14-04-2018, 02:36 AM
The use of the FM limiter valve, or strictly speaking the diode formed by the grid and cathode of the limiter valve, as an AM demodulator was mentioned in connection with the Jason JTV and JTV2 FM-TV sound tuners in the thread: http://golbornevintageradio.co.uk/forum/...p?tid=6373.
When I first saw the Jason JTV2 circuit around 20 years ago I was somewhat surprised about its AM demodulation arrangements. I associated it with grid leak detection which was generally thought to be unsuitable for high quality work. On the other hand, I supposed that Jason would not have done it that way if it did not deliver adequate performance. Also, I noted that the limiter grid was often used as a source of AGC bias in FM receivers and tuners, indicating that it provided a reasonably signal level-dependent output. Then I made the connection with the fact that unfiltered AGC was sometimes used in FM tuners to provide additional AM rejection, indicating that the limiter grid provided a reasonable facsimile of the modulation. The specific example here was the Quad FM tuner. The Series B second iteration (SN 8677 and up) had unfiltered AGC from the limiter grid back to the 1st IF stage. (The original Series A, to SN 4000, and the first iteration of Series B had filtered AGC from the limiter grid back to the RF stage. Series B added AFC and neutralization for the 1st and 2nd IF stages.)
More recently, and also as recorded in the Jason JTV thread, I found some information about the BBC VHF AM/FM comparator receiver designed and built by R.N. Fitton (Ambassador) for the second round of BBC VHF AM vs. FM comparative work, using the Wrotham transmitter. It was briefly described, with mention of its AM detector, in Radio-Electronics for 1951 October, page 51ff. A more detailed description was provided in Wireless World (WW) for 1952 January, page 35ff, although the AM detector was not mentioned. That application certainly suggested a good pedigree for the limiter grid AM demodulation technique. These magazine articles are available at: http://www.americanradiohistory.com/.
F.H. Beaumont of R.N. Fitton wrote an IRE paper on the comparator receiver, namely “The Design of a Comparator A.M./F.M. Broadcast Receiver”, which I have since read. This paper is available here: http://digital-library.theiet.org/conten....1953.0017, and here: https://ieeexplore.ieee.org/document/5258951/. It is most interesting, addressing matters such as valve selection, choice of IF, AGC distribution, etc. In respect of the AM demodulator requirements, Fitton said: “Since the receiver has to compare the two systems without introducing bias in either direction, it is obvious that the a.m. detector should be above suspicion.”
For a description of the chosen AM demodulator form, it is easiest to quite from the paper: “The detector valve is actually the first f.m. limiter stage. The grid functions as a diode anode for a.m. All electrode voltages are maintained for either system, and by so doing no variation in operation is experienced on switching f.m. to a.m. or vice versa. It was observed incidentally that the maintenance of the limiter electrode potentials had the desirable effect of moving the departure from linearity at the bottom of the detector curve from 0.4 V to approximately 0.25 V. Detector distortion at high modulation percentages is thus reduced. It is assumed that the maintenance of the screen and anode voltages, in fact, lowered the "diode" conduction resistance, possibly as a result of space charge reduction. The phenomenon was gratefully accepted without further investigation.”
Added to that, the AC-to-DC load ratio for the specific circuit was stated to be 0.998. So there is little doubt that the limiter grid could be configured as a low-distortion AM demodulator, and that Jason had followed an excellent precedent with its JTV and JTV2 circuitry.
Possibly other British tuner makers had followed the apparent Fitton AM demodulator precedent. Lowther, Chapman and Sound Sales had all offered early (1953 or 1954) FM tuners that also provided for VHF AM reception. Whilst I have no information on the circuitry used, the use of limiter grid AM demodulation would have enabled the addition of the VHF AM facility at minimal extra cost in units that most likely were primarily designed for FM reception. There is some evidence that the prototype Quad FM tuner of 1952 also had VHF AM capability. The nature of its IF strip, two amplifiers and a limiter, rather than say an amplifier, amplifier/limiter and a limiter, aligns with this possibility, including the use of the limiter grid as an AM demodulator.
But the limiter grid AM demodulator was also found on quite another vector. An example I noted a while back without much further thought at the time was from the American GE T1000/T1001 AM-FM table stereo radio receiver. This I think was GE’s first FM stereo model, interesting in and of itself given that the FCC had in 1961 selected the Zenith-GE system for FM stereo broadcasting. Zenith and GE had submitted made separate but basically similar proposals, which the FCC then combined into the final system. (Zenith’s corresponding early AM-FM Stereo table receiver was the MJ1035.)
