confused (and moreover that they WOULD recognize both terms so they wouldn't get confused BY the article when they tried to apply their knowledge in the real world! Isn't that what an encyclopedia is for??)
2. Although in a receiving tube (small tube used in consumer electronics) these terms are rather interchangeable, they do have somewhat different connotations and are NOT used interchangeably for other types such as a CRT where you have 3 ANODES, which are NOT PLATES or the electrons wouldn't hit the phosphor! In other situations "plate" is more appropriate and I spent some time considering each case. In many cases the choice was arbitrary, BUT I DID THINK ABOUT IT.
3. On the other hand, 86.181.51.84 went through it mechanically changing these words and screwed up things like the VFD which does NOT have a plate as he inserted in parenthesis. Nor does the word "anode" apply in the case of the Tektronix scope tube (I'm pretty sure) where he changed it to "a channel electron multiplier anode (plate) " There were a few other poor usages of the term which he shouldnt have just changed
4. I had NO IDEA that this article was written in "British English" as 86.181.51.84 claims, but if so it has many many mistakes, like the spelling of "color" and the use of the word "vacuum tube" rather than "valve". And what's more, I found out (above) that "plate" indeed is used widely in the UK. I think I know more about "British English" than 86.181.51.84 does.
5. Another issue he brought up was the history of the article: "The original article used 'anode' throughout." Well I did some research on previous versions and counted the number of times these terms are used (the numbers are approximate using a macro as they don't count the term used twice in the same line on my editor):

ANODE PLATE Version

10 17 January 2004

13 26 January 2006

46 49 January 2010

53 50 January 2011

26 48 7 april 2011

65 92 My careful edit 9 april 2011

So I really don't know what history he is talking about, or what British English or International English he is talking about and this has become a waste of time for all of us. If the main importance is "consistency" or "rationalization" then you should be careful that articles only use the word "auto" or "car" but not mix them for instance. All the other vacuum tube articles have mixed designations, and that's the way it SHOULD BE. From his edits, I really do not think 86.181.51.84 has the background to determine where one or the other word is more appropriate in cases where there is a difference in connotation, and edits his sort are unhelpful. Interferometrist (talk) 14:56, 9 April 2011 (UTC)

The word 'plate' has never been used in the UK by any electronic engineer that I know (and I know many being one myself who still regularly plays with some of the more powerful thermionic valve equipment). The only time it ever does turn up is in US sourced documentation. Oh, and some of that documentation does occasionally refer to anodes in some of the more exotic tubes as 'plates' even though, as you note, they aren't actually remotely plate shaped (but then, in reality, neither are the ones found in most normal miniature thermionic valves being, as they are, more or less cylindrical - just like in the aforementioned CRT!). Similarly, no one throughout Europe uses 'plate' either, 'anode' being the same word in most languages (unusually including French). The abundance of American terms in the article doubtless comes from the numerous edits that the article has had over the years by US based editors. The article was originally entitled 'Thermionic Valve', but some editor soon moved it to 'Vacuum Tube' without any discussion. —Preceding unsigned comment added by 86.181.51.84 (talk) 17:22, 10 April 2011 (UTC)

You may well be an electrical engineer as am I, but if you were a logician you'd be more careful about using the word "never." Here are what I found in about 2 minutes, all from the UK:

http://itequipment.co.uk/product/ECC83-12AX7LPS-DOUBLE-TRIODE-VALVE-RC.html

UK branch of American company so no surprise there.

http://tubenewbie.com/tube_pinout.html

American site so proves nothing (even the prices are in US dollars!). 86.181.51.84 (talk) 12:00, 11 April 2011 (UTC)

http://www.hificollective.co.uk/valves/other_output_valves.html

Mixed terminology, but clear that they have lifted the words from American translations of the Russian/Chinese data sheets. 86.181.51.84 (talk) 12:00, 11 April 2011 (UTC)

http://www.valvesntubes.co.uk/Vavle-amplifer-terminology

Obviously a rewrite of US sourced info (given the preponderance of US valve types) 86.181.51.84 (talk) 12:00, 11 April 2011 (UTC)

http://livinginthepast-audioweb.co.uk/el8410wampcct.php

The word 'plate' appears only once (presumably to aid non UK readers - but they forgot the parentheses) 'anode' used on its own elsewhere. 86.181.51.84 (talk) 12:00, 11 April 2011 (UTC)

http://www.rossrevenge.co.uk/tx/five.htm

As this is lifted directly from the American products info - no great surprise. 86.181.51.84 (talk) 12:00, 11 April 2011 (UTC)

http://www.svetlana-tubes.com/pdf/3cx3000f7.pdf <<<< Data sheet

An American translation from a Russian data sheet so hardly any surprise. BTW the original Russian language version uses 'АНОДЕ' (cyrillic pronounced 'anode'). 86.181.51.84 (talk) 12:00, 11 April 2011 (UTC)

The Russians do use the word 'anode', but they don't spell it the way you did so you couldn't have got that from the Russian data sheet, as you claim. Since Russian is pronounced exactly as it is spelt (with a couple of wrinkles), what you wrote would be pronounced (as near as I can type it) as 'anodye' The correct Russian word is 'АНОД'. I don't have a suitable data sheet to hand, but this site would seem to agree with me. DieSwartzPunkt (talk) 14:51, 11 April 2011 (UTC)

Most of these use both terms and identify them. But for the sake of any electrical engineer in the UK (or the US) who isn't familiar with both terms (and hasn't read much else about tubes/valves) the Wikipedia article in its current form will be DOING THEM A FAVOR. Enough said. Interferometrist (talk) 18:25, 10 April 2011 (UTC)

So all your 'evidence' is from non UK (or even European) origin thus proving nothing. The article in the form you left it uses 'plate' almost exclusively apart from a few places where both terms have been used. Once again, the term 'plate' is NEVER used in European origined documantation (or indeed Russian - and I doubt if it's used in China either, but I can't read Chinese). The only time it creeps in is US sourced documentation (where it is almost exclusive). The Americans also insist that any documentation that we produce for them is 'translated' into American before accepting it (including spellings).

Since both terms are now used except on the (possibly valid) places you noted, there should be no problem. If I have missed a 'plate' where the term is not used, it is an omission nothing more.

