UK Vintage Radio Repair and Restoration Powered By Google Custom Search Vintage Radio and TV Service Data

Go Back   UK Vintage Radio Repair and Restoration Discussion Forum > Specific Vintage Equipment > Vintage Amateur and Military Radio

Notices

Vintage Amateur and Military Radio Amateur/military receivers and transmitters, morse, and any other related vintage comms equipment.

Reply
 
Thread Tools
Old 9th Feb 2021, 7:07 pm   #101
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 18,578
Default Re: ISB Receivers

Reopened at the request of John KC0G

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is offline   Reply With Quote
Old 12th Feb 2021, 7:02 pm   #102
John KC0G
Pentode
 
Join Date: Mar 2009
Location: Minneapolis, Minnesota, USA.
Posts: 206
Default Re: ISB Receivers

David, thank you for re-opening this thread.

In post #35, I posted a list of articles from Point-to-Point Telecommunication and Point-to-Point Communication which were published by the Marconi Co.. I have recently learned that these magazines and those of the successor publication Communication and Broadcasting are now available online via https://themarconifamily.pbworks.com...and%20Journals There are also searchable indexes which can be very helpful. I notice that the January 1974 issue of Point-to-Point Communication, which contained Sosin's infamous article on HF receiver reception factor, is there, but the revised article is not.

This Marconi archive is not complete - it has very little material from the Marconi Review, but anything is better than nothing. And a lot of the early ones are at www.worldradiohistory.com

Also earlier in this thread there was one or more reference to the Journal of the Institute of Post Office Electrical Engineers, aka POEEJ. These can be found at the web pages of Sam Hallas at http://www.samhallas.co.uk/ The early issues are not there, but some indexes are. I am not a "telecomms guy", but this web site is quite a resource for those who are. The GPO also did some serious work on short wave radio links.

I hope that some of you may find this useful. I am also thinking of compiling a (long) list of online resources for old magazines and journals.

73 John
John KC0G is offline   Reply With Quote
Old 12th Feb 2021, 9:25 pm   #103
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 18,578
Default Re: ISB Receivers

The GPO operated a number of HF links to various countries, and also the land based stations of the ship to shore network on MF/HF.

So they were quite seriously involved.

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is offline   Reply With Quote
Old 29th Jun 2021, 1:46 am   #104
Synchrodyne
Nonode
 
Join Date: Jan 2009
Location: Mt. Maunganui, New Zealand
Posts: 2,700
Default Re: ISB Receivers

Some pertinent POEEJ articles were mentioned in post #61, https://www.vintage-radio.net/forum/...4&postcount=61.


Cheers,
Synchrodyne is offline   Reply With Quote
Old 29th Jun 2021, 1:55 am   #105
Synchrodyne
Nonode
 
Join Date: Jan 2009
Location: Mt. Maunganui, New Zealand
Posts: 2,700
Default Re: ISB Receivers

Returning to the original question, back in 2014, which was:

Quote:
Originally Posted by Synchrodyne View Post
What happened, in a general sense, with point-to-point ISB receivers following the general industry transition from valves to solid state devices?
In a broad sense, I think that the answer as to what happened and what the situation looked like at the end of the transition period might well be illustrated the Marconi H2540 receiver released in 1976, and described in the Marconi journal “Communications and Broadcasting” Volume 4, No. 1, 1977, Autumn, p.12ff, available here: https://themarconifamily.pbworks.com...20Broadcasting.

The H2540 was essentially a single-box ISB receiver that had top-of-the-line performance, with stability that allowed operation of point-to-point links without pilot carriers and the associated AFC, although an AFC facility was available for use with links that were not so frequency-stable at the transmission end. It was also suitable for remote as well as local control. Marconi did not exactly say that the H2540 displaced all that went before it in its point-to-point range, but one gets the impression that as part of an integrated offering including the H1540 transmitter drive, that was the basic intention. The full-height rack MST receivers from the mid-1960s and the more compact Hydrus model from 1968 might have faded from the scene fairly soon after the H2540 appeared.

