Filament voltage standards for tubes (valves)

Anybody know why there were different filament voltages for rectifiers (5 volts) and various small-signal tubes (6.3 or 12.6 volts).

Why didn't one size fit all?

Roby

Reply to
Roby
Loading thread data ...

There are a lot more voltages than those for filaments and heaters for vacuum tubes. Some of the original tubes were designed to run from batteries, and their voltages started at 1.5V, the so-called A battery.

6.3 V tube heaters ran from early automobile batteries with 3 cells in series. The 12.6 V tubes were for the present generation automobiles with 6 cell, (12V), batteries. The 5V for rectifier filaments was just a convenient voltage for a winding on the power transformer.

There are a few that ran at 2.5V, 2.8V, and a bunch of rectifiers at 5V. There were even 18.9, 25, 35, 45, 50, 70, and 117V heaters. The most famous of these were the ones used in the AD/DC five tube receivers in which the heaters were all in series across the 117V line.

When I worked at Bell Labs in 1950, one of my jobs was pumping the tubes that were used in TAT3, one of the early telephone trans-Atlantic cables. It had vacuum tube repeaters about every forty miles. The heaters of all of these were connected in series! It required a few kilovolts DC at each end. The amplifiers were powered by the voltage drop across the heaters. These heaters ran on about 24V DC for each tube.

Reply to
VWWall

That was back in the West Street era. That must have been a fun place to work back then. Just a few minutes by subway to the Theatre District.

Reply to
John Gilmer

Indeed! Just a short walk to Greenwich Village and some of the well known clubs that had lunch specials. The ferry ride from Hoboken was fun also. I moved to Murrey Hill in 1950 and then to California, (North American Aviation), in 1951. I miss NYC, but wouldn't want to live there. :-)

Reply to
VWWall

Popular in Europe (don't know about US, 120V would be a problem) were the 300mA heater tubes/valves, which were all designed to run in series (part numbers all starting 'P'). TV sets with perhaps

15 tubes had them all in series across the mains, with a power resistor to make up any difference between the mains voltage and the total heater voltage and a thermister to limit the inrush current. Different sized tubes/valves used different voltages, so that the power required was delivered from the 300mA current. There was also another standardised current which I don't recall clearly -- might have been 150mA, but the the 300mA ones were far more common in commercial products.
Reply to
Andrew Gabriel

Was the A battery ever a physical size (like AA, C, D cells) or just a designation for the filament battery?

Wow. Whose job was it to replace a burned-out tube? :-) How many voice channels could one circuit handle and how long did they last? I'd hate to calculate a MTBF for a thousands-mile circuit with a tube amp every

40 miles...
Reply to
Michael Moroney

It was the designation for the filament supply. The "plate voltage" was supplied by the B battery, and the bias voltage was from the C battery.

From:

formatting link
TAT-3 Transatlantic No. 3 Out-of-service: retired 1986, 23 years of Service Tuckerton, New Jersey, U.S.A. -- Widemouth Bay, England - 6,515km at 414

  • 414 KHz Maintenance Authorities: AT & T, British Post Office

formatting link
In 1962, they invented SD analog technology, which enabled bi-directional transmissions rather than requiring a separate cable for each "direction" of traffic. This technology, which also increased bandwidth, was used by Tyco Telecommunications to build TAT-3, the third transatlantic telephone cable, and allowed simultaneous transmission of

148 circuits.

Now, of course, fibre optic cables are used.

I skipped that part of technology, but worked in communication satellites since 1960. They still use vacuum tubes--TWTA's, (Traveling Wave Tube Amplifiers), as the RF output amplifiers.

Reply to
VWWall

WECO, the manufacturing arm of the Bell System, made some equipment that used good design engineering to a degree that was sometimes mind boggling.

How about vacuum tube amplifiers where the tubes were *very* likely to last 20 years! I never worked on a submarine cable, so I can't speak to the specifics of the situation above, but I did see tubes that were more than 25 years old and still functioning!

The basic trick was to to use 24 volt filaments, and then actually run them at seriously reduced filament voltages. That doesn't allow for much gain, but when only 1-6 dB was needed needed, they didn't design a circuit with 30 dB of gain and cut it back. They reduced the filament voltage until the tube would barely warm up, barely have any gain, and *never* wear out!

The A4 Channel Bank used with L carrier starting back in the late 1930's was an example where that was done. I don't remember exactly which tube were used, but they were 300 series triodes, like 310's, 311's, or 312's, if I remember right.

Reply to
Floyd L. Davidson

These particular tubes were asembled at Bell Labs West Street, which is why, as a new junior engineer, got the job of pumping and activating them!

These were based on a design that had been on life-test for over twenty years.

The filament or heater voltage has nothing to do with the gain!!! An amplifier with 1-6 dB gain would be worthless!! Read some basic electronic books.

