Talk:Mullard–Philips tube designation

Value Ranges

I found systematic differences in the value-ranges of the "Mullard-Philips Tube Designation" sheme compared to the equivalent in the German Wikipedia.

Examples:

• German WP: "1 bis 9 = Außenkontakt-Sockel (5- und 8-polig), auch Oktal- und Europa-Stiftsockel mit Quetschfußaufbau", Englische WP: "1-10 = Various side contacts, octals, specials (exceptions are ECH3G, ECH4G, EK2G, EL3G, KK2G which have octal bases)"
• German WP: "10 bis 19" = Stahlröhren-Sockel (8-polig), ggfs. Quetschfußaufbau, Englische WP: "11-19 8-pin German octal"

It seems, there are differences in both name and numbering shemes for the european tube designation in the WP:EN and WP:DE. While English Wikipedia refers to the "Mullard-Philips Tube Designation", the German Wikipedia refers to the "Philips und Telefunken Gemeinschafts-Bezeichnungssystem" or "Europäisches Gemeinschafts-Bezeichnungssystem". While this can be easily explained by different historic development of this technology in Germany and the UK, it is commonly belived, the schemes are identical. I checked the main source to be sure the value-ranges in WP:DE are correct. I am offering my help to correct the value-ranges in WP:EN if incorrect. To be absolutely sure, we do not have a minor, but systematic deviation between Mu/Ph and Tf/Ph designation schemes I am asking for help to check the source for Mullard-Philips tube designation. --BEG (talk) 09:07, 7 November 2008 (UTC)

I saw some changes being made by Meranthu. But there is still more to check. Any volunteers? --BEG (talk) 12:04, 8 November 2008 (UTC)

Shouldn't "Mullard-Philips" be called "Pro Electron" in the title? In the USA, I've seen "Pro Electron" used, but not "Mullard-Philips". Regards, Nikevich (talk) 22:15, 17 June 2009 (UTC)

No. The Pro Electron was derived from the Mullard-Phillips in 1966 and has a few differences. For example, under Mullard-Philips, transistors used the first letter of 'O' deignating a 'cold cathode' device (as they have no heater) whereas in the Pro Electron system transistors were designated 'A' for germanium, and 'B' for silicon. There was little room for confusion as both 'A' and 'B' as a heater or filament designation were obsolete by 1966. In any case a pre second world war valve and a transistor are fairly easy to tell apart. 86.176.156.193 (talk) 18:03, 4 December 2010 (UTC)

There is a really serious error and an omission in this article. Consider this sentence: "For example an EF80 manufactured as a special quality tube would be designated 'E80F'." WRONG. The E80F is nothing like the EF80. The former is a random unscreened pentode and the latter is a screened RF pentode. My collection has both, and they don't even look the same. Another example: EL81 vs E81L. The former is a horizontal deflection tube with a top cap and the latter is a small internally screened output pentode for telephone equipment. The physical difference is huge. See respective datasheets.

My point is: Fix the article. The rule is correct (number before tube contents) but a better example would be ECC81 vs. E81CC (I own both types) and mention that the rule does not always hold. Signed, someone who ruined an E80F by sticking it into a socket meant for an EF80. —Preceding unsigned comment added by 83.145.209.56 (talk) 14:38, 7 January 2011 (UTC)

Sorry to dissapoint you, but the article is correct. The special quality version of the EF80 is indeed designated E80F. Similarly the special quality version of the EL81 is E81L (see here here. It may be that another manufacturer has produced a tube with the designation E81L, but if they have, I have no record of its existence (but that doesn't mean that it doesn't exist). The SQ type rarely looks the same as the regular type as there are considerable differences in construction, usually related to the manner in which the electrode structures are supported. Incidentally, the EF80 most definately is a screened RF pentode (see here 86.181.51.84 (talk) 13:28, 9 April 2011 (UTC)

Although the E80F is a SQ tube, and even if some documentation refers to it as a Special Quality version of the EF80, it remains a bad example because the maximum ratings of the E80F are w-a-y lower than the EF80 (e.g. Pa=1.3W instead of 2.5W; Pg2 of 0.4W instead of 0.7W and Ik less, Vg2 less, and characterised at different currents and voltages... as well as being described for different applications and looking different!). Maitchy (talk) —Preceding undated comment added 05:23, 31 December 2012 (UTC)

Not so. All SQ tubes have maximum ratings below that of the equivalent Non-SQ valve. This is one of the means by which the reliability is increased. There are other means. And of course the described applications were different. No one in their right mind was likely to put an E80F costing between £30-50 in place of an EF80 (costing £0.55 - at 1968 prices) in their TV set. Similarly, the army would not replace the E80F (or CV2729 as it would have been) in their radio sets with a common or garden EF80 (CV1376) and hope it survived the arduous conditions encountered in whatever battle they are fighting. — Preceding unsigned comment added by I B Wright (talk • contribs) 16:34, 2 January 2013 (UTC)

P.S. The EF80 vs E80F is covered in the List of vacuum tubes article (note there is quite a bit of overlap in other ways, and a bit of a tidy-up of the two articles would be good). Maitchy (talk) 05:49, 2 January 2013 (UTC)

Nikevitch's question above for correcting the title was answered with a plain "no". Still the issue is that the Mullard-Philips tube designation in the pre-pro-electron phase was in fact defined between Philips (with it's local subsidiaries Mullard, Valvo etc.) and Telefunken, who are not mentioned at all here ( just realized, someone with a US-centered view now removed the Philips' subsidiaries Valvo (Germnay) and Dario (France)...). --Wosch21149 (talk) 22:56, 29 February 2012 (UTC)

Not so much US-centered as someone without the lung capacity of an opera singer or pearl diver - the sentences in this article get extraordinarily long. Valvo and Dario have no visible influence on the history of the numbering system of tubes (as far as we can tell in this article). They seemed to be odd digressions to put in the lead. Anyone who wants to read about Philips corporate structure can read the Philips article. --Wtshymanski (talk) 15:00, 1 March 2012 (UTC)

What does the 'CV' stand for.