Click image for larger version

Name:	Marconi H2540 Block Schematic.jpg
Views:	46
Size:	44.6 KB
ID:	236707

Essentially, it looks as if by 1976 a one-box receiver could do what was previously done by the solid-state MST models and before that by the valved HR92/HR93. And it could it with better stability, and it was easier to operate.

The H2540 was also a general-purpose receiver, in that it covered the range 15 kHz to 30 MHz, and had spin-wheel manual tuning as well as frequency setting. I suspect though that it might have been in a price group that was a level above the general run of general-purpose offerings of the time.

I should imagine that Marconi was keeping pace with general developments in the point-to-point ISB field, was probably at or near the leading edge. In that sense the H2540 was a major marker. Prior to that, the major markers in the Marconi range would have been the MST receivers in 1965, the HR92/HR93 models in 1952, and the CRD150-20B + SSR2 combination in 1948.

Against that, a previously mentioned article may be seen as representing something of a waypoint between 1965 and 1976. It was “A Survey of Communication Receivers”, in Point-to-Point Communication, Vol. 14, no. 3, August 1970, p.110ff (https://themarconifamily.pbworks.com...nt%20magazines).)

In this table drawn from that article one may see a hierarchy of point-to-point ISB/SSB receivers from the MST downwards.

Click image for larger version

Name:	from Point-to-Point 197008 p.115.jpg
Views:	34
Size:	50.5 KB
ID:	236708

Within that group, the H2900, which was intended primarily for point-to-point work, looks like a stepping-stone to the H2540. Of it, Marconi said: “A dual diversity i.s.b receiver with this degree of sophistication would be expected to occupy a 7ft cabinet, whereas the dimensions of the H 2901 are only 7.0 x 17.5 x 22in, weighing only 901b”. From that it may be inferred that the intent was to replace the MST (7 ft high) with a single-box device. But that intent evidently did not come to full fruition until the H2540.

The 1970 table also shows the Eddystone EC958 receiver in a supporting role at the lower end of the price/performance range. This was continued and in fact enhanced, in that an ISB version, the EC958/12, was developed in 1974, based upon the improved EC958/7 of 1973. At this level, there would also have been the ISB versions of the general-purpose single-box HF receiver offerings from Plessey, Racal, Redifon and others. Apparent successor to the EC958/12 was the 1650/3, the ISB version of the 1650 of 1984, Eddystone’s first upconversion general coverage HF receiver. It was described in “Communications and Broadcasting”, Volume 9, No, 3, 1985 February, p.23ff.

The distinction between the Marconi H2540 (and any other similar point-to-point oriented receivers) and the top-of-the-line general coverage receivers of the time was probably a slim one. For example, in the latter category, the Plessey PR2250 of late 1977 was described as being primarily for surveillance and monitoring. Nonetheless it appeared to have had an ISB facility, with carrier AFC, built in, and it could be remote controlled. It was described in Wireless World 1977 December, p.86.

Click image for larger version

Name:	from WW 197712 p.86.jpg
Views:	33
Size:	46.6 KB
ID:	236709


Cheers,
Synchrodyne is offline   Reply With Quote
Old 29th Jun 2021, 5:27 am   #106
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 18,578
Default Re: ISB Receivers

That made interesting reading!

I've still got the H2900 structure in my head, so looking at the downloaded article on the H2450 made for interesting comparisons and good example of how technologies moved quite a bit at that time.

Both receivers were up-converters of course. H2900 to a helical filter around 70MHz, whose width was sufficient to allow the synthesiser to be split with discrete crystal oscillators for 2MHz steps, and finer steps done with the second local oscillator. (Any error in the crystal oscillators being passed to the 2nd LO to offset it to correct the tuned frequency.