You've got the basic physics very wrong here. A cathode, or filament emits electrons, and has to run at a temperature where this happens. If the temperature is too low, the cathode surface will not emit the ammount of electrons to support the tube's current. If too hot, the active part of the cathode is depleted rapidly. Another problem for long life is the poisoning of the cathode surface by residual material within the tube.

Reply to
VWWall

in article Ivzde.3453$ snipped-for-privacy@newsread3.news.pas.earthlink.net, VWWall at snipped-for-privacy@DEADearthlink.net wrote on 5/2/05 5:23 PM:

I happened to be visiting Bell Labs Whipany, in 1956 I think, when the cable was dedicated.

Bill

Reply to
Repeating Rifle

in article Ivzde.3453$ snipped-for-privacy@newsread3.news.pas.earthlink.net, VWWall at snipped-for-privacy@DEADearthlink.net wrote on 5/2/05 5:23 PM:

For most of the tubes, the number 6 in a 6SQ7 or 12 in a 12SQ7 designated the nominal heater voltage. Otherwise, tubes with the same designation except for the heater were supposed to be the same. Some tubes like rectifiers and output amplifiers required higher voltages for the heaters to produce the required electron emission.

In particular, many AC/DC sets use a 35L6 or 50L6 beam tetrodes for the output stage. In this cased, however, they were not analogs of the venerable

6L6.

Without further designation, these tuves had metal tube envelopes. For lower cost, smaller glass envelope tubes were used with a GT designation, such as

12SQ7GT.

Bill

Reply to
Repeating Rifle

For the European numbering, it's the first letter... 'D' is 0.5-1.5V, 'E' is 6.3V, 'G' is 5V. For series operation, 'H' is 150mA, P is 300mA, U is 100mA.

Or the older glass ones with just a G designation, which I once heard someone describe as "female form shape" ;-)

Reply to
Andrew Gabriel

I wouldn't imagine there are too many TWT's left, though there could be. Virtually everything I know of (a few hundred earth stations in Alaska) has long since been converted to Solid State Power Amplifiers (SSPA's).

Reply to
Floyd L. Davidson

Giggle snort...

Some circuit don't require *any* gain. And if you don't think the filament voltage would affect gain (technically it is the transconductance that changes, but that amounts to gain in practical circuits), just try running most tubes at say about

1/2 the normal filament voltage and see how much gain they (don't) have!

Exactly. Now... remember all the nice little "cathode activity" tests that were built into many circuits in L carrier systems? Put a voltmeter across a resistor in series with the plate circuit, thus measuring plate current, and then lower the filament voltage a specified amount. The plate current will drop, and if it goes too far down, the tube should be replaced because the "cathode activity" is too low.

And of course if a circuit is designed to operate at the lowest functional cathode surface temperature, the tube will last

*significantly* longer than if it the normal temperature is maintained.

More significant with some tubes than with others though. "Gassy" tubes were common.

One cause of such failure, with certain types of tubes, was applying high voltages to various elements before the filament/cathode was fully up to temperature. Generally that applied only to transmitting tubes, but was also significant with some receiving tubes such as the 416 lighthouse microwave tubes that were used in the TD-2 microwave systems installed in the late 40's and early 50's which changed the nature of long distance in the Bell System.

For such tubes a switch was provided to turn on filaments first and then, after a delay, another switch for the high voltages. Failure to wait long enough would destroy a tube that had more than perhaps 4-6 months of filament time. Some equipment had timers to ensure the delay. Some equipment used variacs to slowly bring up the voltages and allow precisely setting them too. Filament voltage on many transmitting tubes is very critical, as if it is slightly low performance suffers and if it is slightly high the tube's life is shortened significantly.

Reply to
Floyd L. Davidson

LOL is more like it!!

Some circuits don't even require tubes. :-)

Every drug store had a testor like this in the 40-50's. A very few actually measured gm.

The circuit desigh has nothing to to with the "lowest functional cathode surface temperature", whatever that is.

Which is why getters are used and during activation, induction heaters are used to out gas all tube elements.

The TD-2 tubes has such a small cathode-grid spacing that the grid could pick up cathode material and become emissive. I used similar tubes in the design of the ARSR, (Air Route Survellance Radar).

Really? ;-)

Reply to
VWWall

I don't know of any that use higher voltages to get better emission. Commonly they use low voltages and significant amounts of current though.

Rectifiers typically used 5 volts on the filament. As opposed to the typical "receiving" tube such as 6AK5 etc, which actually has a 6.3 volt filament, many transmitting tubes require 5.0 or

6.0 volts.

The 35L6 and 50L6 are identical to a 25L6, which is a "low-power

6L6" with a higher filament voltage. Of course there are a couple dozen or more "standard" variations on a 6L6, so it's hard to say that the 25L6 wasn't really a 6L6 just because it had slightly lower power ratings instead of slightly higher as most did. :-)

The metal envelopes were a fad that occurred in the 1930's, and continued on momentum only into the 40's. By that time nobody really wanted to make them. But the original 6L6, which was the first "beam tetrode" design and spawned hundreds of other similar tubes, was a metal tube. The first glass tubes to replace metal tubes (as opposed to glass tubes that had no metal counterpart) added a G suffix, but then other variations came along too and there are GA, GB, GC, GN, GS, GT, GX, and who knows how many others. Those would all be, for the most part, direct replacement tubes, though it wasn't uncommon to find circuits that did not work well with certain versions, or for that matter with tubes from certain manufacturers.

