Featured
Latest
Radio
165
Posts
10
Users
39
Reactions
70.7 K
Views
Topic starter
Another RCA receiver of the later 1950s was the AR-8516 marine main receiver. It was not primarily an SSB receiver, although it did include a basic facility for SSB reception. At the time, international marine communications were mostly AM and CW. SSB/ISB was used for links into telephone systems, but typically separate point-to-point type receivers were used for that.
Marine receivers were a complex proposition. With continuous frequency coverage from LF (or in some cases, VLF) through HF, placement of the IF was an issue. Howsoever done, at least two IFs were needed. Added to that were other factors that indirectly affected IF choices, such as the need for tight front end selectivity that tracked with the main tuning, and ease and accuracy of tuning of the HF marine bands, which indicated the need for some form of bandspread, and avoidance of VFOs operating much beyond the low end of the HF band.
In this case, RCA opted for a relatively straightforward superhet format for frequencies below 1.3 MHz. Above 4 MHz, it chose a converter type arrangement in 2 MHz blocks, using a crystal controlled first oscillator. This conferred both stability and ease of tuning. To avoid conflicts, this required the use of two 2 MHz IF blocks, nominally 1 to 3 and 2 to 4 MHz. These also served as the RF bands for the range 1.2 to 4 MHz.
This chart summarizes the tuning ranges:
Here is the block schematic:
And here is the oscillator and IF table that relates to the above schematic:
It is convenient to start the analysis in the middle, as it were. The main signal processing IF was 455 kHz, a familiar standard number. It was the first IF for the tuning range 520 to 1300 kHz, quite logical considering that the range at interest was nested within the MF broadcast band, and was otherwise suitable for most final processing operations. As with broadcast receiver practice, supradyne conversion was used.
Clearly, 455 kHz could not be used as the first IF for the next lower tuning range, 200 to 520 kHz. Instead, 45 kHz was chosen, and logically also used for the lowest tuning range, 80 to 200 kHz, in both cases with supradyne conversion. Narrow IF bandwidths (1.3 and 0.8 kHz) were easily provided at 45 kHz. Thus, where such were required for frequency ranges above 520 kHz, rather than incorporate them in the 455 kHz IF chain, there was an additional conversion from 455 to 45 kHz, so that the latter facility could be used. This conversion used a 500 kHz crystal oscillator. It could have been that the 45 kHz number was derived simply from a preference to use such an oscillator, and found not to be otherwise disadvantageous.
Turning to the “converter” ranges, 2 to 4 MHz was evidently the main block to which all 2 MHz HF blocks above 6 MHz were converted. The block conversion, supradyne, was done by integer MHz crystal oscillators variously using fundamentals or low harmonics. The second conversion to 455 kHz, infradyne, was done using a VFO whose range was 1.545 to 3. 545 MHz.
The 2 to 4 MHz range was converted directly to 455 kHz using the same VFO.
Conversion of the 4 to 6 MHz range to the 2 to 4 MHz range was undesirable because the ranges nominally abutted, and in fact, slightly overlapped given the range overlaps built in. An 8 MHz oscillator frequency would convert 6 MHz to 2 MHz without problem, but it would convert 4 MHz to 4 MHz, which would be somewhat awkward, at least with single-ended valve mixers. Instead, the 4 to 6 MHz band was converted to a range of 1.09 to 3.09 MHz, using a 7.09 MHz crystal oscillator. The apparently odd 1.09 to 3.09 MHz range allowed the use of the same VFO for supradyne second conversion to 455 kHz as used for the other HF bands.
The 1.09 to 3.09 MHz IF band was also converted directly to 455 kHz, thus infilling the 1.3 to 2 MHz gap.
Thus, the receiver as a whole had four IFs, namely:
2 to 4 MHz
1.09 to 3.09 MHz
455 kHz
45 kHz
Depending upon the incoming frequency and the final IF processing required, the IF combinations were:
2 to 4 MHz and 455 kHz
2 to 4 MHz, 455 kHz, and 45 kHz
1.09 to 3.09 MHz and 455 kHz
1.09 to 3.09 MHz, 455 kHz and 45 kHz
455 kHz
455 kHz, and 45 kHz
45 kHz
The maximum tuning range for any band was 2 MHz, evidently providing for fine enough scales for frequency setting and reading.
The maximum VFO frequency was 3.545 MHz, evidently low enough to confer adequate stability. Also, that VFO had a moderate range of 2.3 to 1, which probably helped.