A more recent closer look showed that in this GE receiver, the FM limiter, a 6AU6A, was also used as an AM demodulator, with AM audio taken from the grid. In this case though, the 6AU6A was operated with zero screen and anode potentials when in the AM mode, so basically the cathode and grid were operating as a diode. That prompted a casual search for prior examples. In the GE case, one may go back at least as far as its X-415 table receiver, and immediate post-WWII design that covered both the old (42 to 50 MHz) and new (88 to 108 MHz) FM bands as well as AM and SW. That used a 6SH6 pentode as FM limiter and AM demodulator, with zero screen grid potential in the AM mode. The limiter grid AM demodulator was then used, although not exclusively so, in a wide variety of GE FM-AM equipment into the early 1960s at least. Limiter grid AM demodulation was also found in 1940s and 1950s AM-FM receivers from Motorola, Packard-Bell, RCA and Sylvania, and possibly from others was well. The limiter valve seems to have been operated variously both with and without its usual potentials in the AM mode in these applications.
Anyway, it is apparent that the use of the limiter as AM demodulator was quite a common technique in American AM-FM receivers during the valve era. Whether there were any European or British examples I do not know, but if so, I suspect that they were rare.
Now to the etc. part of the heading, another aspect of the previously-mentioned GE T1000/1001 receiver looked unusual. The FM 1st IF amplifier, also a 6AU6A, also served as an AM self-oscillating mixer. (In this case there was no AM RF amplifier.) One might deduce that a pentode rather than a heptode was used in this position to obtain the best noise performance on FM. The standard American heptode mixer, the 6BE6, was relatively noisy. The 6BA7 was quieter (about the same as an ECH81 heptode section, I think), but this valve was much less common and would still have been noisier than a pentode. Looking back at earlier GE practice indicates the pathway to the 6AU6 self-oscillating mixer. GE originated the 12AT7 double triode in 1947, intended for use as an oscillator-mixer in both FM and (VHF) TV receivers. It was said to be less microphonic than the 6J6, beneficial for both FM and split-sound TV applications. The TV application fell away with the coming of the “high” 40 MHz IF, whereafter pentode mixers were preferred to minimize regeneration on the lower VHF low-band channels. But the 12AT7 remained an important FM valve.
GE used the 12AT7 in FM-AM receivers as an FM oscillator-mixer, typically preceded by a pentode FM RF amplifier. Rather than provide a separate AM front end, it used the oscillator half of the 12AT7 also as AM oscillator, feeding the (usually 6AU6, but sometimes 6BA6) 1st FM IF amplifier which was then used as an AM mixer of the non-self oscillating kind. With late 1950s/early 1960s changes in FM front end design often eliminating the 12AT7, and with it the convenient AM oscillator, apparently the next step was to make the AM mixer self-oscillating. But GE had also used pentode self-oscillating mixers in some of its AM-only receivers, although it used heptodes in others. The pentode case might have been only with receivers that did not have RF stages, whereas heptodes were in receivers with and without. I suppose one might say that GE was in good company, in that pentode rather than heptode mixers were found on some of the higher performance communications receivers of the valve era.
Combined AM-FM front ends were also part of American practice in the 1940s and 1950s, and beyond in some cases. Zenith was a case in point. It migrated from a pentode-plus-heptode combination to a pentode-plus-12AT7 combination fairly soon after the 12AT7 became available. The 12AT7 was used as an oscillator-mixer on both FM and AM. The preceding pentode, often a 6BJ6 was used as a RF amplifier on FM and sometimes also as an RF amplifier on AM. As best I can work out, Zenith offered FM-AM and AM-only (MF-only) receivers both with and without RF amplifiers through to the end of the valve era. Those with RF amplifiers probably supported Zenith’s “long-range AM” advertising statement. Anyway, it was surprising to me to find a triode used as an AM mixer in the post-WWII period. One wonders about the oscillator radiation propensity of the versions without RF amplifiers. With its combined front ends, Zenith moved from what might be called AM-dominant circuitry (e.g. 6BA6 plus 6BE6) that was non-ideal for FM to FM-dominant circuitry (e.g. 6BJ6 plus 12AT7) around which the AM side was accommodated. Later, with changing FM front end designs, Zenith moved to separate front ends. For example, the previously mentioned MJ1035 AM-FM Stereo receiver had a separate AM front end comprising a 6BA6 RF amplifier and 6BE6 self-oscillating mixer.
RCA also used combined FM-AM front ends, using a pentode RF amplifier (sometimes a 6CB6) for both FM and AM, followed by a 6J6 double triode oscillator-mixer. It seems not to have used the 12AT7 for this purpose, perhaps because the 6J6 was one of its own whereas the 12AT7 was from GE. But the 6J6 was superseded in this role by the 6X8 triode-pentode. The latter was designed primarily for VHF TV oscillator mixer service with the 40 MHz IF, but it was also specified for FM and AM mixer-oscillator service. In the FM case the pentode mixer could be used as such, or triode-strapped, as was the case in RCA’s own receivers. In the AM case pentode operation was normal. That arrangement was perhaps the best of both worlds, a triode mixer on FM and a pentode mixer on AM.