However since you have now made three reverts without providing any valid explanation, you are edit warring. 86.181.51.84 (talk) 12:00, 11 April 2011 (UTC)

I can confirm that the term 'plate' is hardly ever used in connection with a thermionic valve in Europe, at least by European engineers. I wouldn't go so far as to say 'never' because of the creeping Americanisation of English. I doubt, but couldn't say for sure, that it is much used outside of the Americas. DieSwartzPunkt (talk) 12:52, 11 April 2011 (UTC)

Most European engineers won't be speaking English. --Wtshymanski (talk) 13:19, 11 April 2011 (UTC)

Actually a surprising number do. But in any case, even when the use their own language, they use the word 'anode' in connection with thermionic valves as every European language including (unusually) French uses the word. The word 'plate' contrary to American belief - doesn't. 86.181.51.84 (talk) 13:36, 11 April 2011 (UTC)

EVERY European language? Even Welsh? Basque? Romansch? Irrelevant in any case, what matters it what it is called in English. American usage is a significant part of the technical literature on vacuum tubes and it would be whimsical of the Wikipedia to pretend otherwise (though not unprecedented). I remain baffled by anti-Americanism here of all places - we don't have tin-pot bicycle-riding Eurotrash nobility to thank for the WIkipedia, do we, now? --Wtshymanski (talk) 14:08, 11 April 2011 (UTC)

Very possibly. Most European and even Asian languages (except the French) tend to adopt the (English) Westernised word for technical terms - even the Russians. Those particular languages just might adopt a completely different (and unique) word, but only because their speakers have a political axe to grind. American usage might be significant part of the technical literature on vacuum tubes - in America. But not elsewhere. In general, the only American usage in Europe tends to be in American sourced literature. I am not as versed in old valve technology as some might be (though they were all the rage when I started in the electronics game back in 1969!), but I still see the odd bit of literature from time to time. DieSwartzPunkt (talk) 14:51, 11 April 2011 (UTC)

The Welsh word for anode is still anode, although it's spelled anod (similarly cathod). This use is consistent in both radio, with valves, (and Welsh radio has a very long history) and also electric arc furnaces. Andy Dingley (talk) 09:02, 14 April 2011 (UTC)

This is getting silly

Listen, you have also made 3 reverts, plus the original edit, which I am sure was well intentioned, but your subsequent edits haven't made any attempt to obtain consensus. Nor have you convinced me that one or the other is prefered in international English. They are both widely used, and if there actually were places that didn't use one then the Wikipedia article will help them learn both. At least some of your claims about the UK websites I looked up is false, saying that the Russian tube datasheet was translated in America and that they don't use "plate" on their own website which I just verified they do. Your latest reversion removes other improvements to the text, contains sloppy errors, and includes sentences such as:

Since secondary electrons can outnumber the primary electrons, in the worst case, particularly as the anode (plate) voltage dips below the screen voltage, the anode (plate) current can actually go down with increasing anode (plate) voltage.

Publications which include both terms do not insult the reader with multiple reminders of these terms' equivalence within the same paragraph, let alone within the same sentence. I already went overboard with that and you made it worse. I cannot see any reason not to revert it again, and ask if you think there can be improvements in the exposition (which there surely can be) you make those edits ONE BY ONE rather than pasting an old version back in as you have been doing. I don't think that I have to be the one to think about each use of the terms in detail while you just use your computer's copy and paste. If you point to specific sentences that are clumsy or unclear or where one term is surely more appropriate, then make those edits and you will have reason to defend those edits (in which case they'll be accepted by everyone else too). Just saying that the engineers you know don't use the term isn't going to wash. I'm reverting this (and will again if you force me) until you try to make edits on the basis of consensus, which requires reaching consensus. The consensus that I was expecting is that the article should use both terms, identify them (but not at every single use of the word), and not favor one over the other since they are in use with approximatly the same frequency. If you don't accept that consensus then try to convince the others here. Interferometrist (talk) 17:17, 11 April 2011 (UTC)

Since you have introduced the subject of consensus, I note that there are currently 3 editors (including myself) stating that the term 'plate' is not in general use outside of America so I would argue that it is ample evidence that that is the case, and that it is you that is not heeding that consensus. Asking around my colleagues here, I find that less than a third knew that the Americans called an 'anode' a 'plate'.

The real problem here is that, in general, most Americans somehow believe that the rest of the world speaks the same version of English that they do. I can tell you that this is most definitely not the case. Indeed, I myself, hadn't even heard of the word 'plate' in connection with this sort of technology until a few months ago. Granted, though: that my experience with thermionic valve technology may be a bit more limited than some on here.

I would suggest, that in the light that most of the English speaking (and non English) world uses the word 'anode' (and apparently so does the US in some contexts), that the article should use the word 'anode' throughout. The alternative word 'plate' should obviously be left in the introduction. This also reflects the fact that British English readers may never have heard the word 'plate'. The word 'plate' can be included in brackets after the first use of the word 'anode' in each section (or possibly paragraph) where it is pertinent. I do not feel it needs to be repeated every time the word appears as most readers memories should be longer than that.

Now since Interferometrist wanted consensus, let's see what everyone thinks? My vote is that article that Interferometrist keeps reverting be restored, but the preponderance of "anode (plate)" be reduced to one per section in those places where 'plate' is used. 212.183.128.39 (talk) 07:46, 12 April 2011 (UTC)

Whoever wrote this clearly doesn't understand what "consensus" means and fortunately we don't decide by voting since if that were the case then the preference of Americans over the rest of the world who speak English would have undue weight. When the issue of terminology was introduced (not on the talk page but by someone making a sweeping edit which was reverted to 3 times without any attempt at consensus) I made an actual attempt to determine the usage in 4 different English speaking countries (see #Terminology: Anode vs. Plate above) and found that "plate" was preferred in the US and Australia, "anode" preferred in the UK and Nigeria, and both terms were widely used in all 4 countries so I thought Wikipedia should do the same. 86.181.51.84 complained that none of the engineers he knew ever used the term, but someone in a different location could well write the opposite and neither would prove anything. I live in Europe by the way, and I don't even know what term people use here because actually almost no one talks about tubes anymore (try asking a newly graduating EE what voltages he would apply to which electrodes!) so most of the references are historical anyway, and historically both terms are used with great frequency.

Also I'm getting wary of giving the same weight to IP addresses rather than real people with a history editing articles to improve them in tangible ways. It's fine if 86.181.51.84 or anyone else wants to improve the current version with specific edits that can be judged on their own merit, that's fine. But as far as a sweeping policy affecting the entire article that hadn't been seen as a problem for the last 7 years until a week ago, there obviously isn't any consensus on the issue and we're not going to call for a "vote" among experienced editors let alone all 4 billion possible IP addresses. If there actually is a Wikipedia policy issue bearing on this discussion I'm happy to hear about it, but not to get lectured as such by 86.181.51.84 whose participation as an editor (at least under that IP) goes back not even one week (and an undue amount of which consists of involvement in edit disputes and issuing warning tags). Interferometrist (talk) 09:31, 12 April 2011 (UTC)

As an anymous editor, I am as entitled to edit Wikipedia and discuss content as you are, whether you like it or not - Wikipedia itself says so. In fact, it is you who are the anonymous one hiding behind a user name (that obviously isn't 'real' as you claim) that could originate from anywhere on the planet. And I have been contributing far far longer than one week (and very likely longer than you have). It's not my fault that I live in a place where ISPs allocate a new IP address everytime a connection is made. I have no obligation to create a user accout with a silly name, and can see no advantage in doing so.