The H2450 is a lot more mainstream to today's eyes. It has a narrow crystal filter at the first IF of 68.6MHz chosen so that a simple 2nd LO of 70MHz does the mix-down to the 1.4MHz final IF. I still have a pair of these filters, and a large collection of 1.4MHz filters for all sorts of modes.... never known to pass a good crystal filter at a radio rally! All steps of the frequency control being implemented at the first LO by a multiloop PLL synthesiser.

The H2540 is more sophisticated than the H2900 in avoiding a wide first IF and getting a narrow filter earlier in the structure. This should help with intermod and overload characteristics. The H2900 is more sophisticated than the H2540 in having a servomotor-tuned narrow preselector, while the H2540 has only sub-octave switched filters, and optional at that.

From an intermod point of view, the H2900 is likely to be better with widely spaced interferers, but the H2540 should win with closely spaced interferers due to the difference in width between a tuneable RF preselector and a narrow, crystal, first IF filter. Assuming some progress in the detail of mixer design happened in the period between the sets. The H2900 mixer was not very advanced. Racal was ahead of the game at that time and I assume the Rafuse patent would have still been in force.

The H2900 may be the more sophisticated one in terms of phase noise with that switched bank of crystal first LOs.

The H2900 also is dual-diversity in one box. H2540 users would simply buy two boxes and get a degree of redundancy as a bonus, losing a number of single points of failure which would take out both diversity channels at once.

So, if it came to a shoot-out between H2900 and H2540 each one would beat the other in different parameters, I expect. The H2540 must have been a better fit to the market. I was told that very few H2900s were sold, that Interpol was the main user.

In the comparison table in the H2540 article the H2900 looks to tick all the boxes, except one: Merchant marine. I doubt if they could have afforded one!

These receivers cover an interesting transition phase in receiver architecture and the H2540 shows things starting to converge on what became a standard approach until the onset of digital conversion at IF.

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is offline   Reply With Quote
Old 29th Jun 2021, 1:12 pm   #107
gm0ekm cecil
Tetrode
 
Join Date: Jan 2008
Location: Shetland, UK.
Posts: 70
Default Re: ISB Receivers

Quote:
Originally Posted by Synchrodyne View Post
Hi David:

Thanks for that.

The BBC situation was described in BBC Engineering #84, 1970 October (http://downloads.bbc.co.uk/rd/pubs/a...neering_84.pdf) which noted that the BBC was expanding the use of HF SSB program links, having started in 1964, with use of ISB receivers for HF AM reception predating that. But the article deals mainly with the SSB transmission, and does not mention the receivers that were in use.

I imagine that the ISB receivers were used both to provide sideband diversity when receiving AM broadcasts, and perhaps more so to mitigate selective fading distortion. As far as I know the BBC did not use ISB program links, although the VOA did (and perhaps still does). Way back then sometimes I used to listen to them using the Liniplex F2.

The Quartz Hill operation is described here: http://www.zl6qh.com/000468.html. From the 1960s, the primary receivers for the BBC World Service were of the Marconi HR21 ISB type. To quote from that website: “Two Marconi HR21 independent-sideband HF receivers were the mainstay of the station’s receiving equipment. They were used primarily for maintaining a round-the-clock feed of the BBC World Service to Broadcasting House. Designed for use on international radio-telephone circuits, they were purchased in the early sixties when the advantages of using the sideband reception technique for short-wave broadcasts were finally recognised. These were the ability to select the sideband with the least interference (especially useful in eliminating heterodynes or whistles), and a substantial reduction in the audio distortion caused by selective fading. Each HR21 had separate 6kHz bandpass filters and amplifiers for the upper and lower sideband, and its two audio output channels were fed to faders on the control panel, thus enabling a smooth changeover to the better channel. Each receiver employed about 65 valves (vacuum tubes), and comprised several separate units occupying a complete rack about two metres high. They were relatively difficult to tune and to service, but their shortwave reception quality was superior to that of the nine GEC BRT-402 general-coverage receivers that continued to be used for medium-frequency and less important short-wave reception.”