Some kind soul keep changing the reference to CV valves to 'common valve'. This is in fact a common error. I have worked with CV valves all my working career working, as I did for the UK Ministry of Defence since the day I left school. 'CV' is the abbreviation for 'civilian valve' and refers to a device that was designed and developed outside of the military field (usually by a commercial valve company) that has been adopted for military use. Valves that were developed inside the military environment (usually by a government owned R&D establishment were given initially an 'E' designation (for 'experimental') and when perfected, an 'M' designation (for 'military). To add to the confusion, when unified part numbering schemes were adopted such as the A.M. Ref no. scheme (Air Ministry part numbers) and later the NATO stock numbering scheme, the stock number was based on the CV number. Consequently, all in service 'M' designated valves were given a 'CV' number so that they could be incorporated into the numbering schemes.

For example, an rugedised high reliability R.F. pentode was developed by one such government R&D establishment and was given the designation 'M8084'. It was placed for manufacture by the Mullard company, who produced a commercial version of it under the designation 'EF91'. In order to incorporate it into the A.M. ref no. system it had to be given a CV number (mainly because the system insisted that rugedised reliable valves had to have a 4000 series number, but mainly to avoid duplication of numbers from two unrelated schemes). The CV number therefore became CV4014 (the next number off the rank). The valve therefore became 'A.M. REF No. 10CV/4014' and subsequently NATO stock no. '5960-99-000-4014' [1].

Life was made even more bizarre because with the introduction of semiconductor devices, they too were enrolled into the scheme. Thus a CV7010 although having a 'civilian valve' designation, is actually a silicon rectifier (NATO stock No. 5960-99-000-7010: the Air Ministry having long gone the way of many government reorganisations).

[1] For the curious: '5960' means 'Valve, Electronic'. 99 means the code was generated by the U.K. (America uses '01'). '000' is the sub class of which 'Valve, Electronic' has none. And '4014' is the recycled CV number. 109.145.22.224 (talk) 16:45, 14 April 2012 (UTC)

This would be a great addition to the article if you could please find some references for it. There's got to be some reliable tube-collector's handbook that explains British numbering schemes. --Wtshymanski (talk) 16:29, 15 April 2012 (UTC)

I do have my old training notes which cover the point very well. Unfortunately, they are Ministry of Technology (which the Air Ministry became before it in turn became part of Ministry of Defence) training notes and were only used internally for that ministry's training and are thus very unlikely to meet the Wikipedia requirement for verifiability (Unless anyone else has a copy or something similar that is). 109.145.22.224 (talk) 13:53, 17 April 2012 (UTC)

Same number, different heater and characteristics

Removed from article:

The PL84, HL84 and UL84 varied from the EL84 in more than just heater voltage/current; their characteristics were altered to make them more suited to the lower high tension voltages found in typical television sets where series heater chains would be employed (although still giving comparable output power when used as an audio amplifier). Their RETMA type numbers were 15CW5, 30CW5 and 45B5 respectively. The corresponding 6.3V heater and 0.45A series heater versions were the EL86/6CW5 and LL86/10CW5 respectively.

The contributor of this piece has confused characteristics (from data sheets) with 'typical operating conditions' as found in pocket data books. All the *L84 have identical characteristics (apart from heater voltage and current). See here, here and here for the EL84, PL84 and UL84 respectively. Examination of the characteristic curves shows that all three devices are electronically identical. Pocket data books provide operating parameters that reflect the typical usage to which the device is put. For the EL84 this would be as an audio output valve which is operated typically at a V_a = 250 volts. The PL84 is more likely to find itself used as a TV frame output valve with V_a = 170-200 volts and data books give typical conditions to reflect that usage (with the reduced HT voltage actually used in AC/DC TV sets). The EL84 could just as easily operate at that anode voltage and vice versa. Similarly, the UL84 often has typical operating conditions in pocket data books showing a V_a of 100 and 200 volts. This is not because the valve is any different electronically, but reflects the usual use to which that valve is put (i.e. audio output in an AC/DC radio designed to run from 115 volts or 230 volts respectively). All three valves can operate with any anode voltage up to 550 volts (the specified maximum for all of them). No manufacturer is going to establish 3 different production lines for what is essentially the same valve. All the L84s were assembled on the same line with the heater being changed to suit the exact valve being produced at that time.

I can find no reference to the existence of an HL84 (or any H*80 series valve except an HM85). I B Wright (talk) 15:57, 2 January 2013 (UTC)

Sorry, but... this time you are wrong. I made the same mistake at first, because it is very common for the whole family to be produced with only heater differences, as you mentioned, but the PL84 and UL84 were redesigned with different cathode geometry, etc. See http://www.diyaudio.com/forums/tubes-valves/154372-pl84-tubes.html and various other comments on the web you can find from a google search for EL84 and PL84 (e.g. the following helpful exchange:

"But isn't the convention that the first letter just differentiates the heater voltage, when the rest of the designation is the same?" Convention hasn't really been established sufficiently for this to be a reliable guide. Usually you're right, but this is an exception (and the DM71 / EM71 is another striking exception, as you point out). As well as different heaters, the EL84, PL84, UL84 have different cathode areas, and probably different grids, for their differing gm's etc. The anode is, as far as I know, the same, so they are a bit difficult to tell apart if the markings get rubbed off."