Cheers,
Steve
When I first saw the Jason JTV2 circuit around 20 years ago I was somewhat surprised about its AM demodulation arrangements. I associated it with grid leak detection which was generally thought to be unsuitable for high quality work. On the other hand, I supposed that Jason would not have done it that way if it did not deliver adequate performance. Also, I noted that the limiter grid was often used as a source of AGC bias in FM receivers and tuners, indicating that it provided a reasonably signal level-dependent output. Then I made the connection with the fact that unfiltered AGC was sometimes used in FM tuners to provide additional AM rejection, indicating that the limiter grid provided a reasonable facsimile of the modulation. The specific example here was the Quad FM tuner. The Series B second iteration (SN 8677 and up) had unfiltered AGC from the limiter grid back to the 1st IF stage. (The original Series A, to SN 4000, and the first iteration of Series B had filtered AGC from the limiter grid back to the RF stage. Series B added AFC and neutralization for the 1st and 2nd IF stages.)
More recently, and also as recorded in the Jason JTV thread, I found some information about the BBC VHF AM/FM comparator receiver designed and built by R.N. Fitton (Ambassador) for the second round of BBC VHF AM vs. FM comparative work, using the Wrotham transmitter. It was briefly described, with mention of its AM detector, in Radio-Electronics for 1951 October, page 51ff. A more detailed description was provided in Wireless World (WW) for 1952 January, page 35ff, although the AM detector was not mentioned. That application certainly suggested a good pedigree for the limiter grid AM demodulation technique. These magazine articles are available at: http://www.americanradiohistory.com/.
F.H. Beaumont of R.N. Fitton wrote an IRE paper on the comparator receiver, namely “The Design of a Comparator A.M./F.M. Broadcast Receiver”, which I have since read. This paper is available here: http://digital-library.theiet.org/conten....1953.0017, and here: https://ieeexplore.ieee.org/document/5258951/. It is most interesting, addressing matters such as valve selection, choice of IF, AGC distribution, etc. In respect of the AM demodulator requirements, Fitton said: “Since the receiver has to compare the two systems without introducing bias in either direction, it is obvious that the a.m. detector should be above suspicion.”
For a description of the chosen AM demodulator form, it is easiest to quite from the paper: “The detector valve is actually the first f.m. limiter stage. The grid functions as a diode anode for a.m. All electrode voltages are maintained for either system, and by so doing no variation in operation is experienced on switching f.m. to a.m. or vice versa. It was observed incidentally that the maintenance of the limiter electrode potentials had the desirable effect of moving the departure from linearity at the bottom of the detector curve from 0.4 V to approximately 0.25 V. Detector distortion at high modulation percentages is thus reduced. It is assumed that the maintenance of the screen and anode voltages, in fact, lowered the "diode" conduction resistance, possibly as a result of space charge reduction. The phenomenon was gratefully accepted without further investigation.”
Added to that, the AC-to-DC load ratio for the specific circuit was stated to be 0.998. So there is little doubt that the limiter grid could be configured as a low-distortion AM demodulator, and that Jason had followed an excellent precedent with its JTV and JTV2 circuitry.
Possibly other British tuner makers had followed the apparent Fitton AM demodulator precedent. Lowther, Chapman and Sound Sales had all offered early (1953 or 1954) FM tuners that also provided for VHF AM reception. Whilst I have no information on the circuitry used, the use of limiter grid AM demodulation would have enabled the addition of the VHF AM facility at minimal extra cost in units that most likely were primarily designed for FM reception. There is some evidence that the prototype Quad FM tuner of 1952 also had VHF AM capability. The nature of its IF strip, two amplifiers and a limiter, rather than say an amplifier, amplifier/limiter and a limiter, aligns with this possibility, including the use of the limiter grid as an AM demodulator.
But the limiter grid AM demodulator was also found on quite another vector. An example I noted a while back without much further thought at the time was from the American GE T1000/T1001 AM-FM table stereo radio receiver. This I think was GE’s first FM stereo model, interesting in and of itself given that the FCC had in 1961 selected the Zenith-GE system for FM stereo broadcasting. Zenith and GE had submitted made separate but basically similar proposals, which the FCC then combined into the final system. (Zenith’s corresponding early AM-FM Stereo table receiver was the MJ1035.)
