Tube types
kruesi asked:
Seeing confusion in the ever-proliferating tube types, the (American) Radio Manufacturer's Association (RMA) started registering tube types in 1934. This was a voluntary system: the manufacturer sent the critical specs of a tube type to the RMA which then forwarded them the other RMA members. If no objections were raised then the type was officially "registered" and any manufacturer who made tubes marked with the registered number had to meet the published specs. This might seem dis-advantageous to the manufacturers who would like to keep state of the art tube types to themselves, but the consumer electronics market demanded multiple sources. The US military also required RMA registration since it, too, desired 2nd sources. The RMA later became the EIA (Electronics Industries Association).
Prior to RMA registration, the tube manufacturer would request a tube type number for their proposed design. The RMA would then assign a number using the scheme outlined below. The assignment of a number was not binding - thus many numbers were assigned that never were registered - this explains the many "gaps" in the numbering sequence.
There are three basic RMA numbering schemes: original, early industrial, and later industrial. This can be summarized by the start of each series: 2A3, 1B21, and 5500, respectively. There are also two systems used for CRTs and picture tubes.
The original series started with the 1C6, registered on Feb. 5, 1934. The first digit represented the filament voltage, the last number was the number of "elements" inside the tube, and the letter(s) were assigned sequentially, starting with A, then rolling around to AB and so forth. The general rules and exceptions are:
1. Rectifiers started with Z and went backwards.
2. Doubled-up letters (i.e. AA) were excluded.
3. The letters I and O were excluded.
4. If a different heater voltage of an existing type was to be made, the original letters would be used, unless they had already been assigned. This is the source of most of the exceptions and confusions.
5. Loctal tubes started with 7 or 14 to keep them separated from regular types.
6. The number-of-elements rule can get confusing. The octal metal shell counted as an element, so both a 6C4 and 6C5 are single triodes, the later is a metal octal type. However, the G and GT glass types still had the extra "element" (i.e. 6C5G, 6C5GT).
7. The single-ended octal types were initially indicated by an extra S ahead of the original double-ended number, i.e. 6SK7 vs 6K7. However, single-ended types that had no double-ended predecessor, such as the 6SG7, still used the S notation.
8. The G suffix meant an ST glass envelope version of a metal type and the GT suffix meant a T (tubular) glass version. However, up until the early 1950s, all glass versions were given G or GT suffixes, even if no metal version existed, i.e. 6SN7GT. Leaving a G or GT of the type number is a convenient shorthand but doesn't correctly name the tube.
9. Improvements or enhancements were indicated by adding a letter suffix, starting with A. The later types were generally downward compatible, i.e. a 6L6GC could be used in an amp designed for a 6L6GB, but a 6L6GB should not be used in an amp designed for a 6L6GC.
10. Although not sanctioned by the RMA, the following suffixes are occasionally used: W means ruggedized, often to mil specs, Y means low-loss micanol (brown) base, X means ceramic base.
The last registered receiving tube was the 6MJ6 by RCA on Dec. 4, 1973. The highest registered number was the 6MY8 by Toshiba, registered on April 17, 1971. This shows how when the number was "pulled" versus when it was registered can vary a lot. Picture tubes were registered well into the late 1990s, ending when TV manufacturing in the US ended.
I won't go into detail on the industrial numbering system here. However, it should be noted that the 1N and 2N semiconductor series are part of the early industrial numbering scheme, and all semiconductors with the 1N, 2N, 3N, etc. prefixes were registered with the EIA.
It wasn't necessary to have an RMA-supplied number to be RMA registered. During and shortly after World War 2, many existing transmitting and industrial types were registered en mass, including the RCA 800-series, Eimac tubes, and many Western Electric tubes.
While the RMA numbering system is essentially American, foreign companies registered some of their tubes as far back as the 1930s. Usually this was to assist in selling their tubes into the American market, but sometimes they were for tubes only sold into their own markets, such as the 6AR7GT in Australia, the 6M7G in France, and the 6GV7 in the UK.
I have gotten most of this information from the "TCA Data Cache" - a collection of nearly all the RMA/EIA registrations since about 1947, including semiconductors. Luckily there are indexes sorted by sponsor, type, date, and registration number to help sort through this vast collection of data. It can be purchased from the Tube Collector's Association for $50. See: TCA Notices & Announcements for details.
- John Atwood
kruesi asked:
Seeing confusion in the ever-proliferating tube types, the (American) Radio Manufacturer's Association (RMA) started registering tube types in 1934. This was a voluntary system: the manufacturer sent the critical specs of a tube type to the RMA which then forwarded them the other RMA members. If no objections were raised then the type was officially "registered" and any manufacturer who made tubes marked with the registered number had to meet the published specs. This might seem dis-advantageous to the manufacturers who would like to keep state of the art tube types to themselves, but the consumer electronics market demanded multiple sources. The US military also required RMA registration since it, too, desired 2nd sources. The RMA later became the EIA (Electronics Industries Association).