My April 1969 Philips "Electron Tubes" handbook, part 4, gives different limiting values for the PL84 (and the UL84) - such as a higher I_k and slightly lower anode and g2 voltage and g2 power ratings; the PL84 and UL84 involved changes to the EL84 design in noticeable ways for the type of work they were doing, and the EL86 was (I guess) later produced as the 6.3V heater version of these.

It is possibly interesting that the US (RETMA) devices mentioned in the 14th Ed of Sylvania Technical Manual (reprinted by Audio Amateur Press) are the 6BQ5/EL84 and 8BQ5 and 10BQ5 all mentioned together on page 124, while the 6CW5/EL86, 8CW5, 10CW5/LL86, 15CW5/PL84 and 30CW5 are mentioned on page 167 with different ratings (slightly different again to the Philips PL84 even!). Maitchy (talk) 05:52, 3 January 2013 (UTC)

And oddly, I had meant to write HL86 since it is mentioned in the October 1974 (3rd revision) Babani "Handbook of Radio, TV, Industrial & Transmitting Tube & Valve Equivalents" as an equivalent for the 30CW5. It also mentions an XL84(=8BQ5) and an XL86(=8CW5). However the HL84 can be found on the Internet easier (I tried to put tinyurl dot com /HL84-86 here but it seems it is black-listed by wikipedia.org), e.g. finding on the very good radiomuseum.org website at: www.radiomuseum.org/tubes/tube_hl84.html Maitchy (talk) 05:52, 3 January 2013 (UTC)

The reference you provided is a forum which are notoriously unreliable as sources of information - as betrayed because the contributors cannot agree.

I had hoped to find the three data sheets from the same manufacturer, but alas I could not. Some differences between manufacturers in data sheets are not uncommon, this is usually down to how the characteristics are measured rather than any real difference. If there is likely to be any difference in the three types it will invariably be in the heater/cathode assembly (the rest of the valve being made on the same production line). The characteristic curves shown in all three of the data sheets I referenced are (to all intents and purposes) identical. The limiting values of the three are also practically identical. The one difference is that the EL84 shows an I_k max of 65 mA, whereas the PL84 and UL84 have a I_k max of 100 mA. This doubtless comes about because the cathode has to accommodate the longer heater wire required for the P and U variants and thus is most likely slightly larger. This will be driven by necessity rather than any desire to change the characteristics. But the actual I_a/V_a against V_g characteristics themselves of the valves are completely unchanged. This means that (subject to changing the heater power source) that both the PL84 and the UL84 can be exchanged for an EL84 operating at 250 volts without any other circuit modification. The reverse is also true provided the I_k max limit of the EL84 is observed (which it almost certainly will be if you do not want to exceed the P_a max limit of 12 watts). I B Wright (talk) 13:20, 3 January 2013 (UTC)

There are some problems comparing datasheets, but still the PL84 and UL84 should be thought of as "other heater" versions of the EL86/6CW5, not the EL84. There is enough data from databooks to back this view, over and above the little differences you'd expect to get between manufacturers and so on. The fact the EL86 exists is the clincher, I think. If you don't agree then take note of the way the US number allocations went (what gets a -CW5 and what gets a -BQ5). If you try to explain away differences in characteristics too much you end up having to say an EL86 is an EL84. :-) The 3 datasheet pdf's you gave links for, and my data books, don't have easy-to-compare plate curves for the same Vg2, but from the perfect match in maximum ratings the EL86 is obviously the closest family member to the PL84 and UL84... the chat on various forums (and many conclude the EL84 and PL84 differ) may or may not be accurate in suggesting the EL86 came from the PL84, and it is reasonable to say the PL84 and EL84 characteristics are pretty darned close, but still they are different enough to say they haven't simply changed the heater resistance. I wouldn't mind building an EL84 circuit substituting PL84's and a different heater supply, but I wouldn't expect it to sound exactly the same - not like PCL86's in an ECL86 design for example.

It isn't the only naming quirk, but one worth noting in the main article (but with different wording perhaps). And I'd say the E80F vs EF80 differences (vastly different ratings, characteristics, applications and even different base pinning diagram) that it should be noted as a trap-for-young-players as well. I wouldn't go so far as to include the DM71/EM71 differences since the directly-heated types always are different to the indirectly-heated ones (although some sentence to the effect might be a good idea for the benefit of those who don't see how obvious it is :-) ). Maitchy (talk) 22:42, 4 January 2013 (UTC)

As I explained elsewhere, the different rating for the SQ version of a valve comes about because the SQ version is required to meet minimum MTTF targets. By stressing the valve less it lasts longer. The different pinout is not a deliberate attempt to make the valve different in any way. It is merely the result from having to support the electrodes in a more robust manner (to withstand greater vibration etc.). The improved supporting sometimes necessitates a revised pin out arrangement. Generally the SQ version of a valve will have the same pin out if it is possible. As you note the E80F does have a revised pin out compared to the non SQ version, the EF80. On the other hand, the SQ version of the EF91 (the M8083) has exactly the same pin out and one can be substituted for the other in applications that do not demand the SQ version. Indeed, I have a sine wave oscillator with 2 M8083s and a 300 volt variable power supply with one in lieu of the specified EF91s. Sometimes, you get an SQ version of a valve that is completely different. The 12E1 (a series pass beam tetrode) has an SQ equivalent, the S11E12. Not only are the valves a diferrent physical size, but the construction is sufficiently different that the former has a top cap connection to the anode, but the latter does not.