As mentioned, provision was made for receiving SSB as well as AM and CW signals. This was done using a 3 kHz bandwidth 455 kHz mechanical filter and the same BFO as used for CW, albeit with increased injection amplitude to minimize SSB distortion. So, RCA had deployed its mechanical filter technology here, as well. RCA had said “The limits of the frequency range, at the present state of the art, are from approximately 50 to 600 kc. From a production standpoint, however, the preferred frequencies are from 200 to 250 kc.” In that context, 455 kHz was within the range but away from the optimum frequency. However, for the application at interest, where a variable BFO was used, and SSB reception was anyway secondary, the high stability required for SSB generation purposes was not needed, so the 455 kHz filter was acceptable.
As well as the 455±2 kHz BFO, for use with the 455 kHz IF demodulator, there was also a 44 kHz BFO for use with the 45 kHz IF demodulator, but only for the two lowest frequency bands that were converted directly to 45 kHz. For the higher bands where 455 kHz was translated to 45 kHz, both the 455±2 kHz BFO signal and the 500 kHz conversion oscillator signal were fed into the 45 kHz IF system final stage after the selectivity circuits and so appeared at the 45 kHz demodulator, where inter alia, a 45±2 kHz BFO signal was created. By this means, for signals that used the 455 kHz IF, the same BFO control was used regardless of whether the 45 kHz IF was also used. The pertinent section of the block schematic is shown here:
Overall, the AR-8516 was necessarily quite complex in respect of its IF system. (It was complex mechanically, as well.) It did though use the standard 455 kHz as its main IF. Converter type receivers anyway tend to have individual block IF selections, this case being typical. The 45 kHz IF could be thought of as being a variant in the 50 Hz group.
Cheers,
Steve
Posted : 26/03/2025 6:39 am
Topic starter
An RCA SSB unit of the later 1950s that did not use the 250 kHz final IF was the AN-NRC-54, a military aviation HF transceiver. Rather, it had a 300 kHz final IF, using a mechanical sideband filter. This though was not an independent choice. Rather, it was carried over from the existing AN-ARC-21 AM military aviation transceiver, the objective being that the latter could be modified to the SSB form. Evidently that was done on quite a large scale, alongside new building.
Correction:
The designation for the aviation SSB transceiver was incorrect. It was AN/ARC-65, not AN/ARC-54.
Cheers,
Steve
Posted : 27/03/2025 9:45 pm
Topic starter
Another RCA SSB/ISB receiver of the late 1950s was a model (designation unknown) developed for marine HF applications. It was first deployed on two American vessels, the “Brasil” and the “Argentina”. These were the first applications of HF SSB/ISB to American flag vessels. The first UK flag vessel application had been the “Caronia” in 1949.
These vessels had a pair of RCA AR-8516 main receivers in their main radio consoles as part of their “compulsory” equipment. The SSB/ISB transmitters and receivers, along with a separate radiotelephone console, were “voluntary” equipment.
The RCA marine SSB/ISB receiver had a three-step IF structure that was somewhat different to any of those previously discussed here. In part that arose because the receiver was divided into two parts. One was the remotely controlled head end, mounted aft and adjacent to the receiving aerials, the latter placed as far from the transmitting aerials as reasonably possible. The other part was in the radiotelephone console. The latter also included another AR-8516, whose SSB facility was probably used in this case.
The head end of the SSB/ISB receiver contained the RF, first IF and second IF sections. There was a separate RF amplifier and first frequency changer for each of the five marine HF bands, 4, 8, 13, 17 and 23 MHz. For each of the five bands, there was a crystal-controlled oscillator with ten selectable crystal frequencies, making for a total of 50 remotely selectable frequencies.
The first IF was 3 MHz. In isolation, this was reasonable considering the 4 MHz lowest RF input. But it does contrast with the 1 MHz of the SSB-R3. In that case the lowest RF input was 2.8 MHz, which would have allowed say a 2 MHz IF. But the 1 MHz number may have been the best fit for first oscillator synthesis system, which constraint, if it was such, did not apply with direct crystal control. The marine receiver had a single RF amplifier stage in each of the five RF units, with bandpass tuning on the input as well as the interstage, giving 90 dB of image rejection. The bandpass input traded signal-to-noise ratio in favour of selectivity. Excellent front end selectivity was a key requirement for marine receivers, which were operated in closer proximity to transmitters than most other types. (The SSB-R3 had three RF amplifier stages ahead of the first mixer, each with single-stage tuning at its input and ahead of the mixer.)
The five RF sections were switched to the 3 MHz IF section, which included infradyne conversion to 455 kHz using a 2545 kHz oscillator. The followed a 455 kHz crystal roofing filter of 9 kHz bandwidth (-6 dB points), a 455 kHz IF amplifier and a cathode follower that provided a 200 mV, 75 R output for the coaxial cable that connected the head end to the console. (There was also a local monitoring demodulator and AF amplifier, and an AGC rectifier.)