I drove past the Quartz Hill site a few times back in the old days when it was fully operational, and once I think up the hill to the gate, but never saw inside it. It certainly had a bleak, remote feel about it, even though it was maybe only 40 minutes from the Wellington suburbs.

Cheers,
A photo of my former radio museum showing the Marconi HR21 ssb receiver on the right. .... GM0EKM.
Attached Thumbnails
Click image for larger version

Name:	Display 001.jpg
Views:	105
Size:	51.6 KB
ID:	236726  
gm0ekm cecil is offline   Reply With Quote
Old 30th Jun 2021, 1:54 am   #108
Synchrodyne
Nonode
 
Join Date: Jan 2009
Location: Mt. Maunganui, New Zealand
Posts: 2,700
Default Re: ISB Receivers

The HR21 was rather large as compared with the H2540! And it was Marconi’s second-line point-to-point receiver, below the HR93, although I think fairly close in performance. I imagine that the H2540 equalled or bettered the HR93.

Re the H2900, in hindsight one might say that it was something of a “concept” receiver, with the H2540 the production realization of that concept, as refined by market response, intervening technology developments, etc.

The H2540 employed the same synthesizer as used in the H1540 transmitter drive, the latter being described in Communication & Broadcasting Volume 3, No.1, 1976 Summer, p.13ff. According to an article “Versatility in h.f. systems” in Communication & Broadcasting Volume 4, No.2, 1978 Spring p,16ff, that whole product line was referred to as “MFT”, Marconi Fast Tuning, and does appear to have been the successor to the MST product line.

The next step was the incorporation of microprocessor control, in what was called the MFT2 range, MFT 2nd generation. This was described in an article “MFT2 h.f. systems – synthesized drive and receiver” in Communication & Broadcasting Volume 6, No.2, 1981 March p.19ff. The receiver was the H2541, which appears to have been an updated H2540, retaining the same basic layout and IFs.

Click image for larger version

Name:	Marconi H2541 Block Schematic.jpg
Views:	28
Size:	48.0 KB
ID:	236767



Cheers,
Synchrodyne is offline   Reply With Quote
Old 30th Jun 2021, 2:52 am   #109
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 18,578
Default Re: ISB Receivers

What had happened was that H2900 to H2540 had two steps in technology that were inter-related.

The H2900 had some of the synthesiser functionality exported from the synthesiser and implemented by the choice of receiver stip architecture.

The H2540 made the step change to an up-converter with a narrow crystal roofing filter in the first IF. It also changed the synthesiser to multi-loop to implement all the tuning steps at the first LO.

This receiver structure then stood the test of time and was used up to the point where DSP techniques took over the later second IF functions, rolling backwards towerds the front end as it became more capable and ADCs progressed.

There was still progress to be made in the synthesiser area. The multiloop design was probably faster tuning than the ultra-slow loop design in the H2900 but it was also likely to be worse on phase noise.

Receiver design went through several epochs where different performance parameters took their turn as the primary limitation. Once it was getting enough gain, then it became getting narrow enough selectivity, then it became getting enough frequency stability to make those narrow bandwidths useable without needing constant adjustment by a real-time skilled operator. The synthesiser fixed the stability problem while introduceing the last problem: Phase noise!

The addition of a microprocessor made remote control easier and introduced user conveniences like memories but it also freed sysnthesiser design from having to follow the partitioning of the decimal format tuning number. It became possible to use simpler binary counters in programmable dividers and to implement more complex DAC operated pretune schemes which eased limitations on frequency-adding loops.