I regard 212.183.128.39's suggestion as probably a better compromise than the one I made as it avoids, what I conceed, is repetition of the parenthetical 'plate' but at least explains it at the section level (not all links go to the head of the article).

Oh - and Wikipedia does work by concensus. 86.178.14.98 (talk) 13:05, 12 April 2011 (UTC)

If we're already talking about consensus, put my !vote in with Interferometrist and Wtshymanski. I've been silently reading the comments here for days and their arguments are solid. Binksternet (talk) 13:39, 12 April 2011 (UTC)

It's a gas, isn't it? I once read that Usenet news (remember Usenet?) was like a particle accelerator, in that it made it possible to observe opinions of an intensity and transience that don't exist in the natural world. "Consensus" is agreement among several interested parties. The Wikispelling "concensus" is a code word for "all the sensible people gave up long ago, and an uninterested admin finally closed it". There's what, half a million "contributors" here? How many of them weigh in on anything? Here we are on the Wikipedia, virtually eating hot dogs on the White House lawn, and pretending that the Americans have no more effect on the English language than does a Kiribati mission school. Self-deception is one of the less atractive qualities of WIkipedia editors. --Wtshymanski (talk) 14:00, 12 April 2011 (UTC)

I fail to follow your logic here. In the light of the original objection noted above, I have just had another look at the article in question. The terminology is indeed somewhat mixed up. For example, this fragment,

"Since the tube requires a vacuum to operate, convection cooling of the plate is not generally possible (except in special applications where the anode forms a part of the vacuum envelope; this is avoided in consumer products due to the shock hazard it entails). Thus anode cooling occurs mainly through black-body radiation."

So reading this, it is not clear that the plate refered to at the beginning of the first sentence is, in fact, exactly the same part as the anode refered to twice towards the end of the fragment. This example is by no means alone in this regard. The terminology desparately requires rationalisation. I could not find the word 'plate' in a UK published dictionary of technical terms (only 5 years old). However, the word 'anode' does turn up in an American published similar dictionary (and defines it incorrectly - but that is another issue). On this basis, 'anode' would seem to be the more universally recognised term, if not necessarily the most used. An encyclopeadia should be as accessible and understandable as possible to as many readers as possible. DieSwartzPunkt (talk) 15:50, 12 April 2011 (UTC)

At last, somebody who understands the point that I have been making for some while. And a 'real' (as in anonymous) user as well! How can the opponents believe that an article with mixed terminology can be comprehensible? I think another revert beckons - when I've got the time. 86.183.175.94 (talk) 15:35, 13 April 2011 (UTC)

Yes, I still intend to rationalise the terminology, but there is a heck of a lot of it, and I just haven't had the time. 86.183.175.94 (talk) 17:10, 21 April 2011 (UTC)

Triode voltage gain: don't edit if you don't understand it

The recent attempt by DieSwartzPunkt TWICE to remove the primary distinguishing characteristic (among triodes) of the 12AX7 is very irritating given that he doesn't seem to know what he is talking about. I don't know if his remark contained a typo or two, but is wrong in any case:

Vacuum tubes do not feature current gain. so 'voltage gain' is tautologous)

Of course tubes have a HUGE current gain (almost infinity at DC) and pentodes are rather characterized by their transconductance. Triodes however also have a limited voltage gain, mu, and if anyone needs to they can look up the spec sheets on the 12AX7 versus the 12AT7 (they even have pages on Wikipedia) and you'll see the difference. In short, if you don't understand the subject, don't edit, and certainly don't revert when you neither understand nor have a reliable source to the contrary. I feel like I'm wasting a lot of time dealing with incorrect/unhelpful/disruptive edits which an inconsiderate person spends not one minute on, and I expect an apology in this case. -- Interferometrist (talk) 17:04, 19 April 2011 (UTC)

Well, clearly, you don't understand the subject as well as you think. Mu is not actually voltage gain as you claim (so there is no tautology). Mu is a dimensionless theoretical value extrapolated from the tube's characteristic curves and is called 'amplification factor' (though it is calculated from delta voltages within the curves) . The term 'amplification factor' doesn't mention voltage amplification in any way. Voltage gain of a tube stage will always be less than mu (and for a tube stage properly matched to the following stage will always be less than half of mu). As you note (and DieSwartzPunkt apparently didn't realise), the current gain of a tube is indeed extremely high. In a practical circuit the current gain of (let's stick with the tube at hand) a 12AX7 based cathode follower circuit can easily exceed 1.5 billion. The current gain is also related to mu, but can be larger or smaller in a practical circuit. The spec sheets invariably give the value of 'mu' for a triode tube but they never claim that it is the voltage gain of the tube (especially as the current gain, and indeed power gain, are also related to it). As you nearly got right, the characteristic invariably given for pentode tubes in lieu of mu is the mutual conductance (I have never seen the term 'transconductance' used in connection with any tube). However that merely is a convention that has been adopted for the two types of tube. The mu of a pentode is easily found by dividing the anode resistance (R_a) by the mutual conductance (g_m) - both of the latter of which are usually quoted in a pentode spec sheet. Similarly for a triode, the mutual conductance can be found by dividing R_a by g_m.

The article is best served by deleting the phrase 'high voltage gain' and leaving the tube to be described as a 'high amplification factor (mu)' which is what all spec sheets claim (e.g. [here]). If you wish to revert the phrase 'high voltage gain' then please provide a citation that the two terms are interchangeable in accordance with WP:VERIFY and WP:BURDEN - though I doubt you will find one. 86.183.175.94 (talk) 15:19, 21 April 2011 (UTC)

I'd still favour some wording that expresses "voltage gain". This isn't a technical article, it's an entry-level piece in an entry-level encyclopedia. About all we might expect from 'typical' readers here is some vague knowledge of bipolar junction transistors. In that case, the contrast between a voltage gain (even when this is really mu) and the current gain exhibited by a BJT is an important contrast. Andy Dingley (talk) 15:56, 21 April 2011 (UTC)

Listen, if I were to take you seriously then I would indeed find (and it wouldn't take long) a RS regarding terms that YOU haven't heard, but as far as I can tell you are just here to waste my time and I'm getting annoyed. I know I'm supposed to assume good faith and I did when the issue of plate/anode terminology first came up, but my assumption has been proven incorrect.