455 kHz was a logical choice here. It was, nominally at least, a clear channel, so was not subject to direct interference. It was low enough to minimize losses in the coaxial cable link. And it was a convenient frequency for the local monitoring receiver.
At the console end, there was a further frequency conversion, in this case to 100 kHz. This third change conversion was needed in order to provide a frequency change oscillator upon which the AFC system could operate. Why 100 kHz was chosen over say 250 kHz is not clear. Certainly, RCA had developed a full 250 kHz final ISB processing system for the military HF receiver described upthread, in that case with conversion from 600 kHz. But it also had the 100 kHz system as used in the SSB-R3. In the marine case, crystal 100 kHz filters were used for the two sidebands and for carrier extraction.
Under remote control, sideband-derived AGC was also provided from the console unit and fed back to the headend. (Possibly carrier-derived AGC was also available.) The head-end local AGC would have been switched out when remote control was used.
The use of crystal filters at 455 and 100 kHz was somewhat surprising, given that RCA was invested in mechanical filter technology. Possibly it was a bandwidth issue, against a possible constraint that existing filters were preferably used for what was likely a low production number receiver. Existing 455 kHz mechanical filters were likely of lesser bandwidth than the required 9 kHz. And existing 100 kHz mechanical filters may have been limited to the 6 kHz bandwidth types that were used for the SSB-R3. It would appear that where narrower sideband width was required in that receiver, it was achieved by using a suitable roofing filter, say 6 kHz wide, in the 100 kHz IF channel ahead of the sideband separation point, whilst retaining the 6 kHz sideband filters,
That RCA sometimes based designs upon the use of established components and assemblies is illustrated by the SSB/ISB transmitter used on the two vessels mentioned above. This, thought to have been the ET-8063A model, generated the sidebands at 250 kHz, using mechanical filters of 3 kHz bandwidth . The exciter from this transmitter was used for the later and larger IST-5K land-based SSB/ISB transmitter, but the same ISB generator could not be used as 6 kHz sidebands were required, for which 250 kHz mechanical filters were not available. Thus, generation was done at 100 kHz using available 6 kHz crystal filters, with subsequent conversion to 250 kHz for use by the exciter.
Anyway, we may record the use of the 3 MHz/455 kHz/100 kHz combination for an HF SSB/ISB receiver.
Given that ISB/SSB sideband filter bandwidth issues have come up in this discussion and appear to have influenced IF choices in some case, I thought that I would take a quick look at this as a sidebar item.
International telephone voice bandwidth, as established by the CCIF, was 300 to 3400 Hz. I understand that the rationale for the 3400 Hz upper limit was that intelligibility in the presence of noise did not improve for any further reduction in the upper frequency. Nonethless, the need for bandwidth economy in radiotelephone systems often tended to dictate a lower upper limit. This was recognized by the CCIR and recorded in the output from its 1951 Geneva Plenary meeting, thus:
__________
2. Circuits below 30 Mc/s.
2.1 That, since it becomes necessary to economise in the use of the frequency spectrum when considering intercontinental circuits which consist mainly of single long-distance radio links operating on frequencies less than 30 Mc/s, it is desirable to use single-sideband transmission to the maximum extent possible, to employ a transmitted band less than the 300 to 3400 c/s recommended by the C.C.I.F. for land-line circuits and preferably to reduce the upper frequency to 3000 c/s or below but to not less than 2600 c/s, except in special circumstances;
__________
In respect of marine SSB/ISB, although this was voluntary at the time, its use was recognized by the ITU in its 1959 Radio Regulations (from its Geneva meeting). In fact, it was recommended for use in the marine HF bands, using USB and exceptionally ISB, with two SSB channels accommodated in each of the existing AM channels, and a voice frequency range of 350 to 2700 Hz. It was also recommended that coast stations be equipped for both SSB and AM. That voice bandwidth remained in place when SSB later became mandatory for marine MF and HF communications. Interestingly, when SSB became mandatory for aviation HF communications, somewhat before the same happened to marine HF, the voice band upper frequency limit was set as 3000 Hz. (What the lower limit was I do not know.)