Around this time in history, I was designing a 5-loop synthesiser at HP for a fast-tuning synth with very low, and carefully shaped for the application, phase noise. That's what's in the HP journal article by Guy Douglas and myself. So I followed the changes in sysnthesiser technology with quite a lot of interest as it was happening. The generation of receiver synths after the multiloop saw a DDS replacing the lower order loops in a 5-loop structure or the single lower order loop of a 3-loop structure. This was good as it reduced cost, size and power consumption while making tuning quicker and reducing phase noise. The bandwidth of the PLL following the DDS acted to clean up the DDS' main limitation, quantisation noise (Quantisation both from truncation of the phase word, and of the amplitude word getting truncated by the DAC)

Later still synthesisers were steps back from the performance the PLL/DDS hybrid gave. On one hand the implementation of a single loop synthesiser with noise-shaping fractional-N, and on the other hand doing the entire mix down with a DDS..... The all by a very fast DDS is what lives in DSP inside the very latest architecture where the whole HF band is stuffed into an ADC and everything after is DSP. To keep the look-up tables finite and the multipliers finite in bit width, truncation in phase and magnitude both go on.

Potted history, but I lived through it and made a couple of contributions along the way.

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is offline   Reply With Quote
Old 30th Jun 2021, 10:36 pm   #110
Bill MCSL
Diode
 
Join Date: Jun 2016
Location: Whitby, North Yorkshire, UK.
Posts: 2
Default Re: ISB Receivers

This is my first post, and having followed this topic for most of the last 7 years I have to say that the quality of the discussion is superb.

I possess none of the receivers under discussion (although a Marconi ICS3 is winging its way as I write!) but I do have a H1541 drive, which has the same architecture as the H2540 receiver, ie a narrowband IF of 68.6Mhz and a second IF of 1.4Mhz.

For the later H2550 receiver again a similar structure is used, but with figures of 62.5MHz and 2.5MHz respectively.

From the earlier period, the ICS3 and H2900 have identical architectures, and probably at least similar circuitry. In fact, I suspect that the H2900 did indeed turn out to be a "concept" design, produced in very limited numbers, but probably sporned the slightly later ICS3, which was produced in relatively large (for Marconi) numbers.

But talk of architectures ignores the important related topics of mechanical construction and intended market. Take for example the H2540 and H2550. Assuming the former construction is similar to the related drive, it is quite clear that the H2540 is intended for a fixed location only, whilst the H2550 has a much more rugged construction, suitable for the shock and vibration environment of shipborne operation. I have only seem rather poor photographs of the H2900. but again I would say it is designed for a fixed location, whilst the ICS3 was widely used by the Royal Navy and US Navy
Bill MCSL is offline   Reply With Quote
Old 1st Jul 2021, 10:14 am   #111
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 18,578
Default Re: ISB Receivers

The H2900 I had was a prototype and the main case was milled out of a single block of aluminium.

Bizarrely, the input output connectors were all accessible from the front panel. Two knurled screws released the panel section with the meter on it, so it could come away on a double hinged strut ro support it. Behind this panel were the input, output and interface connectors. There was a cable duct machined into the right hand side of the big alloy lump, with a sheet metal cover over it.

There was no anti shock or anti vibration measures taken, and with those connections, it was obviously destined for a rack. So maybe they assumed the whole rack would be on shock mounts?

Shipborne radios, particularly warships endure RF hell with multiple radios and multiple systems having to share antennae based on a single mast. The need for excellent blocking and intermod performance is legendary. Apart from the motor tuned preselector, the circuitry in the H2900 isn't going to be wonderful in these respects. It suffers from a lot of the optimism common in early transistor designs. Racal made a similar assumption with their RA217/RA1217 design. The RA17 wasn't great at dynamic range, but the transistor versions were a significant step backwards.

I decided I'd never use the H2900. Slow thumbwheel tuning isn't fun.