Right, "mu" which is ALSO called the "voltage gain" in the spec sheet of a triode, or the "amplification factor" in the spec sheet you provided indeed is the MAXIMUM voltage gain that you can get, but as you so kindly pointed out it's possible to make an amplifier out of it which has less voltage gain. And I can't help but noticing that that spec sheet DOES use the word "transconductance" (but if you're careful not to read it then you can keep claiming that you've never heard the term, just like you've never heard a UK EE say "plate") and these terms indeed are SYNONYMOUS but you don't want to hear that so keep ignoring me if you must. I'm glad you now realize that the current gain of a tube is hardly an interesting parameter (especially at low frequencies) so that power gain is likewise, and the interesting figures are transconductance and, for a triode, mu=voltage gain (at maximum). But since current gain and power gain are not relevant parameters, any reference to "amplification factor" (like in the spec sheet) could ONLY be referring to voltage gain (as is even mentioned in the introductory text of the spec sheet which you didn't have time to read)! The reason the article on vacuum tubes says "voltage gain" rather than mu (which is mentioned in parenthesis) is so the "casual" reader can be told something about that tube without digressing into a detailed explanation of what mu means and how you solve for the voltage gain of a particular circuit. It's like calling a Ferrari a "fast car" without giving a detailed description of its engine, drive train, and tires and mentioning all of the factors that determine the speed that it will reach in practice. Get it? Interferometrist (talk) 16:42, 21 April 2011 (UTC)

The actual practical voltage gain can never be as high as mu. Mu is relevant to current gain, because the calculation for current gain includes mu as a parameter. I now require a valid citation that 'mu' is compltely and uniquely synonomous with 'voltage gain' as stated above. In my view the term 'amplification factor' was a perfectly acceptable term for the purpose of the article, given that that is what 'mu' really means. No citation, and I am at liberty to remove the term 'voltage gain' (In accordance with WP:BURDEN, the onus is on you to prove that it is, not on me to prove that it isn't). I can easily find a citation that mu means 'amplification factor'.86.183.175.94 (talk) 16:58, 21 April 2011 (UTC)

Oh, and I'll conced the transconductance point. It turns out that this a language problem. The correct English term is 'mutual conductance'. 'Transconductance' is a foreign term (American). All of the valve ('tube' is a foreign word as well) data sheets that I use day to day (and there aren't too many these days) are of non American origin. 86.183.175.94 (talk) 17:08, 21 April 2011 (UTC)

"Foreign term"? Maybe you should point that out on the Wikipedia article then. You seem to think a lot of English terms are "foreign." Where are you really from, Mars? Interferometrist (talk) 17:18, 21 April 2011 (UTC)

Ignoring the grave threat of being blocked (!) I have taken out the {CN}, as I assume there are no autoconfirmed users (with one exception who I don't expect to hear from) who would belabour this point. Interferometrist (talk) 17:32, 21 April 2011 (UTC)

What does 'autoconfirmed' users have to do with it? You have come up with that before and it still is nonsense. Any editor can challenge your rubbish. Seeking article protection to prevent someone from challenging your material is a gross abuse of that process. 86.183.175.94 (talk) 17:42, 21 April 2011 (UTC)

I expect the dispute will be quickly resolved by finding textbook descriptions of the 12AX7 and citing it here please. —EncMstr (talk) 17:48, 21 April 2011 (UTC)

Well this is about a particular part number and there are thousands of tubes: these are not dealt with individually in text books, or encyclopedias, though Wikipedia does have a page on it whose first sentence is "12AX7 is a miniature dual triode vacuum tube with high voltage gain." If it had said almost anything else (besides "high mu," a more obscure technical term) that article would have been remiss. This page linked to by the article:

http://www.amplifiedparts.com/tech_corner/12ax7_comparison_of_current_made_tubes

also has a column called "gain" and as is understood by anyone with an understanding of tubes (as pointed out above) that could only refer to voltage gain (mu) which is why they left out the qualifier "voltage" in that table (the other possibilities would have been "current gain" or "power gain", neither applicable to a tube). I could research further, but as I've been complaining in the LAST 3 TOPICS above, this person (regardless of identity) has become a tremendous waste of time for us who are trying to write CONTENT!! Interferometrist (talk) 18:02, 21 April 2011 (UTC)

Thank you Binksternet for adding the citations! However I don't think you REALLY needed to add them to the article and I am still dissatisfied with matters inasmuch as a lot of time has been wasted and there are hundreds of other terms or little facts that could be {CN} tagged and I'll be damned if I'm going to waste an hour explaining what each term means let alone finding a RS just to comply with bad faith requests by a SINGLE person! Interferometrist (talk) 18:18, 21 April 2011 (UTC)

You're welcome. Mu is amplification factor, or voltage gain, per textbook-type sources. Binksternet (talk) 18:51, 21 April 2011 (UTC)

No it isn't. Mu is amplification factor, a figure theoretically derived that provides a method of calculating voltage gain in a particular stage (and, so I discovered, current gain). I had previously missed current gain because valves are seldom (if ever) called upon to perform such a task, but it is obvious (once it is pointed out) that they must. Also 86.183.175.94 is incorrect in his assertion that the voltage gain of a single valve circuit can never be greater than mu. It most certainly can - if positive feedback is applied. This was a popular technique in the early days of radio - I even possessed one back in the 1960's). I am not aware of its use in AF stages, but it was frequently employed in the RF stages. The amount of feedback was variable (by means of the 'reaction' control), and the technique was to adjust the reaction control until the RF stage just failed to oscillate. Substantial increases in gain well above the 'mu' figure could be achieved thus way. The technique became obsolete somewhere around the point that the superhet was introduced.

By the way, one of your sources is obviously a self published source without a credible backing (a shop perhaps) - unacceptable by wikipedia standards. I can't locate a copy of the other, but maybe the author has made a common mistake - it does happen. None of my electronics training notes from my early days refer to mu as anything other than 'amplification factor'. I was probably unfortunate, that I was I was in the last year that the training was based on valves. Apart from a brief flirtation in the late 70's (25 kW output amplifiers) I haven't touched them since. DieSwartzPunkt (talk) 08:37, 22 April 2011 (UTC)

I included the online store page because it was immediately accessible while the book source was not. I just now replaced it with a better online source showing the same article: "Mu, Gm and Rp and how Tubes are matched", by Roger A. Modjeski, former Chief Engineer at Harold Beveridge, Inc., and the founder of RAM Tube Works and Music Reference.