Prior to RMA registration, the tube manufacturer would request a tube type number for their proposed design. The RMA would then assign a number using the scheme outlined below. The assignment of a number was not binding - thus many numbers were assigned that never were registered - this explains the many "gaps" in the numbering sequence.
There are three basic RMA numbering schemes: original, early industrial, and later industrial. This can be summarized by the start of each series: 2A3, 1B21, and 5500, respectively. There are also two systems used for CRTs and picture tubes.
The original series started with the 1C6, registered on Feb. 5, 1934. The first digit represented the filament voltage, the last number was the number of "elements" inside the tube, and the letter(s) were assigned sequentially, starting with A, then rolling around to AB and so forth. The general rules and exceptions are:
1. Rectifiers started with Z and went backwards.
2. Doubled-up letters (i.e. AA) were excluded.
3. The letters I and O were excluded.
4. If a different heater voltage of an existing type was to be made, the original letters would be used, unless they had already been assigned. This is the source of most of the exceptions and confusions.
5. Loctal tubes started with 7 or 14 to keep them separated from regular types.
6. The number-of-elements rule can get confusing. The octal metal shell counted as an element, so both a 6C4 and 6C5 are single triodes, the later is a metal octal type. However, the G and GT glass types still had the extra "element" (i.e. 6C5G, 6C5GT).
7. The single-ended octal types were initially indicated by an extra S ahead of the original double-ended number, i.e. 6SK7 vs 6K7. However, single-ended types that had no double-ended predecessor, such as the 6SG7, still used the S notation.
8. The G suffix meant an ST glass envelope version of a metal type and the GT suffix meant a T (tubular) glass version. However, up until the early 1950s, all glass versions were given G or GT suffixes, even if no metal version existed, i.e. 6SN7GT. Leaving a G or GT of the type number is a convenient shorthand but doesn't correctly name the tube.
9. Improvements or enhancements were indicated by adding a letter suffix, starting with A. The later types were generally downward compatible, i.e. a 6L6GC could be used in an amp designed for a 6L6GB, but a 6L6GB should not be used in an amp designed for a 6L6GC.
10. Although not sanctioned by the RMA, the following suffixes are occasionally used: W means ruggedized, often to mil specs, Y means low-loss micanol (brown) base, X means ceramic base.
The last registered receiving tube was the 6MJ6 by RCA on Dec. 4, 1973. The highest registered number was the 6MY8 by Toshiba, registered on April 17, 1971. This shows how when the number was "pulled" versus when it was registered can vary a lot. Picture tubes were registered well into the late 1990s, ending when TV manufacturing in the US ended.
I won't go into detail on the industrial numbering system here. However, it should be noted that the 1N and 2N semiconductor series are part of the early industrial numbering scheme, and all semiconductors with the 1N, 2N, 3N, etc. prefixes were registered with the EIA.
It wasn't necessary to have an RMA-supplied number to be RMA registered. During and shortly after World War 2, many existing transmitting and industrial types were registered en mass, including the RCA 800-series, Eimac tubes, and many Western Electric tubes.
While the RMA numbering system is essentially American, foreign companies registered some of their tubes as far back as the 1930s. Usually this was to assist in selling their tubes into the American market, but sometimes they were for tubes only sold into their own markets, such as the 6AR7GT in Australia, the 6M7G in France, and the 6GV7 in the UK.
I have gotten most of this information from the "TCA Data Cache" - a collection of nearly all the RMA/EIA registrations since about 1947, including semiconductors. Luckily there are indexes sorted by sponsor, type, date, and registration number to help sort through this vast collection of data. It can be purchased from the Tube Collector's Association for $50. See: TCA Notices & Announcements for details.
- John Atwood
John, you supply a wealth of information- your comments are truly appreciated.
I do remember the RMA (becoming EIA later) - it's good to refresh the memory now.
Yes, the last "number of elements" number can't be taken TOO seriously... plenty of latitude for poetic license here. The suffixes I get, along with most, I suspect- G, GT specifying envelope shape, and "A", "B" specifying improvements.
I'm surprised to hear the 6MJ6 was the last registered type- I kinda figured the last one must have been a Compactron or one of the ultra miniature military / communication types.
I'm surprised Compactrons weren't more popular. Some of them were really pretty good tubes- but then, some were just plain weird.
Also surprised that Locktals didn't last that long. They were the first effort (AFAIK) by US manufacturers to eliminate the "pinch" and get the elements closer to the base- an idea that just might catch on, if you're interested in RF... and they eliminated the hollow metal pins and the soldering thereto. They were REALLY hard to unseat, but they were at least keyed and easy to insert, a feature that the later (and cheaper!) 7- and 9-pin miniature types lacked.