A more recent closer look showed that in this GE receiver, the FM limiter, a 6AU6A, was also used as an AM demodulator, with AM audio taken from the grid. In this case though, the 6AU6A was operated with zero screen and anode potentials when in the AM mode, so basically the cathode and grid were operating as a diode. That prompted a casual search for prior examples. In the GE case, one may go back at least as far as its X-415 table receiver, and immediate post-WWII design that covered both the old (42 to 50 MHz) and new (88 to 108 MHz) FM bands as well as AM and SW. That used a 6SH6 pentode as FM limiter and AM demodulator, with zero screen grid potential in the AM mode. The limiter grid AM demodulator was then used, although not exclusively so, in a wide variety of GE FM-AM equipment into the early 1960s at least. Limiter grid AM demodulation was also found in 1940s and 1950s AM-FM receivers from Motorola, Packard-Bell, RCA and Sylvania, and possibly from others was well. The limiter valve seems to have been operated variously both with and without its usual potentials in the AM mode in these applications.
Anyway, it is apparent that the use of the limiter as AM demodulator was quite a common technique in American AM-FM receivers during the valve era. Whether there were any European or British examples I do not know, but if so, I suspect that they were rare.
Now to the etc. part of the heading, another aspect of the previously-mentioned GE T1000/1001 receiver looked unusual. The FM 1st IF amplifier, also a 6AU6A, also served as an AM self-oscillating mixer. (In this case there was no AM RF amplifier.) One might deduce that a pentode rather than a heptode was used in this position to obtain the best noise performance on FM. The standard American heptode mixer, the 6BE6, was relatively noisy. The 6BA7 was quieter (about the same as an ECH81 heptode section, I think), but this valve was much less common and would still have been noisier than a pentode. Looking back at earlier GE practice indicates the pathway to the 6AU6 self-oscillating mixer. GE originated the 12AT7 double triode in 1947, intended for use as an oscillator-mixer in both FM and (VHF) TV receivers. It was said to be less microphonic than the 6J6, beneficial for both FM and split-sound TV applications. The TV application fell away with the coming of the “high” 40 MHz IF, whereafter pentode mixers were preferred to minimize regeneration on the lower VHF low-band channels. But the 12AT7 remained an important FM valve.
GE used the 12AT7 in FM-AM receivers as an FM oscillator-mixer, typically preceded by a pentode FM RF amplifier. Rather than provide a separate AM front end, it used the oscillator half of the 12AT7 also as AM oscillator, feeding the (usually 6AU6, but sometimes 6BA6) 1st FM IF amplifier which was then used as an AM mixer of the non-self oscillating kind. With late 1950s/early 1960s changes in FM front end design often eliminating the 12AT7, and with it the convenient AM oscillator, apparently the next step was to make the AM mixer self-oscillating. But GE had also used pentode self-oscillating mixers in some of its AM-only receivers, although it used heptodes in others. The pentode case might have been only with receivers that did not have RF stages, whereas heptodes were in receivers with and without. I suppose one might say that GE was in good company, in that pentode rather than heptode mixers were found on some of the higher performance communications receivers of the valve era.
Combined AM-FM front ends were also part of American practice in the 1940s and 1950s, and beyond in some cases. Zenith was a case in point. It migrated from a pentode-plus-heptode combination to a pentode-plus-12AT7 combination fairly soon after the 12AT7 became available. The 12AT7 was used as an oscillator-mixer on both FM and AM. The preceding pentode, often a 6BJ6 was used as a RF amplifier on FM and sometimes also as an RF amplifier on AM. As best I can work out, Zenith offered FM-AM and AM-only (MF-only) receivers both with and without RF amplifiers through to the end of the valve era. Those with RF amplifiers probably supported Zenith’s “long-range AM” advertising statement. Anyway, it was surprising to me to find a triode used as an AM mixer in the post-WWII period. One wonders about the oscillator radiation propensity of the versions without RF amplifiers. With its combined front ends, Zenith moved from what might be called AM-dominant circuitry (e.g. 6BA6 plus 6BE6) that was non-ideal for FM to FM-dominant circuitry (e.g. 6BJ6 plus 12AT7) around which the AM side was accommodated. Later, with changing FM front end designs, Zenith moved to separate front ends. For example, the previously mentioned MJ1035 AM-FM Stereo receiver had a separate AM front end comprising a 6BA6 RF amplifier and 6BE6 self-oscillating mixer.
RCA also used combined FM-AM front ends, using a pentode RF amplifier (sometimes a 6CB6) for both FM and AM, followed by a 6J6 double triode oscillator-mixer. It seems not to have used the 12AT7 for this purpose, perhaps because the 6J6 was one of its own whereas the 12AT7 was from GE. But the 6J6 was superseded in this role by the 6X8 triode-pentode. The latter was designed primarily for VHF TV oscillator mixer service with the 40 MHz IF, but it was also specified for FM and AM mixer-oscillator service. In the FM case the pentode mixer could be used as such, or triode-strapped, as was the case in RCA’s own receivers. In the AM case pentode operation was normal. That arrangement was perhaps the best of both worlds, a triode mixer on FM and a pentode mixer on AM.
Cheers,
Steve