I could have put other references in besides, but I felt that two were enough. Apparently not! Here are some more:

• "mu is the voltage gain". Audio cyclopedia, by Howard M. Tremaine.
• "The mu of a tube is the amplification factor, or voltage gain." 22 radio and receiver projects for the evil genius, by Thomas Petruzzellis
• Glen Ballou in the comprehensive Handbook for Audio Engineers defines mu as "amplification factor". Directly following this definition he equates the terms under discussion, writing, "Amplification factor or voltage gain is the amount the signal at the control grid is increased in amplitude after passing through the tube, which is also referred to as the Greek letter μ (mu) or voltage gain (V_g) of the tube."

I consider the Ballou reference the last word. Binksternet (talk) 18:04, 22 April 2011 (UTC)

Thanks for doing the research Binksternet! But actually you went too far. This didn't need an in-line citation in the first place. Citations are only needed in the article to back up facts, and this wasn't a question of fact but one of terminology, whether these terms are the same or, as this guy was claiming, that they were different. It was only the latter (false!) claim that would have made the wording in the article problematic. EncMstr had only requested that the appropriateness of the term be cited HERE (since he wasn't familiar with it himself) so that there wouldn't be any question that the article's terminology was proper.

Furthermore I am dismayed that this guy went ahead and edited after all we went through, forcing you to revert him once again and again look up the identity of these terms! The page is now semi-protected (the result of the "complaint" he tried to lodge against me!) in order to keep him from screwing with it, and WASTING OUR TIME AGAIN AND AGAIN :-( -- Interferometrist (talk) 19:14, 22 April 2011 (UTC)

Oh also, thanks for dealing with my embarrassing spelling errors (!) in addition to the citations. This isn't like urgent, but I really don't think those citations are needed there (should instead copy them to this page as a matter of record) but the more useful pages you looked up (where there is web content) could well be added to the External Links with some accompanying description. I haven't yet had a chance to judge them myself (I'm still too pissed about being asked to prove that A=A!). -- Interferometrist (talk) 19:25, 22 April 2011 (UTC)

Stupid complaint against me

Just for anyone's entertainment, here is the official complaint against me ;-)

WP:3RRNB#User:Interferometrist reported by User:86.183.175.94

I'm hoping the judge will let me off easy so I can avoid serious jail time..... Interferometrist (talk) 19:01, 21 April 2011 (UTC)

It's more like hockey - if you don't get the whistle blown at you once in a while, you're not playing hard enough. Difference is, every player has a whistle! --Wtshymanski (talk) 19:43, 21 April 2011 (UTC)

"Tube sound"

Any but the briefest mention of "tube sound" is out of place on this page (it has an article of its own). It's not particularly relevant to electron tubes in themselves, and is far too complex to summarise well—unnecessary as it has its own article. The text said that tube sound was due essentially to the use of signal transformers. It's also said that a semiconductor amplifier has inherently high distortion; reduction to a very low value by massive negative feedback results in very low thd but with a higher proportion of high, and audibly very dissonant, harmonics. The nature of clipping also plays a part. There were transformer-coupled transistor amplifiers which didn't satisfy those who preferred tubes, as far as I remember. And of course there are people who say that tube sound is not better. I'm not trying to pick an answer (and am out of date on this anyway), but it seems that it's impossible to summarise the situation in a brief but neutral way, beyond saying what I did "Many people prefer tube-based to semiconductor audio equipment, citing a more desirable "tube sound". Some companies specialize in high-end tube audio amplifiers to serve this market." And, given the "tube sound" article, it's not particularly relevant. Pol098 (talk) 18:24, 25 April 2011 (UTC)

Ok, this isn't a big deal, but I thought (still think) that it deserves more than a tiny mention because use of tubes in hi-fi (AND guitar) amps is one of the most visible and well-known modern uses of tubes. Yes, there are more vacuum tubes in microwave ovens, but you don't take those to the drugstore to test (are you old enough to remember that?!). So I think it's rather significant if we're talking about modern applications.

Yes, there is an article on that, but in case you haven't noticed there is a lot of duplication of material in WP articles and always will be. Sure, insert a {main} tag to that, but still mention it here (in less detail). And whether and where tubes have advantages (real or perceived) certainly is relevant to this article. I myself have a POV but tried hard to present it in a balanced manner. Namely that there IS a difference in sound, yes, largely due to the transformer. And also because a single-ended (class A) tube amp when overdriven produces even order (not just odd) harmonics, though I would note (if I were arguing the point!) that many tube enthusiasts don't seem to mind push-pull tube amps as long as they can see the pretty glow of the filaments ;-) There is also the higher output impedance of a tube amp compared to the very low dynamic output impedance of a solid state amp (due to negative feedback) which is cited as an advantage in other contexts, called a "high dampening factor" (but you never see this as a point of debate). This changes the realized frequency response given a speaker's impedance vs. frequency.

I also carefully wrote that claims for differences in sound using tubes in pre-amp / signal amp stages are "more controversial" but left it at that.

Also, your edit said "many people" which isn't right. They are a rather small number within the audiophile community but have a disproportionate amount of spending power (they need it!), enough to keep these companies in business. (I used to work next to a place making these: they probably shipped one or two amps a week and their business was doing alright!). So I think the emphasis isn't misplaced and if you see specific problems I would invite you to address those individually. How's that? -- Interferometrist (talk) 19:49, 25 April 2011 (UTC)

I'm in favor of a summary paragraph here about tube sound, with a prominent "Main article"-type link to Tube sound. Binksternet (talk) 20:09, 25 April 2011 (UTC)

Well give it a quick edit then and ask me again (in other words, don't spend time polishing it but indicate what you approximately propose). But the relative importance of tubes needs to be properly reflected, and I'm thinking..... but I do not believe there is one other product in existance which you can order made using tubes as an option! (Maybe transmitters, but that's not a normal consumer item). In other words, by almost any criteria audio has to be among the most significant uses of tubes today, and this isn't specifically a history article. Interferometrist (talk) 22:07, 25 April 2011 (UTC)

Tube sound IS a big deal to some people! However it is an artifact of Class B push-pull output stages generating Even Harmonic Distortion which many people perceive as pleasant and harmonic (F to 2F is one octave higher). Recently there was a computer motherboard manufacturer that placed a tube (12AX7?)on their motherboard for that reason. Shjacks45 (talk) 09:11, 2 August 2011 (UTC)

Use in computers and reliability

Well it's funny, because I saw the new edits about tube computers and found so much conflicting information, that I was getting ready to rewrite it so that Pol098 would see that he was wrong and I wouldn't rub it in. But THEN you took it so much further and further! So look, I hardly wanted an edit war (we've already had enough of that!) and just want to get to the truth, so I'll bring it up it here.