Great information, John- and thanks for the reference! I'm not a collector, merely interested in the path of technical progress.
I do remember the RMA (becoming EIA later) - it's good to refresh the memory now.
Yes, the last "number of elements" number can't be taken TOO seriously... plenty of latitude for poetic license here. The suffixes I get, along with most, I suspect- G, GT specifying envelope shape, and "A", "B" specifying improvements.
I'm surprised to hear the 6MJ6 was the last registered type- I kinda figured the last one must have been a Compactron or one of the ultra miniature military / communication types.
I'm surprised Compactrons weren't more popular. Some of them were really pretty good tubes- but then, some were just plain weird.
Also surprised that Locktals didn't last that long. They were the first effort (AFAIK) by US manufacturers to eliminate the "pinch" and get the elements closer to the base- an idea that just might catch on, if you're interested in RF... and they eliminated the hollow metal pins and the soldering thereto. They were REALLY hard to unseat, but they were at least keyed and easy to insert, a feature that the later (and cheaper!) 7- and 9-pin miniature types lacked.
Great information, John- and thanks for the reference! I'm not a collector, merely interested in the path of technical progress.
tubes
The registering of industrial tubes trickled-on to the 1980s. The last industrial registration was for the 9021, an Amperex 35kW power tetrode, on Oct. 7, 1986. I'm sure that specialized industrial and transmitting tubes are still being designed, but no one seems to go to the trouble of registering anymore.
The problem with compactrons is that tube and equipment designers are notoriously conservative - if an older type works and is "qualified" by some agency then new types are avoided. This is particularly true with military designs where the types used up through the 1980s seem stuck in a time warp from 1949.
Loctals were a superior tube for its time (the first one was registered in 1938), but got screwed by RCA, since RCA was promoting their metal and later their miniature types. RCA scored a major coup during World War 2 by pushing a "Preferred Types" list, in the interest of efficiency. Notably absent from the list were loctals and many other non-RCA types. Sylvania made a valiant attempt after the war to push loctals (as well as subminiatures), but by then the miniatures were coming on too strong.
Other manufacturers of loctal tubes were National Union, Raytheon, and Landsdale. Landsdale was owned by Philco and made tubes up through the 1960s, but never were sold under their own name. They also were the sponsor of the first loctal registrations in 1938, so may have had a hand in developing them.
Interestingly, the single-ended dual triode was first registered as a loctal, the 7N7, on April 15, 1940, whereas the ubiquitous 6SN7GT was registered almost a year later in March, 1941 by RCA.
One thing I still don't know about is the link-up between Sylvania and Philips regarding loctal technology. Philips came out with the loctal in Europe about the same time as Sylvania and Landsdale did in America, and since Philips was essentially the RCA of Europe, the format was not suppressed. In fact, Philips came out with a 9-pin loctal (such as used on the EF50) and the 8-pin "rim-lock" design seems to draw a lot from loctal technology. I suspect that Sylvania and Philips swapped technology and patents, but I don't know who did the heavy lifting in developing the loctal technology.
- John
The registering of industrial tubes trickled-on to the 1980s. The last industrial registration was for the 9021, an Amperex 35kW power tetrode, on Oct. 7, 1986. I'm sure that specialized industrial and transmitting tubes are still being designed, but no one seems to go to the trouble of registering anymore.
The problem with compactrons is that tube and equipment designers are notoriously conservative - if an older type works and is "qualified" by some agency then new types are avoided. This is particularly true with military designs where the types used up through the 1980s seem stuck in a time warp from 1949.
Loctals were a superior tube for its time (the first one was registered in 1938), but got screwed by RCA, since RCA was promoting their metal and later their miniature types. RCA scored a major coup during World War 2 by pushing a "Preferred Types" list, in the interest of efficiency. Notably absent from the list were loctals and many other non-RCA types. Sylvania made a valiant attempt after the war to push loctals (as well as subminiatures), but by then the miniatures were coming on too strong.
Other manufacturers of loctal tubes were National Union, Raytheon, and Landsdale. Landsdale was owned by Philco and made tubes up through the 1960s, but never were sold under their own name. They also were the sponsor of the first loctal registrations in 1938, so may have had a hand in developing them.
Interestingly, the single-ended dual triode was first registered as a loctal, the 7N7, on April 15, 1940, whereas the ubiquitous 6SN7GT was registered almost a year later in March, 1941 by RCA.
One thing I still don't know about is the link-up between Sylvania and Philips regarding loctal technology. Philips came out with the loctal in Europe about the same time as Sylvania and Landsdale did in America, and since Philips was essentially the RCA of Europe, the format was not suppressed. In fact, Philips came out with a 9-pin loctal (such as used on the EF50) and the 8-pin "rim-lock" design seems to draw a lot from loctal technology. I suspect that Sylvania and Philips swapped technology and patents, but I don't know who did the heavy lifting in developing the loctal technology.