If you want to substitute a PCL86 for an ECL86 and want it to sound the same, you would need to take care that the two valves are built the same way. Mullard (or Philips) produced ECL86s where the pentode part really was a pentode, but also produced almost equally as many where the pentode was actually a beam tetrode. They did this for many pentode designs. The characteristics were different but this was not reflected in the data sheets.

Any comparisons that you make between data sheets cannot be used to support a claim in the article that E, P, U or whatever versions are in any way different because that would be synthesis and violates the WP:SYNTHESIS rules because the conclusion is yours not that of the data sheet producer. Of course, if you can find a data sheet for (say) a PL84 that specifically states that the characteristics have been deliberately changed from the EL84 vesion, then that would be a valid supporting cite. I B Wright (talk) 12:25, 5 January 2013 (UTC)

Some considerations on the previous discussions:

1 – In no way the E80F and the E81L tubes are SQ versions of the EF80 or EL81 tubes, respectively. They differ too much in their characteristics, their maximum ratings, their pinouts and the purposes they've been designed for. Even in their visual approach they’re very different, e.g. by comparison of photos of a typical EL81 (http://www.mif.pg.gda.pl/homepages/frank/images/Mul/e/EL81_MUL_HM.jpg) and a also typical E81L (http://www.mif.pg.gda.pl/homepages/frank/images/Phi/e/E81L_Phi_IMG_8198_FP.jpg).

2 – Decreasing maximum ratings might have been a manner to obtain longer life for SQ tubes. The Russians often did so. But this, by far, cannot be considered as a general rule. The E84L (http://www.mif.pg.gda.pl/homepages/frank/sheets/009/e/E84L.pdf), e.g., which indeed is a SQ version of the EL84 (http://www.mif.pg.gda.pl/homepages/frank/sheets/010/e/EL84.pdf), has extended ratings for all three mean parameters, as of plate voltage, plate dissipation and cathode current.

3 – Besides their heaters, the EL84 on one hand and the PL84 and UL84 on the other are, especially by comparison of their plate curves, different tubes! You’re simply not (W)right here! How else could you explain the obvious fact that one tube (PL84, UL84) reads, at the same plate and control grid voltages, but at a lower screen voltage of 200V vs. 250V, very much higher plate currents than the other one (EL84), other than both are electronically very different? At least in Europe, from the early beginnings (e.g. CL4 vs. AL4) series heater output tubes for radio sets have been developed contemporaneously to their parallel heated versions with the same maximum plate dissipation. They were given, in comparison to their parallel companions, a lower internal resistance and a lower screen-to-grid amplification factor (µg2g1). Thus they could cope more easily with the usually lower plate voltages in transformerless AC/DC radios, providing sufficient output power even at 100V plate/screen voltage. This applies to all series heated valves for radios that were developed simultaneously to the parallel heated variants (CL4 vs. AL4, UL12 vs. EL12, UL41 vs. EL41, UL84 vs. EL84 etc.). So we can say that EL84 vs. PL84 and UL84 share nothing else than three characters within their numbers and the maximum plate dissipation. Also the look of their internal structure differs significantly, PL84 and UL84 being more slim than an EL84 from the same manufacturer.

But there are exceptions as well: ECL82, ECL86, EL36 and EL81, e.g., indeed are electronically identical to their series heated P versions. And there are much more examples, especially in tubes for TV sets.

Conclusion: I dare to claim that two tubes having different numbers in fact have to be considered as two different tubes. Similarities can occur, but sometimes are not more than accidentally. For closer information on each tube we have to carefully read the data sheets, especially the plate curves. --Menrathu (talk) 18:48, 30 July 2013 (UTC)

If you wish to compare physical appearance of valves that are notionally the same, you need to find valves that are not only from the same manufacturer, but the same factory and the same era. If these three condidtions are not met, it is quite likely that even valves with the same type number can look different.

Your claim that the EL84 and the PL84/UL84 being different sizes is a nonsense. Comparison of the Philips datsheets for the three types shows that are all exactly the same size - 77 mm high by 22 mm diameter.

Yes, the envelopes' size is the same for both lines of tubes. But did you read carefully what I've written? The point is the different visual appearance of the internal structures, i.e. the plates, which are more slim for the PL/UL84 than for the EL84.--Menrathu (talk) 10:20, 14 March 2015 (UTC)

Here are some examples, to point out what I've meant: PL 84, EL86, and EL84. For comparison: The PL84, the left hand sided EL86 and both right hand sided EL84's are made by Philips or one of it's subsidiaries. --Menrathu (talk) 17:49, 22 March 2015 (UTC)

The interesting bit that you wrote was your observation that the E81L was not a special quality version of the EL81. Indeed an examination of the two types shows that the EL81 is a power pentode for line time base use, whereas the E81L is a signal high slope pentode. This would suggest that the later type number has not been derived from the Mullard-Philips numbering scheme. I eventually found a data sheet, and the clue was at the very top. This is a special long life valve designed for use in telephone equipment. The telephone industry, which has always had its own way of numbering parts, has used various numbering schemes usually based on serially allocating letters and numbers without their meaning very much. An 'L' appended to the part number identifies any valve that has a design MTTF exceding 10,000 hours. The potential for confusion between two numbering systems that yields the same valve number is obvious (as has been demonstrated here) but in this case, luckily a special quality version of the EL81 was not produced. –LiveRail < Talk > 15:02, 10 February 2014 (UTC)

Interestingly, I have a few old volumes of the Mullard Technical Handbook (essentially a designers' guide to using valves). Volume 1 covers the normal run of the mill line of valves, and the data sheet for the EL84 does has 'characteristics' that differ from the PL84 and the UL84 (most likely because they are quoted at different anode voltages. Also the limiting cathode current is larger for the PL84 and UL84 than for the EL84 (100 mA versus 65 mA) though the output power is lower for the UL84 (5.3 W versus 6.0&nbspW for the EL84). The output power of the PL84 is different yet again, but as its intended application is for a frame output pentode, it would not be comparing like with like. The interesting thing is that neither the PL84 nor the UL84 have any characteristic curves on their respective data sheets. Both refer the reader to the EL84 for the characteristic curves.