Suffixes with A also had a significant meaning, and for the life of me I can't remember what it was! Probably the ability to do a "slow warm up" with less of current surge.

WECO of course did not label tubes that way. Their designs were numbers, for example a 408, and then there might be be a 408A,

408B and so on as "improved" or sometimes "selected" variations were produced.

Tubes like the 6AK5 mentioned above (a 7 pin miniature) did not ever have a metal counterpart, and therefore did not have a G suffix. However, sometimes there were special selected variations of such tubes that might have suffixes like W for a ruggedized version. And there were identical tubes with 4 digit numbers too. A 5654 was a 6AK5, for example.

Some tubes were "selected" according the customer too! There were lots tubes that either had a prefix or a suffix of "WA", which were selected for the White Alice Communications System in Alaska. If I remember the number right, the 416B tubes we used were all labeled WA6280 (it might have been 6280WA though), and I seem to remember 5654WA tubes, but it may have been some other type.

Reply to
Floyd L. Davidson

Ain't that the truth. At least you gave up on being serious, for the most part.

Being silly, are you? There are many applications where tubes are used in circuits that do not require gain. And as little a

1-6 dB is of course *very* common in telecom designs.

Yes, many of them did in fact test tubes with something very similar to that.

The point to learn though, is that your original statement above is incorrect, because *you* got the physics wrong.

Look, if you don't have a clue about how any of these designs worked, and never saw one of them, just say so. But cease making abjectly stupid statements.

The purpose of reducing the filament voltage was *precisely* to lower the filament/cathode (whatever the electron emitting element was) to a temperature that would just barely allow it to function. That resulted in an extremely stable long lived vacuum tube circuit, as opposed to one that required the tube to be changed every few years.

Yep.

That and the extremely thin layer of oxide on the cathode. That layer was something like 0.0005 inches. The cathode to grid spacing was only 0.0006 inches.

Yes, 416B's were popular. But none of the other uses had the massive effects that the design and implementation of the TD-2 microwave network that was installed by the Bell System. The difference was that in the 30's and 40's, as we can still see in vintage movies, everyone who made a long distance call to grandma had to shout "Hello Hello, can you hear me!" into the phone three times before grandma realized that someone actually was on the line. By 1955 or so it was possible to almost speak normally on a cross-country telephone call from just about anywhere to anywhere else in the country. That was enabled by the 416B tube.

(And is was Pulse Code Modulation, finally fully implemented

1980's made it possible to whisper into a phone and be heard!)

Matter of fact. Even if you had forgotten about it (or never knew for all I can tell).

Reply to
Floyd L. Davidson

How many systems have *you* designed?

There is an optimum temperature and it's not where the cathode "just barely functions".

Close, but no cigar! See:

Morton,J.A., "A Microwave Triode for Radio Relay". Bell Laboratories Record, May, 1949.

The L-3 Coax system worked pretty well, too.

One of the neat things about the TD-2, was that it was single sideband and put into use with only one sideband used. When traffic warranted it, a simple change doubled the capacity.

The three stage 416A amplifier had a gain of 18 db at a bandwidth of 20 megacycles at the 0.1 db points. The output was set at 27 dbm.

Reply to
VWWall

in article snipped-for-privacy@barrow.com, Floyd L. Davidson at snipped-for-privacy@barrow.com wrote on 5/3/05 5:00 PM:

The "classical" AC/DC set used a string of tubes requiring 0.15A for the heater. The ususal low power tube in the string ran at 12V. The output tube and rectifier tube required more heater power than could be obtained at 12 (actually 12.6 nominal) volts and 0.15A. Thus, the typical 50L6GT output tube used 50V and the typical rectifier 35Z5 tube used 35V. Three 12V tubes and these higher heater voltage tubes gave a string that required about 120V to run. The 35Z5 had a tapped heater that was used to obtain power for a pilot light.

The classical AC/DC used "bantam" octal tubes that were keyed to orient them with respect to the socket. The 6AK% was a later tube that did not use an octal base.

The metal tubes had the advantage of not requiring separate shields.

Reply to
Repeating Rifle

in article snipped-for-privacy@barrow.com, Floyd L. Davidson at snipped-for-privacy@barrow.com wrote on 5/3/05 5:43 PM:

A lot depends upon what you mean by "gain." A cathode follower, the equivalent of a modern emitter follower, had a voltage gain less than one. Nevertheless, it provided a power gain from a high impedance source because it can drive a lower impedance load than the source could.

Bill

Reply to
Repeating Rifle

PolyTech Forum website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.