- John
In 1925 the standard 60 Hz 110 volt line was not fully adopted, and there were places with no power, and there might have still been some places that still had DC power. Most went AC by 1900.
Many old radios were built to operate on battery power only. The first product sold by the Galvin Manufacturing Corporation (now Motorola) was a "battery eliminator". It converted "house" power to "A", "B" and "C" power to run the radio. Today we would call it a power supply.
Yeah, they were often used in car radios. They came in cars that were made before all cars had air conditioning. Here in south Florida that means if the loctal is in the socket for more than a year or two, it will become "one with the socket". Yes, they rust!
Just in case someone wants to know. The ceramic RF tetrodes from the 4X150, 4CX250 and 4CX350 family will fit into a Loctal socket.
Yes, it took quite a while for frequencies and voltages to be standardized- What a mess! -It's like trying to decide between Beta and VHS...
I've seen a few loctal tubes that began to rust here in Illinois- I imagine it's REALLY a problem in S Fla for mobile use! -Still, the whole loctal thing was a pretty good idea- I suppose it pointed the way to miniature bases and higher-frequency applications.
John- I'm not sure what you mean by "single-ended dual triode" -single ended as opposed to ---??? -having the cathodes tied together? -Aren't ALL triodes single ended? -I don't mean to criticize your terminology- I just don't understand what you mean.
There are books on various company stories - Hallicrafters, R.L. Drake, Collins, etc. -I'll have to look for one on RCA. Quite a good/bad story there- They did much to further electronics development, and as much to squelch it (Armstrong/FM, various other patent wars, the whole Sylvania thing). I'll have to look for a book- would make an interesting read this winter.
But once again, that's a topic for another thread.
I've seen a few loctal tubes that began to rust here in Illinois- I imagine it's REALLY a problem in S Fla for mobile use! -Still, the whole loctal thing was a pretty good idea- I suppose it pointed the way to miniature bases and higher-frequency applications.
John- I'm not sure what you mean by "single-ended dual triode" -single ended as opposed to ---??? -having the cathodes tied together? -Aren't ALL triodes single ended? -I don't mean to criticize your terminology- I just don't understand what you mean.
There are books on various company stories - Hallicrafters, R.L. Drake, Collins, etc. -I'll have to look for one on RCA. Quite a good/bad story there- They did much to further electronics development, and as much to squelch it (Armstrong/FM, various other patent wars, the whole Sylvania thing). I'll have to look for a book- would make an interesting read this winter.
But once again, that's a topic for another thread.
Don't forget the added complication of 25Hz AC power well into the 1950s in places like NY state where there was a lot of antique hydro power.
My 1932 Atwater Kent was designed for 25Hz - 60Hz operation for this reason. (All 2.5V indirect heated tubes)
My dad has mentioned that there was noticeable 25Hz flicker visible even in the incandescents of the day. (Buffalo NY early 1930s)
I wonder if some early idhts were designed with longer thermal time constants to deal with this issue...
My 1932 Atwater Kent was designed for 25Hz - 60Hz operation for this reason. (All 2.5V indirect heated tubes)
My dad has mentioned that there was noticeable 25Hz flicker visible even in the incandescents of the day. (Buffalo NY early 1930s)
I wonder if some early idhts were designed with longer thermal time constants to deal with this issue...
single-ended dual triodes
kruesi said:
I meant single-ended in the physical sense, rather than the electronic sense. Before the 7N7 and 6SN7GT there was the 6F8G, which had one of the grids come out of the top of the tube. And, technically, the 7N7 wasn't the first single-ended dual triode, since there were a bunch that had their cathodes or filaments tied together, such as the 19, 6A6, and 6N7. The 7N7 and 6SN7GT were the first dual triodes with separate cathodes (making them much more useful) and no grid cap (making them easier to use and more rugged).
By the way, some years ago I wrote an article on the 6SN7GT and gave it to Single Power Audio, where it still can be found on www.6sn7.com. (I did this article for free.)
- John
kruesi said:
I meant single-ended in the physical sense, rather than the electronic sense. Before the 7N7 and 6SN7GT there was the 6F8G, which had one of the grids come out of the top of the tube. And, technically, the 7N7 wasn't the first single-ended dual triode, since there were a bunch that had their cathodes or filaments tied together, such as the 19, 6A6, and 6N7. The 7N7 and 6SN7GT were the first dual triodes with separate cathodes (making them much more useful) and no grid cap (making them easier to use and more rugged).
By the way, some years ago I wrote an article on the 6SN7GT and gave it to Single Power Audio, where it still can be found on www.6sn7.com. (I did this article for free.)
- John