I also managed to dig up a data sheet from a company called Tesla (apparently defunct). While I have no idea what language the data sheet is written in (possibly Hungarian or somewhere around that part of the world), the data sheet adresses the EL84 and the UL84 on the same sheet, apparently noting the different applications and limiting values, though there is only one set of curves. It also seems to be common practice for US valve suppliers to publish one data sheet to cover all the various filament/heater arrangements (e.g. this one). DieSwartzPunkt (talk) 11:23, 11 February 2014 (UTC)

Formatting and Consistency

This article on the Mullard-Philips naming scheme needs to be able to be read in conjunction with similar articles such as on other naming schemes (e.g. RETMA tube designation and especially Pro Electron) and to some extent mesh in with the List of vacuum tubes because that duplicates some of what is said here and because it is sensible to use examples to illustrate the naming conventions. I think all of these articles could be tidier, have less duplication, be easier to read, and follow more consistent formatting patterns.

So to kick off some discussion on this (and it is obviously better to get some agreement before mucking around with things) I will suggest a few thoughts here, and then look to get some talk going for other articles. (Which ones? That needs discussion too).

1. Section names could be the same, and in the same order between all articles relating to tube (and transistor) designations; e.g. this article has a section "2.1 Further patterns and exceptions" while RMA tube designation has a section "1 Oddities" doing a similar job in a different position in that article, with a different name.
2. The List of vacuum tubes has a section "European designation (with American equivalents)" that duplicates a lot of this article
3. There is difficulty getting across important information to the reader, because the information being written about is messier than we can deal with simply by quoting some book... I have several references that describe the Philips-Mullard system (calling it all sorts of different things, and with details that vary because some filament types came and went over time, and because "rules" were broken in reality). Also, at present, the wording could give the reader the impression that the type letters and digits defines the device more than it really does - we have to keep saying things like "just because an initial K meant a 2 Volt battery filament and a second letter T under one (later) version of the scheme meant a thyratron, it doesn't mean to say a KT66 is going to be a 2V thyratron!". I suggest a good way to deal with the reality is to have a table of what types were available (a bit of work!) rather than try to use words to refine rules.
4. Having said that, I introduced into the article a series of "Historical progression" pictures of the 'standard' that corresponded to how many digits were in the name (being easier to pin down than dates, although it does correspond with dates), because it is misleading to give a list of all (say) first-letter heater types that were ever allocated and put them beside a list of all number (socket and serial number) code patterns, because they did not all occur at the same time. The down-side to doing this is that the original references over the years use messy phrases when they mention what letters have been discontinued (and it will not reflect the exact date they were discontinued - probably decades afterwards).
5. Formatting the output is important for presenting information and this example (although crude) is better than what is being used in the article at the moment... can we do better still? This is from the Pro Electron article:

• See Mullard–Philips tube designation for details. A brief summary of just the more common letters is:

    ECC81
   /  \ \\__ last digits give serial number
  /    \ \__ first digit(s) indicate base (3=octal, 8 or 18 or 80=Noval (B9A), 9=7-pin(B7G).
 /      \___ one letter per valve in the tube:
D=0.5-1.5v          A=single-diode (low power)
E=6.3v*             B=double-diode (usually shared cathode, but not always)
P=300mA             C=triode
U=100mA             F=pentode (low power)
                    L=pentode (high power)
                    Y=Single-phase rectifier
                    Z=Full-wave rectifier
* Note: some 6.3 Volt heater types have a split heater allowing series (12.6Volt; the
  default for Noval pins 4 to 5) or parallel (6.3Volt) operation.

Further thoughts, anyone? Maitchy (talk) 20:30, 5 January 2013 (UTC)

Just because a type number appears to conform to the Mullard-Philips system (or indeed any system) does not necessarily mean that it does. The KT66 numbering that you mentioned above was not part of, nor derived from the Philips-Mullard scheme. It was an entirely separate scheme used for (what was at the time) an entirely new type of valve developed by the MO Valve company. In this case, the letters 'KT' simply meant 'kinkless tetrode' and were used to identify the (then) new beam tetrodes that were developed by that company.

Any article in Wikipedia is always capable of improvement. If you believe that you can improve the presentation of the information in this article then go right ahead and have a go. However, as this is an encyclopedia, the meanings should not be limited to the common types. All variants should be documented. Your list has a couple of errors. The first letter 'D' does not mean a 0.5-1.5 volt heater. It is officially documented that it represents a heater or filament operating from 1.4 volts or less. Also the second letter 'B' is a double signal diode specifically having a common cathode. I B Wright (talk) 15:38, 6 January 2013 (UTC)

With the "KT" example I was trying to say it looks like it should be part of the system, but isn't. Other examples are: AA61, AA91E (well,that's a variant of EAA91 but...), AB150, AB4, a whole lot of Arcturus types like AC22 to AC46 (that happen to be triodes quite often, as it happens). And the list goes on and on. The point I was trying to make was that it is hard to come up with a written rule to say when a type number that matches the rules for the Philips-Mullard(-Telefunken/etc) or Pro Electron naming schemes is actually part of that series 9and so the heater voltage, triode/pentode/etc is predictable). Plus... the manufacturers played games with the naming system, breaking the rules a bit to start with (e.g. when is a tetrode a tetrode? And AA vs B for diodes with separate cathodes). hence the conclusion: a table of data is a far better way to go. And that requires a bit of imagination (or experimentation) though.