Sometime around 1970, I got a tour of a Canadian warship on a goodwill tour, anchored just off Whitby harbour, and I'll be passing through on Saturday, heading for Scarborough.
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is offline   Reply With Quote
Old 4th Jul 2021, 3:07 am   #112
Synchrodyne
Nonode
 
Join Date: Jan 2009
Location: Mt. Maunganui, New Zealand
Posts: 2,700
Default Re: ISB Receivers

I imagine that the construction of the H2900 and its tuning method were oriented to its apparent primary end-use, which was in fixed point-to-point service, where frequency changes were not that frequent and involved resetting to a defined channel rather searching for a desired channel. At least that application was given at the start of the pertinent Point-to-Point article:

Click image for larger version

Name:	Point-to-Point 197001 p.04.jpg
Views:	30
Size:	50.6 KB
ID:	237046


The Marconi applications list from Point-to-Point 1970 August shows the H2900 as being suitable for shore-to-ship generally, and for naval ship-to-shore, but not for mercantile ship-to-shore operations. I have read that shore station receivers did need to have some search tuning capability given that mercantile ship transmissions could vary somewhat from their nominal frequencies. So that may have been a reason for excluding the H2900. That said, I also understand that the GPO shore stations were equipped with SSB/ISB receivers well before SSB was introduced for what might be called “formal” marine communications. As noted upthread, passenger vessels had been using SSB for passenger radiotelephone services since 1949.

Click image for larger version

Name:	from Point-to-Point 197008 p.115.jpg
Views:	23
Size:	59.4 KB
ID:	237047

All the receivers in the shore-to-ship list, namely the Hydrus, H2900, EC958 and H2310 Argo would I think have been additional to whatever approved receivers were carried by a vessel since as far as I know, none of that group were actually approved.

Certainly, the Eddystone EC958 was designed with the then-forthcoming marine SSB requirements in mind, but not all versions would have been fully compliant. The EC958/5 was the fully compliant marine variant, rebranded as the Marconi (Mimco) Nebula, and released in 1971 at the same time as the Marconi Apollo, which was designed by MWT specifically for Mimco, although MWT also sold it as the N2050 general-purpose receiver. The H2310 Argo was a modified version of the EC958, with further improved stability (beyond that provided by the high-stability mode Wadley loop in the original) to suit naval applications. It was also said to have had some constructional changes. These might have included an element of “ruggedizing”. I should guess that shipboard use of the Hydrus would have been for “business” SSB/ISB voice, VFT, etc., communications.

A couple of Point-to-Point articles illustrate the diversity of shipboard radio equipment.

“A Report on Trials of a Ship-Shore Data Transmission System” - 1964 October p.5ff;

“Q.E.2 – Its Newspaper and its Communications” – 1971 May p.125ff. (Note that the QE2 had both Piccolo and Lincompex equipment, and that its HF receiver complement included the EC958, GEC RC411 and Eddystone 830/12.)


Cheers,
Synchrodyne is offline   Reply With Quote
Old 20th Nov 2021, 3:26 am   #113
Synchrodyne
Nonode
 
Join Date: Jan 2009
Location: Mt. Maunganui, New Zealand
Posts: 2,700
Default Re: ISB Receivers

In respect of the Marconi H2900, as well as the previously mentioned Point-to-Point 1970 January article, there is another, also by B.M. Sosin, in “The Radio & Electronic Engineer”, 1971 July.

This covered much of the same ground, but did go into more detail on some aspects, such as the choice of the triple conversion scheme and its associated intermediate frequencies.

It also included a signal level map through the receiver, showing minimal gain and AGC action until the 3rd IF strip:

Click image for larger version

Name:	Marconi H2900 Signal Level Map.jpg
Views:	17
Size:	41.8 KB
ID:	246179


In comparison, here is the map for the HR92/HR93, Marconi’s top-of-the-line point-to-point SSB/ISB receiver (valve) of 1952. The HR92 was the double-diversity version, occupying two 7 ft high racks.