Yes, I read that interview too, but I actually don't believe it. According to the Wikipedia article on ENIAC:

In 1954, the longest continuous period of operation without a failure was 116 hours (close to five days).

So I did some quick math, and if the MTBF actually was 2 days, then there would have been MUCH longer periods of continuous operation. That guy was bragging about his success 40 years later, but this statistic (if correct, it's not referenced) belies any such possibility. When I researched it I found a RS (sort of....) who wrote:

"The ENIAC had almost 18,000 vacuum tubes, resulting in an overall MTBF of 7 minutes..... The key to keeping it running was to reduce repair time (MTTR): a staff of 5 people were patrolling the innards of the 30-ton monster on roller skates and replacing tubes as soon as they blew out." - http://www.pld.ttu.ee/IAF0530/basics-1.pdf

Now he sounds like a reliable source, but not on the ENIAC in particular so I guess it's possible that he picked this up anecdotally. Still, the comparison of numbers, 7 minutes vs. 2 days is rather striking. Another source mentioned the Whirlwind as having a MTBF of 20 minutes I believe (don't see it now). Presumably that was using the "special-quality" tubes lauded in the article.

The line:

By the late 1950s it was routine for special-quality small-signal tubes to last for hundreds of thousands of hours, if operated conservatively

has two problems. First, by the late 50's vacuum tube computers were just becoming obsolete. Secondly, a tube that lasts 200,000 hours, say, would imply a MTBF of only 12 hours in the ENIAC which used 17,000 tubes, so that statistic isn't very impressive. That's taking the 200,000 hours as the MTBF assuming a poisson process. It could NOT have meant that individual tubes wear out consistently after 200,000 hours because it would have taken 30 years to measure that!

My guess is that the 7 minutes may have applied to the early ENIAC in 1946 (or may have been an incorrect number) and that reliablity improved greatly between then and 1954 when the machine worked for almost 5 days without a failure (and when that guy "remembered" that tubes only failed every two days, which again is inconsistent with the other statistic). So let's get some more accurate information, no?

Also, this:

In 1934 Flowers built a successful experimental installation using over 3,000 tubes in small independent modules;

is a nice historical note about this guy who I'm sure was very smart and made tubes much much more reliable. BUT it doesn't reflect on the ability of tubes to be reliable in a computer where the failure of a SINGLE tube brings the machine down, whereas a phone line going down doesn't bring down the phone system. So it's sort of misleading if it's trying to make a point.

Oh also, you didn't have to change the title of the section from early computers: Section 3 is about history and discussion of reliability per se might rather go under section 3.7. Of course the unreliability of tubes has a lot to do with their (non-) use in computers, but in a HISTORY section it makes sense to have a subsection on computers. OF COURSE the content of it would concern the (un) reliability of tubes, but please change the section title back. Thanks, Interferometrist (talk) 22:07, 25 April 2011 (UTC)

I hadn't realised you disagreed so fundamentally otherwise I'd have acted differently. I always make bold edits, but am happy to revert if necessary, and know that eventually anything suspect (and most of the valuable stuff too) is susceptible to reversion by someone over time. Reliability: quite a lot of the literature is fairly clear that tube equipment was reliable enough to be every useful. The literature on Colossus and Flowers is quite clear that the equipment he built worked pretty reliably. But ultimately the issue is, I think, whether a tube computer was reliable enough to do useful work, rather than being a plaything. In my own experience I found it very useful, and everyone managed to get usable work done. It was down occasionally, but I don't remember huge frustration—people said things like "it's taken 20 minutes to do this job, but it would have taken me years by hand". The machine was built into narrow cabinets running along the length of a long room; I don't know how long it took to identify a problem, but everything was very accessible, and I expect a simple failed tube would have been quick to locate.. Basically the machine was a useful tool, worth its (high) cost, something practical. In hindsight the small amount of storage and limited programming language were problems; reliability wasn't.

Personal experience isn't considered valid for Wikipedia, but it does match written reports I've quoted.

I was very careful to distinguish Flowers's modular 1934 equipment, which was accepted by the Post Office as reliable enough to do its job, from Colossus, where everything had to work together; the early work suggested that reliability would be sufficient, and the history of Colossus was very successful as a reliable system. Pol098 (talk) 00:33, 26 April 2011 (UTC)

Mentioning reliability in the header for the section on tube computers: I added it there quite deliberately, as reliability only becomes a serious issue for equipment using thousands of tubes, so is very much tied to computers. As there are objections I'll revert this. I do think that we should speak of tube, rather than early, computers. The earliest computers long predate tube computers; and tube computers were being sold 6 years after the first transistor computer was made.Pol098 (talk) 00:45, 26 April 2011 (UTC)

A point that changes the figures here: One tube failure does not equal computer failure. Bear in mind Colossus was programmed by rewiring it, not by software, so at any one time not all the valves would have been in the circuits being used for calculations. Tabby (talk) 16:55, 18 June 2011 (UTC)

Misc.

By the way, most of your edits are for the better, improved wording etc. so I didn't mention that and am just mentioning the things where I differ. Because you made so many edits there are a number of those, but don't think I don't appreciate the edits I didn't have particular reason to mention. :-)

One little thing: the "Classification" section had only been introduced in order to delineate the scope of the article, that is, to steer people AWAY from it if they were thinking of any non-vacuum (gas filled) tube OR if they were thinking of CRT's, phototubes, etc.. It's called classification because it had no other obvious title, and delineating the subject natter would have been too involved for the lede, but really the content of the article starts with "Description". At least that's the way it had been intended (but perhaps it's beside the point now).

"the the whole field of digital electronics is analog electronics (which it is)"

Yes, that's something I always point out myself: that all digital circuits are analog circuits, and this becomes clearer and clearer when you go to higher switching rates, that is, when you really try to take advantage of the maximum "switching" speeds of a device, at which point is ISN'T really switching, it's spending half the time ramping up and down (otherwise you'd be able to use it at an even higher speed, after all). So people draw waveforms as if they were square, but the only way they get away with that is by complying with the setup and hold times which takes into account the analog behaviour of so-called digital electronics without actually thinking about it as it exists.

Well, I still think adding the word "switching" at that place is unneeded since you seem to agree with me, that switching is amplifying, and the lede is supposed to be made simple and concise. Like, why didn't you include "oscillating"? Because it's (also) a form of amplification, of course. The irony is that there were at least 1,000,000 times as many pieces of equipment which used tubes for oscillating than switching.