I'm sure you are (W)right about the D being "1.4V or less"; even though I don't think any "D" types with filament voltages below 0.6V are in my databooks, the standard (in the 1950's and 70's at least) said "up to 1.4V". I'll change it in the original article. But I have to leave the usually in the description about the "B" for double diode because there are exceptions. The references that say what the conventions should be are simply not correct all of the time... and the standard has changed over time (both in what types were produced and what the Philips books themselves said were the rules). Maitchy (talk) 01:24, 7 January 2013 (UTC)

Confusion reigns as ever here. All my sources show that the 'D' designation in the Mullard-Philips scheme designates a filament or heater voltage of 1.4 volts or less. As you note, filament voltages (were there any 1.4 volt or less indirectly heated valves?) were seldom, if ever, less than 0.6 volts. However, in the Pro electron scheme, the 'D' designation is defined as between 0.5 and 1.5 volts. In practice, it makes little difference as both encompass valves that are generally encountered, but this is an encyclopedia and the correct definition should be given at each article. I B Wright (talk) 16:39, 7 January 2013 (UTC)

Yes, there were! Think of all the diodes designed for rectifying the CRTs'high voltage in b/w TV sets, such as DY80, DY86, DY802 etc., with their cathodes being heated by a separate heater. The fact that their cathodes were strapped to one end of their heaters doesn't affect that they indeed were indirectly heated. --Menrathu (talk) 19:28, 31 July 2013 (UTC)

Possible way to show historical progression (and tackle Pro Electron differences)

The differences over time (and especially when moving from Mullard-Philips to Pro Electron) that I and IB Wright mentioned above should be made clear to the reader, but what is the best way? I already started with subsections based on the number of digits (a back-door method of slicing the rules up by era). That might be better done using tables... but rather than muck around with the main article page I'll put something below for comments first.

Also to consider: having a succession of sets of rules might be no more help than one big, complicated rule set if the reader is coming to Wikipedia with the aim of finding out what some type number (e.g. ABC345) means. Well, the first response to that might validly be "it isn't a good idea to read through the various articles on naming schemes until you find one that mentions a naming pattern that matches the type you are interested in!", but I suspect we can still go some way to clarifying the facts for such a reader.

It also raises the question: how do we know that such-and-such a type number was made under this or that naming scheme? Just because it has letters followed by digits isn't enough to say it is Mullard-Philips. We could list exceptions (the KT example is in quite a few different articles on vacuum tubes and their names! e.g. in Vacuum tube#Beam power tubes and again in Vacuum tube#Names; in List of vacuum tubes#Other letter followed by numerals and List of vacuum tubes#KT as well as in this article: Mullard-Philips tube designation#Double-digit numerical sequences).

One option

Single-digit numerical sequences

e.g. : K 1st Letter	F 2nd/subsequent letters	4 number	G optional letters
(Filament or Heater Ratings)	(type classification)	(serial number)	(variant)
A: 4Vh AC B: 180mA DC C: 200mA AC/DC D: up to 1.4Vf battery E: 6.3Vh heater K: 2Vf battery type	A: Single Diode B: Twin diode C: Triode (low power) D: Popwer Diode E: Tetrode F: Pentode (low power) H: Hexode or Heptode K: Octode L: Power Pentode M: Tuning Indicator X: Gas-filled Rectifier Z: Full-wave High-vacuum Rectifier	Pinch type construction valves fitted with European 5-pin (V base) or 8-pin (P base) side contact bases or international octal bases with European basing connection sequences. Single-digit numbers imply some 1930's European pre-octal base such as B5 or Ct8.	blank: early European bases G: Glass Octal version A,B,C: other variants X: no top-cap

Double-digit numerical sequences Table with heater options A/B/C removed (as they were obsolete around then) plus new ones like "P" that only ever had 2-digit sequences or longer. The number column would then have a list like:

• 11-19: Europrean type metal valves and glass valves fitted with European metal bases
• 20-29: All-glass valves fitted with 8-pin Loktal type American bases.
• 30-39: Pinch type construction valves fitted with international octal bases with American basing connection sequence.

etc...

Advantages and disadvantages

1. Similar to what is there already
2. makes it reasonably clear that only some heater and number sequences (bases) go together
3. does allow both Pro Electron naming schemes (normal 2/3-digit and professional 4-digit) to be described this way

but...

1. the change-over from the "original" system happens with (in theory at least) 10 included in the early version; the method of having the number sequence indicate the base starts at 11, not 10.
2. the change-over to Pro-Electron happened while there were still 2-digit sequences being allocated; there would have to be two tables for Double-digit sequences.
3. Hard to match official documentation to the tables (they may or may not say certain heater types were discontinued at any given time, but the real test is whether any were being allocated).

Another option

Have one big table with all the heater letters ever allocated, all the different uses of the type classification letters, all the number ranges of 1,2,3 (and 4?) digits (i.e. pretty similar to what we have now in the article, and repeated with a few errors in List of vacuum tubes#European designation (with American equivalents), but in tabular form as above)... but.... use something like different background colours to indicate the eras in which they were used.

Background colours are not always a good idea - not everyone can see them. There are probably other ways of marking groups of letters/digits as belonging to an era (e.g. I used asterisks) but it is messy to go too far with adding detail that way.