Point-to-point SSB/ISB generally used either ±3.5 kHz of ±6 kHz sidebands. The former accommodate a full international telephone channel, 300 to 3400 Hz, in each sideband. The latter accommodated two 3 kHz voice channels in each sideband. One could occupy the space 250 to 3000 Hz. The other, limited to the same range, was modulated on to a typically 6250 Hz carrier (although other frequencies were used), from which the lower sideband, sans carrier, was selected, this occupying the range 3250 to 6000 Hz, and so forming the outer part of a 6 kHz transmitter sideband. At the receiving end, after signal demodulation, it was filtered out and demodulated as the lower sideband of a 6250 Hz signal. There was, so as to speak, SSB within SSB, with 6250 Hz as a final processing frequency, although not strictly speaking an IF, because it was predetermined, not chosen. Perhaps it could be viewed as a quasi-IF, in rather the same way as TV intercarrier sound frequencies. The full 6kHz sideband channel could also be used for broadcast relays, in which case the lower cutoff frequency was sometimes extended downwards.
Anyway, one may see why nominally 3 kHz, 3.5 kHz and 6 kHz sideband filters were the common numbers associated with SSB/ISB receivers (and transmitters).
Cheers,
Steve
Posted : 30/03/2025 2:15 am
Topic starter
The group of RCA receivers recently discussed, all from the later 1950s, and either primarily designed for SSB or having SSB capability, show quite a diversity of IFs and IF systems.
All tuned part or all of the band 2 to 30 MHz, and one also tuned down to 80 kHz. So, they were essentially HF receivers.
Variously:
Initial IFs used were:
3 MHz
2.4 MHz
2 to 4 MHz
1.8 MHz
1.4 MHz
1.09 to 3.09 MHz
1.0 MHz
600 kHz
455 kHz
Intermediate IFs used were:
600 kHz
455 kHz
Final IFs used were:
1.4 MHz
455 kHz
300 kHz
250 kHz
100 kHz
In that list there were some existing and future standard and quasi-standard numbers, namely 1.4 MHz, and 455, 250 and 100 kHz. Of these.455 and 100 kHz were well-established, whereas 1.4 MHz and 250 kHz were apparently new.
The IF diversity from one maker probably, and over a relatively short period of time, probably reflects a number of factors, such as different design teams optimizing designs for performance and cost specific targets, perhaps without a lot of intergroup exchange, carryover of established practice in the case of derivative designs, and in some cases, the use of existing components such as filters. It does reinforce the notion that for HF receivers, unlike say broadcast receivers, there was not a single best answer to the question of IF structure.
With SSB transceivers, intermediate frequencies were also used on the transmit side, which was not usually the case for AM transceivers. To the extent that common IFs were used for the transmit and receive functions, then choices were often influenced more by the transmit requirements. But as has been noted upthread, independent SSB/ISB receivers sometimes used IFs that were originally derived for transmitter purposes, 3.1 MHz being an example.
Cheers,
Steve
Posted : 09/04/2025 12:36 am
Topic starter
Quite by happenstance, I recently found came across some information about the IF structures of four early 1960s, valve-era American Citizen’s Band (CB) equipment. Accordingly, I have compiled the following. I do not know whether, in IF terms, these equipments were in any way typical or representative of the CB equipment of that era.
Polytonics Poly-Comm Senior 23:
This included a dual-conversion receiver with IFs of 6.0 MHz and 455 kHz.
The first conversion was supradyne, with the local oscillator signal generated by two-stage crystal synthesis. The second conversion was infradyne, using a 5.545 MHz crystal oscillator.
The receiver first local oscillator signal was also used on the transmitter side, being mixed down with the output of a 6 MHz crystal oscillator to achieve the correct frequency.
In respect of the synthesizer, the first stage used six crystals over the range 37.60 to 37.85 MHz at 50 kHz intervals. The second stage used four crystals ranging from 4.595 to 4.635 MHz at 10 kHz intervals. The difference between the first and second stages was used as the receiver first oscillator signal, which thus ranged from 32.965 to 33.255 MHz in 10 kHz steps.
I imagine that consideration of both placement of the crystal oscillator frequencies within the spectrum and the likely spurs led to the choice of the first IF as 6 MHz.
An interesting point is the first stage crystal range, 37.60 to 37.85 MHz, was proximate to that which used by itself as the receiver first oscillator signal, would have provided a first IF of 10.7 MHz. (The actual oscillator range required for a 10.7 MHz first IF was 37.665 to 37.955 MHz.) Possibly a 10.7 MHz first IF was the conceptual starting point, with, as it were, the 10.7 MHz range subdivided into a workable decrement allocated to the subtraction of the second crystal oscillator frequency, and a lower, but still acceptable first IF. A possibility is that there was a preceding, non-synthesized model that had a single set of crystals similar to those used for the first stage here, but did not have the second set, and so had 10.7 MHz first IF. This was then modified to suit the desired synthesis format for the later model. But that is speculation on my part.