Anyway this isn't a big deal, just an unneeded word. Interferometrist (talk) 22:35, 25 April 2011 (UTC)

It's quite normal and correct to discuss only what you disagree with, it's not a mutual congratulation society. Thanks for mentioning your approval too. I see what you mean about "classification", but it is pretty misleading as a section title; "scope" or "scope of this article" (maybe a sentence in the introduction rather than a section) would be better. While we agree that switching can be considered an application of amplification, we disagree fundamentally on its significance - I go along fully with the conventional distinction between linear and digital semiconductor components, with considering switching and amplification as fundamentally distinct, with classifying transistors as mainly switching or mainly amplifying devices. So I think it's necessary to include "switching" as much an application of tubes as amplification (historically, not today). Let's see what others think. Anyway this isn't a big deal, just a needed word. Pol098 (talk) 00:42, 26 April 2011 (UTC)

I don't have much time to deal with this page or wikipedia at all right now, and you're doing a fine job editing it (of course I will have various disagreements or ways I would have written it differently, but that can wait). But I'll just answer your latest remarks (also from the previous section):

Not important for the article (and non/inclusion of the word isn't so important), but the main difference between analog and digital electronics has to do with the circuits surrounding them, not the amplifying devices themselves. Except that device linearity is obviously of no importance when used for digital circuits, but nor is it so important in modern analog circuits which are dominated by negative feedback (or which are very small signal levels so that anything is linear). Analog circuits rely on negative feedback even when it isn't explicit, while digital circuits if anything have positive feedback which is what a flip-flop is, and are supposed to be overdriven. And my point before was that at the highest speeds a flip-flop might not even have time to have fully latched into a different state when one of its outputs must already be set up for the next stage, and that the waveforms aren't square at all but barely reach the right logic level in time for the next clock pulse to come along. So they are still "analog" circuits in that sense, just having a digital purpose. And as far as semiconductor devices themselves recovery from reverse biasing or saturation is an issue not relevant to signal amplification. But for tubes, were there any specific "switching" tubes or particular characteristics associated with digital applications? (You'll surely forgive me if my knowledge of electronics doesn't include using tubes in computers! I'd hardly know where to even look it up.)

VERY interesting that you have personal experience with tube computers! I don't remember ever seeing one though its possible that I had seen a tube computer and perhaps couldn't have told the difference. After all, large computers were in "computer rooms" with locked doors (but big plate glass windows) and you had to be special to even get in there (and certainly no children!), and the main thing you could see were the old tall tape drives jerking their reels and maybe some blinking lights. So you WERE in contact with a tube computer? I'd be interested if you could mention the model (and anything else about it). You say they were still selling tube computers 6 years after there were transistorized machines? Surprising... so that must have been well into the 60's?

"Personal experience isn't considered valid for Wikipedia.... "

Oh of course it is! It just doesn't qualify as verifiability. Your knowledge is very valued whether it's based on personal experience or what you subsequently learned or read. It saddens me to think that in 20 or 30 years none of the Wikipedia editors writing the page on vacuum tubes would ever have waited for a radio or TV to warm up, let alone taken the tubes out and tested them. On the other hand your personal experience can be overgeneralized and it sounds like your use of tube computers was at the time of their maturity (and obselecence). In other words they were so much more reliable by then that you can hardly imagine what they faced in the early 40's. Did you ever resolve the gulf between 7 minutes and 2 days? (Again, the last figure couldn't possibly be right as I've shown)

"I added it there quite deliberately, as reliability only becomes a serious issue for equipment using thousands of tubes, so is very much tied to computers. As there are objections I'll revert this."

Well not as a favor to me obviously but they are two different aspects of history. I think tube reliability affected all of their applications, the article mentions stowing them in underwater cable repeaters for instance, but also for every other use even if they were easier to replace but still a nuisance (and the source of 90% of electronic failures, which alone makes it important). It's only that tube computers made unreliability a show-stopper. But figure out how to present the history logically. I consider the use of tubes in computers of great interest even regardless of the reliability issue. (For instance, I was shocked to have learned from this that tube computers were clocked much slower than 1 MHz, even though tubes were being used at VHF by then.)

Also, I tagged it but will just ask you now since you added this wording:

Where the mains voltages was in the 100-120V range, the limited voltage was only suitable for low-power receivers. Television receivers still needed a power transformer for anode voltages

Now I lived in the US then and DO seem to remember that the last generation of tube TV's DID use direct rectification of the 120VAC just like the 5-tube radios. And that technicians would use isolation transformers to avoid a hot chassis when working on it. But it was also possible that these used voltage doubling power supplies (and silicon rectifiers were already in use) and I didn't have direct experience with them (I fixed the audio gear instead). It's possible I'm only thinking of early transistorized sets which had power supplies running from the flyback transformer (what we'd now call a switching supply) from transistors operating on 150v using a directly rectified supply (and hot chassis). This is all a bit fuzzy now! So try to decide whether you really based that on reliable information (regardless of whether you can find a citation). Ok, that's all :-) -- Interferometrist (talk) 12:25, 26 April 2011 (UTC)

I think we're actually in total agreement about what switching circuits do, but disagree on nomenclature: I think devices specialised, optimised, or mainly used for switching should be distinguished from linear devices. However, examples of switching-only and non-amplifying tubes: Mullard describe the E88CC as being for use in cascode circuits and as a multivibrator and cathode follower in computers. It will work as a linear device, but is clearly identified as suitable for switching. Krytron, sprytron, thyratron. I don't know if the magnetron can be used for amplification. The beam wotsit (the one used for synchronous demodulation by switching the beam between two anodes). Other beam-switching tubes (cyclophon), trochotron. Gas voltage stabilisers/references (I've forgotten what they're called). The [tubedata.tubes.se/sheets/137/7/7AK7.pdf 7AK7] was designed specifically for computer use. I don't know what was actually used by computers I used.

I used a Ferranti Mercury. Paper, not magnetic, tape, and magnetic drum, not disc. The first transistor computer that I know of (a transistorised version of the Mercury's predecessor) was made in 1953; the Mercury was sold from 1957 until about 1960. (This information looked up recently, not from memory). A downloadable manual with lot of information on hardware and autocoder in Spanish. Much of what I "remember" is wrong (always was, not memory failure); I thought it used mercury tube memory, and was very surprised to see it used "modern" core! Autocode (high-level language) was thought to be for wimps; "real" programming was done in machine language (not assembler, there wasn't one, everything used fixed locations anyway). You needed to be aware of the time taken by each operation. If you had the right connections you could hang around in the computer room, the only air-conditioned place available. On of the great feats of those who ran it was a program which made it play simple tunes. Regarding reliability of earlier computers, I only know what I can find now, same as you. I do get the impression that Colossus was pretty reliable, the only reason I can see why the Mercury would have been significantly better would be that it replaced all the diode tubes (50%!) by semiconductor diodes (I don't know if it used special-quality tubes, though it's reasonable to think so if they were available).