Yet another Option

Make little attempt in the tables (perhaps still have 3 historical ones?) but handle the "invalid combinations" by listing what type numbers were allocated. This allows us to stick closer to official references in defining the standard, and the examples aid in highlighting where they bend the rules in practice, and most of all, it could help readers that wish to know about specific types to see what is relevant.

A very primitive example (with errors, but hopefully enough to show the idea):

		Directly Heated		Parallel (Voltage-Source) Indirectly Heated		Other	Series (Current-Supplied) Indirectly Heated						No Heater
	Diagram	up to 1.4V DC	2V	4 Vh	6.3Vh (or 12.6Vh with centre-tap)	Vh/Ih	600mA Series	450mA Series	300mA Series	200mA Series	150mA Series	100mA Series	Cold Cathode
single diode	B7G	DA90 DA101			EA960 EA961 EA962
	Rimlock				EA40
	other	DA50			EA41 EA50 EA52 EA53 EA71 EA76 EA111 EA766	GA560 SA100 SA102
Single Triode	B7G	DC90 DC93 DC96 DC193			EC91 EC92 EC93 EC94 EC95 EC96 EC97 EC98 EC900 EC903		XC95	YC95 LC900	PC92 PC93 PC95 PC96 PC97 PC900			UC96
	Rimlock				EC40 EC41
	Loctal	DC25
	B9A	DC80			EC80 EC81 EC84 EC86 EC86C EC88 EC806S EC866 EC8010 EC8020		XC88	YC88	PC80 PC86 PC88			UC88
	IO			KC3	EC31 EC360 EC362 EC55 (coaxial)
	other	DC11 DC51 DC70 DC760 DC761 DC762	KC1 KC3 KC4 KC50 KC51	AC2 AC50 AC100 AC101 AC102 AC104 AC701 AC761	EC2 EC50 EC52 EC53 EC54 EC56 EC57 EC70 EC71 EC157 EC162 EC561 EC760					CC1 CC2
Double Triode	B7G	DCC90			ECC90 ECC91 ECC91W ECC99 ECC900 ECC960 ECC962
	B8A				ECC40
	B9A				ECC81 ECC82 ECC83 ECC84 ECC85 ECC86 ECC87 ECC88 ECC89 ECC180 ECC186 ECC189 ECC282 ECC801 ECC802 ECC803 ECC804 ECC805S ECC807 ECC808 ECC812 ECC813 ECC863 ECC865 ECC8100		XCC82 XCC89 XCC189	LCC189 YCC89 YCC189	PCC186 PCC805		HCC85	UCC84 UCC89
	IO				ECC31 ECC32 ECC33 ECC34 ECC35
	other				ECC171 ECC230 ECC70
Triode plus Pentode	B9A				ECF80 ECF801 ECF802 ECF804 ECF805 ECF82 ECF83 ECF86			LCF80 LCF802 LCF86
	IO		KCF30
	other	DCF60			ECF1 ECF174 ECF200 ECF201			LCF201
Triode plus Hexode or Heptode	B8A				ECH113 ECH41 ECH42 ECH43							UCH41 UCH42 UCH43
	B8G	DCH22 DCH25			ECH22 ECH71							UCH71
	B9A				ECH80 ECH81 ECH83 ECH84	HCH81						UCH81
	IO	DCH31			ECH32 ECH33 ECH35 ECH3G					CCH35
	other	DCH1 DCH11	KCH1	ACH1 ACH1C	ECH171 ECH200 ECH3 ECH3F	BCH1				CCH1		UCH5
Triode plus Output Pentode	B8A				ECL113
	B9A				ECL80 ECL802 ECL805 ECL81 ECL82 ECL83 ECL84 ECL85 ECL86		XCL82 XCL84 XCL86	LCL84 LCL85 YCL82 YCL84 YCL86	PCL80 PCL81 PCL82 PCL85 PCL86 PCL801 PCL802 PCL805 PCL88			UCL81 UCL82 UCL86
	other				ECL200	VCL11

There can be little doubt that there is considerable room for improvement in the manner in which information of this type is presented (the current effort having much redundancy). It is also clear that you have put a great deal of thought and effort into coming up with options which is highly commendable. It might be of interest to note that I have a few references detailing the in's and out's of several nomenclature systems, but they all have the common feature that the options are just presented as a series of lists in the same manner as the article. One aspect that needs to have an eye kept on it is not to overwhealm the essential information with too much side detail.

I would prefer that the article did not go as far as option 3. We already have a List of vacuum tubes article, that I do not believe fulfills any valuable purpose (and it is not particularly well laid out anyway). Your option 3 is attempting to provide another list of valves which we do not need. A simple table along the lines of your option 1 would seem to fit the bill. However, trying to arrange the table to show what letters were and were not used together is asking for trouble because whatever you include, other editors will keep finding obscure valves that expands the list. Where does one draw the line? Even the Mullard-Philips scheme as generally documented has its quirks. One such quirk is the Philips EF8. A 6.3 volt heatered pentode with a Ct8 side contact base. Except that it has FOUR grids?!? And, no, it is not a hexode mixer. I B Wright (talk) 14:27, 9 January 2013 (UTC)

References

How do sources handle this?