Hammarlund CB-23:
This also had a dual-conversion receiver. Here though the first IF was variable according to the selected CB channel, being one of three values, namely 1.65, 1.75 and 1.85 MHz. The second IF was 262 kHz, infradyne converted from the first IF with oscillator frequencies of 1.388, 1.488 and 1.588 MHz respectively.
The use of three first IF values suited the two-stage transmitter synthesis process. Here, the first stage was an eight -crystal oscillator producing 25.315 to 25.405 MHz. The second stage was a three-crystal oscillator producing 1.65, 1.75 and 1.85 MHz. This was added to the first oscillator output to generate the transmission frequency. The first stage output alone was used as the receiver first oscillator frequency.
Here one could say that the first IF was a variation on the 1.65 MHz (or thereabouts) theme that was sometimes used for HF receivers, in this case modified to suit the chosen synthesis process. With three different, albeit quite close first IFs, a second conversion to a single second IF was desirable, although in any event, it was probably also desirable for selectivity reasons. With possible mobile use of the transceiver in mind, possibly 262 kHz was selected for the same reasons that it was still used in American car radio receivers of the era.
Regency Range Gain:
This had a dual-conversion receiver with IFs of 7.5 MHz and 260 kHz. It had three-stage additive synthesis of the transmitter frequency. The first crystal stage delivered 8.615 to 8.655 MHz, the second stage added 10.85 to 11.1 MHz, and the third stage added 7.5 MHz. The second stage output was used as the receiver first oscillator frequency. Again, one would suppose that the pattern was chosen in order to minimize deleterious spurs and to end up with an acceptable first IF.
Mark Products Sidewinder SSB-27:
As the model name suggests, this was an SSB unit. The receiver was of the single-conversion type, with a 7.5 MHz IF. Here is the block schematic:
As may be seen, adequate selectivity was obtained at 7.5 MHz using a 3 kHz bandwidth sideband filter, thus obviating the need for a second conversion. This was actually an upper sideband filter. Sideband inversion was obtained by choosing either infradyne conversion, for USB, or supradyne conversion, for LSB. Nominal oscillator frequencies were respectively 19.5 and 34.5 MHz, these being separated by 15 MHz, or twice the 7.5 MHz IF.
The channel oscillator was 19.5 MHz nominal, with variation to cover the whole 27 MHz CB band. This was used for infradyne conversion. For supradyne conversion, 15 MHz was added, this being obtained by doubling the 7.5 MHz carrier insertion oscillator output.
The transmitter process was essentially the reverse, with generation of USB at 7.5 MHz, followed by upconversion to 27 MHz nominal using 19.5 MHz for USB, and 34.5 MHz for LSB.
In this case, the IF, the channel oscillator nominal frequency and in turn the conversion injection frequencies were interrelated. Possibly the 7.5 MHz was a given, if for example if had prior use in CB equipment, or suitable sideband filters were already available, although that seems less likely. Or perhaps it was a case of find a pair of first oscillator frequencies, symmetrically disposed about 27 MHz, that were workable, and for which half of their difference was workable as an IF.
In summary, from four CB transceivers, we find IFs as follows:
7.5 MHz Two cases, one as 1st IF, one as only IF
6.0 MHz One case as 1st IF
1.65, 1.75, 1.85 MHz One case as a set of 1st IFs
455 kHz One case as 2nd IF
260, 262 kHz Two cases as 2nd IF
The use of the standard 455 and 260/262 kHz numbers as final IFs is hardly surprising. For example, it would have allowed the use of standard, quantity-produced IF transformers. But 260/262 kHz would not have been suitable as a first IF at 27 MHz, and 455 kHz would have been at best marginal. Hence the use of double conversion with higher first IFs. Above 455 kHz, 1.65 MHz or thereabouts was the “next stop”, avoiding the MW broadcast band, and had some history in HF receivers, so was not unexpected.
Transceivers with synthesized frequency generation – all of those considered above – used the same synthesis circuitry to generate both the transmitted signal and the receiver first oscillator signal. The latter was obtained by adding or subtracting the first IF to the transmitter frequency. The increment/decrement would have needed to be large enough not to be problematical. I suspect that 1.65 MHz was probably about the minimum workable separation there, with higher numbers being seen as preferable. That was evidenced by the 6.0 and 7.5 MHz cases. That the latter appeared twice might in a sample of four, but it might also have been a reasonably common choice. Perhaps surprising is the absence of the 10.7 MHz standard number as a first IF, but given the small sample, that might not be indicative, and the Poly-Comm model discussed provides at least a vague suggestion that it had been used.
Cheers,
Steve
Posted : 12/04/2025 4:39 am
Page 9 / 9
Prev