Regarding AC/DC TV: in the article on AC/DC receiver design) perhaps a couple of years ago I wrote that there were AC/DC TV receivers (the ones I knew worked off 220-240VAC). I was reverted and criticised by someone from the 100-120VAC area who said, I think, that some circuits ran directly off mains, but a transformer was needed additionally to generate high DC voltages required; I rewrote accordingly. If you find what I wrote in that article at that time (the older versions may retain more than the current one) and its Talk, it might be of interest. I linked the circuit of an AC/DC set with details of how the power supply worked. I only know that I was told there were no true 100-120VAC AC/DC TVs; I have no reference and it's actually someone else's OR! If you have good information contradicting what I wrote (and was told), by all means correct it or delete as appropriate. I think it was said by User talk:Wtshymanski, whom you could contact. He also objected to my saying that 200-240V was more dangerous than 100-120V, so that was deleted, though I did provide references supporting that (which took quite a lot of searching). Pol098 (talk) 15:07, 26 April 2011 (UTC)

Very interesting about the tube computer you used. I guess I was thrown off because I had run across something saying that the first transistor computer wasn't until 1957 (but you say 53) so I added 6 years to that. I very much believe that they'd be using semiconductor diodes (germanium I assume) for making gates since those were perfected sooner than transistors. I imagine with the improved reliability of tubes and the poorer reliability of early transistors meant that transistorization might even have meant a reduction of reliability before a certain point, so that might explain the continued sales of tube computers until 1960 as you say. Interferometrist (talk) 18:30, 26 April 2011 (UTC)

From Ferranti Mercury#Mark I: "One team decided to produce a much smaller and more cost-effective system built entirely with transistors. It first ran in November 1953 and is believed to be the first entirely transistor-based computer." More in Metrovick 950. See also 1953 - Transistorized Computers Emerge Pol098 (talk) 18:55, 26 April 2011 (UTC)

Modern Manufacturers

I've noticed that the Modern Manufacturers section has been deleted by some guy on the grounds that "this articles is about the device, and not about where to buy it". Acceptable reason?? First of all, given the current situation of vacuum tubes perceived by many as somewhat antique technology (specially in non-RF applications), this section was (& is) highly valuable and essentially encyclopedic. Plus, this section was there for quite a long time (years I believe) and nobody never objected. (I wonder somebody completely unfamiliar to the topic deleted it). What do you think? --Kourosh.Alavi (talk) 07:22, 13 May 2011 (UTC)

I'm sorry, but I really agreed with that deletion and the user who did so is an extremely experienced editor who didn't do so without due consideration. I had left it there simply because it had been there for so long, but appreciate that someone did the obvious thing. Lists of businesses are not "encyclopedic" however if someone would want to write a general blurb on the current production of vacuum tubes and where they occur (perhaps mentioning the name of one or more companies, but only where they are particularly notable) I would welcome that. Lists of a commercial nature also have the problem that they attract covert advertising which is contrary to the interests of the encyclopedia and degrade its neutrality, and that is a more important concern than including as much information (even if encyclopedic) as one might. -- Interferometrist (talk) 14:56, 13 May 2011 (UTC)

Thanks for the quick reply, and the info. I'm a newbie but kind of knew what the likely answer would look like, and I'd like to say once again that the mentioned section (in this particular case) was of encyclopedic use (well at least to me, I was not aware who would produce such a thing and where, also for the reason I mentioned in my earlier post). With such a tiny market, and with such unknown manufacturers, there's no reason to be afraid of covert advertising. I believe that was the case during the years the section was there. After all, there are hardly any new tube manufacturers. But with that said, I was not going to change anything, just wondered why it suddenly disappeared after such a long time, I thought a random editor might have done that.--Kourosh.Alavi (talk) 20:10, 13 May 2011 (UTC)

Apparently the 100,000 year magnetic pole reversal cycle has begun (mixed North and South domains in central African and over Oceans). The collapse of the Earth's Magnetic Field and ensuing lack of protection from Solar protons would not bode well for Semiconductor devices (like a giant EMP). Assuming humans survive, we will probably use vacuum tubes. It only took us 1500 years to rediscover the "lost wax" process.Shjacks45 (talk) 09:01, 2 August 2011 (UTC)

Traditionally

1. Getter are supposed/designed to remove Nitrogen and Argon also, like Titanium and Zirconium. (By the way what is the in vacuo vapor pressure of Barium?) Often see blue glow of Argon in older Rectifiers and Beam power tubes e.g. Type 80 or 5U4. Glass tubes especially when hot (please remember that Glass is actually a liquid) rapidly diffuse Helium, and Rectifiers with internal low pressure hydrogen lose that in operation. Ions boiled off the Cathode etc will return to strike the Cathode as high energy positive ions which causes deterioration of CRT emission. 2. Tube amplifier Transconductance was not mentioned. Order of magnitudes lower for vacuum tubes than MOSFET transistors. Exception is grid design in 2C40 type tubes with perforated metal foil grid. "Nuvatron" (in IEEE) was attempt at sealing these tubes in tiny compartments in ceramic sheets. Modern Plasma Televisions are use similar technology for their screens (tiny conducting channels). Research into other display devices has yielded Field emission display, emitting electrons non-thermionicly. 3. Cathode/Filament material could be more elaborate as it applies to modern CFLs and CCFLs. Shjacks45 (talk) 08:52, 2 August 2011 (UTC)

oops I was wearing a mu-mu: re my comment of not including transconductance. please link to other wiki article.

I realized the huge brouhaha about mu, and I just wanted this wiki to be scientifically correct. First Ohm (Ω) is a the unit of resistance. On the other hand, Mho is derived from spelling ohm backwards and is written with an upside-down capital Greek letter Omega: $\mho$, Unicode symbol U 2127 (℧); equal to 1/Ω or in layman's terms, Conductance. A Tube or MOSFET take a voltage input and changes Plate current or Drain current due to that changed voltage: that is called Transconductance. Call me Old but RCA et al have manufactured ICs to be perfect transconductance amplifiers, the RCA product was CA3060. Transfer Conductance. The old unit of conductance, the mho (ohm spelled backwards), was replaced by the SI (International System) unit, the siemens, with the symbol S (1 siemens = 1 ampere per volt). The term "micro mho" or "mu" is used with tubes because except for the likes of 2C40 etc tube transconductance is quite low. Even the oldest MOSFETs (VN66, IR130) are tens of thousands of "micro mho". Or should we end it all and just start talking abouut Siemens? Shjacks45 (talk) 09:44, 2 August 2011 (UTC)