How do the sources handle the description of the Mullard-Phillips designations? And surely we could leave the ProElectron descriptions to that article. Is it necessary to spend a lot of effort here to describe what is fundamentally a list of part numbers assigned by the manufacturer for their convenience? Give the gist of the scheme, warn that it evolved over time, warn that there are quirks and exceptions, list any of these quirks that are particularly notable...there's your article. --Wtshymanski (talk) 15:11, 9 January 2013 (UTC)

I completely agree about Pro Elecron. This article merely has to note that the Pro Electron was derived from this system and leave it at that. I B Wright (talk) 12:55, 10 January 2013 (UTC)

I partially agree, but what I am thinking of is that there are two ways this article could be viewed:

1. as an historical account of how the designation system evolved, in which case the different versions of the scheme, and when they occurred, is important;
2. to help people understand what a particular type number is describing, in which case having multiple tables giving "snapshots" of what the system looked like at different times is just going to make it harder to read. And for this type of user of wikipedia, having the 3- and 4-digit information in the same place as the 1- and 2-digit iterations of the scheme is very helpful, as is the inclusion of some mention of the potential ambiguity with "A","B" and "O" prefixes for semiconductors, even though this article is aiming to talk about tube designations.

That said, the division of the information into tables for 1-digit serial numbers, 2-digit serial numbers, etc up to either 4-digit or whatever Philips go up to by the time the system became Pro-Electron can be easy to find what you want and relates to the historical progression in a reasonably useful way.

The question of what the sources had opens up a can of worms: no one source covers the entire system properly; I have quite a few that display the system with either lots of information missing because that part of the system came later, or has comments like "The use of letters A(4V), B(180mA),... has been discontinued" and may still can miss some details of the scheme other versions had. From my point of view, it isn't helpful to have one big table with all possible heater letters, all possible 2nd/etc letter options and all base/serial number options, because some combinations never occurred together, and because the table would be uncomfortably big (although maybe not?), and because our synthesized conglomeration takes us a long way from the original source data.

I am thinking of (and still playing around with what it might look like before presenting it for discussion, but...) an approach that addresses the needs of somebody approaching this with a question like "what is an XYZ123?" but also is good for encyclopedic treatment of the topic. It also tackles the duplication of material between this article and the list of tubes. Basically what I am suggesting is to have one article listing tubes by name in a table with links to articles on the naming schemes where appropriate, meaning subsection names within this article and the RETMA, Mazda, etc tube designation articles should follow a pattern (not very much different to the present situation though), and people looking for information should be able to find it easily (at present I feel the reader has to pretty much know the answer already to know where to look!). I will follow up with examples for discussion (before actually changing articles) as soon as I can. Maitchy (talk) 23:17, 5 February 2013 (UTC)

Four-digit numerical sequences

What's the source for this information? The German RTT tube vademecum says special-quality tubes were issued four-digit numbers, the first digit of which denoting the base, after the function/base swapping ceased. I see this holds for the 8000's, while 1000's normally have wires.

Examples: EC1000/1030/1031, EC8010/8020, ECC2000, ECH8000, ED8000, EF5000, EL3010, EL5070/8608

I've never heard of a thermionic or cold-cathode tube with a Pro-Electron designation. Can't describe tube functionality with Pro-Electron letters, too, so I wonder what the tubes part in the Pro Electron article is all about. "1^st letters A, B, C, F, T, U were out of fashion in M-P and reallocated; 2^nd letters A, C, D, F, L, Y are descendants from the M-P system" could establish historic precedence as well.

Or is Pro-Electron just the authority that issues MP designations? Then I think infos about them belong here. -Mkratz (talk) 23:08, 20 April 2013 (UTC)

• See pages 89 - 90 of Babani's handbook of Radio, TV, Industrial & Transmitting Tube & Valve Equivalents as one source of the 4-digit sequence number classification system. Although many of the Mullard and Philips manuals I have looked at don't specifically name the naming system, at least the May 1969 Philips "Electron Tubes" Part 5 does say: "NEW CODE SYSTEM (PRO-ELECTRON TYPE DESIGNATION CODE)" for the naming system applied to thermionic valves in that book, so it isn't just for semiconductors (which makes sense too because the Pro-electron system basically took over what the Mullard-Philips system had been up until then, and carried on in the same direction). I'm not sure what you mean by "Can't describe tube functionality with Pro-Electron letters..." - the letters (like B for silicon) used by Pro-Electron for semiconductors is just the subset of the naming schem for those devices... it is a bigger system, embracing picture tubes, logic ICs, etc, etc. 202.78.155.231 (talk) 04:42, 8 May 2014 (UTC)

Difference between "Screening" and "Screen"

In the paragraph "Nomenclature systems" we have an example to explain the difference between SQ tubes and non-SQ tubes: "for example the E80F was more suitable for audio and electrometer applications, lacking the RF screening of the EF80, with anode and screen grid power ratings roughly half the EF80". An IP repeatedly asks for citation, because he does see a "screen" in the E80F as well ("Reference shows the valve with a screen despite claim that it has not."). I think this is a simple misunderstanding. "Screening" means a special testing (in this case suitability for RF-use), while the "screen" is one of the grids. Of course, both tubes have this "screen", but only the "commercial" version underwent a "screening" for fulfilling special RF requirements. This was not needed for the LF applications of the E80F. Or am I (a non-native English speaker) wrong with this interpretation? --Wosch21149 (talk) 14:12, 23 October 2014 (UTC)

It refers to an R.F screen around the electrode structure. The EF80 has such a screen as can be seen in the valve illustrated at [1]. The corresponding E80F lacks the screen as can be seen at [2]. However, confusion may arise because data sheets show pin 2 of the E80F as being a screen connection. The screen is question is a small screen that shields some of the wires leading from the base, these can just about be seen in the photo of the E80F. The EF80 lacks this later screen. 86.153.135.110 (talk) 17:33, 2 June 2015 (UTC)