Click image for larger version

Name:	Marconi HR92 & HR93 Signal Level Map.jpg
Views:	16
Size:	55.2 KB
ID:	246180


There was also a discussion of reciprocal mixing, including this chart:

Click image for larger version

Name:	Marconi H2900 Reciprocal Mixing Chart.jpg
Views:	13
Size:	42.6 KB
ID:	246181


That looked somewhat familiar; I had previously seen the similar chart that Racal had used in its 1974 articles on the RA1772 (*):

Click image for larger version

Name:	Racal Reciprocal Mixing Chart.png
Views:	18
Size:	31.6 KB
ID:	246182


Racal went one down on Marconi, as it were, including a 100 dB line (representative of the RA1772) as well as the 70 and 90 dB lines used by Marconi.



Cheers,


(*) The two articles, both by R.F.E. Winn (apposite initials), were:

“Operational Aspects of HF Receiver Design”; Electronics & Power 1974 June 13.

“Synthesized Communications Receiver”; Wireless World 1974 October.
Synchrodyne is offline   Reply With Quote
Old 20th Nov 2021, 9:52 am   #114
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 18,578
Default Re: ISB Receivers

Having very little AGC until the 3rd IF has the consequence of needing dramatic dynaic range figures deep into the receiver from the antenna socket onwards. But ISB receivers are stuck with this. The gain of the stages before the channel-defining filters split the two signal paths are common to both USB and LSB so any AGC action there will affect both channels and they may have different needs, so to stop cross-channel AGC effects, the dynamic range bullet had to be bitten.

Modern crystal filter developments now permit SSB and even CW bandwidths to be achieved at 60-odd MHz IFs. The Icom IC7700 in the shack here has switchable mode-related first IF filters. This does wonders for dynamic range, but there is no possibilities for this in an ISB receiver, unless the sideband split was at a high first IF.

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is offline   Reply With Quote
Old 23rd Nov 2021, 1:24 am   #115
Synchrodyne
Nonode
 
Join Date: Jan 2009
Location: Mt. Maunganui, New Zealand
Posts: 2,700
Default Re: ISB Receivers

The block schematic for the H2900 shows the separate AGC systems for each sideband, the two being combined to provide (evidently very mild) AGC for the RF and 30 MHz IF amplifiers.

Click image for larger version

Name:	Marconi H2900 Block Schematic.jpg
Views:	10
Size:	33.9 KB
ID:	246407

In the case of the HR92, the distribution of AGC bias was not shown on the block schematic, but was described as being fed back to the 2nd (of 2) RF amplifier stage, both 1st IF (1.6 MHz) stages, and as being fed forward to the 1st sideband amplifier stages. Thus there were three earlier stages with AGC, and just one in each sideband amplifier, suggesting that a reasonable proportion of the AGC action happened quite early on.

Click image for larger version

Name:	Marconi HR92 Block Schematic.jpg
Views:	4
Size:	56.2 KB
ID:	246408

Between the HR92 and the H2900 was the H2102/2112 MST receiver, in 1964. As this apparently used silicon bipolar devices, I’d guess mostly of the small-signal type, with limited dynamic range, I should imagine that quite a proportion of the AGC action took place in the RF and 1st IF (2 MHz) sections. It occupied one 7 ft rack, so was half the size of the HR92.

Before the HR92 was the CRD150/20B +SSR2 triple diversity combination of c.1948, which like the following HR92, occupied two 7 ft racks. It was more or less a receiver/adaptor combination with some integration, with the AGC bias fed from the adaptor side. The CRD150 was the diversity variant of the CR150, in this case the version with 1.2 MHz rather than 1.6 MHz 1st IF (CR150/4?). In this block schematic, the adaptor section starts at FC3.

Click image for larger version

Name:	Marconi CRD150-20B- SSR2-2 Block Schematic.jpg
Views:	5
Size:	36.6 KB
ID:	246409


Cheers,
Synchrodyne is offline   Reply With Quote
Old 23rd Nov 2021, 6:48 am   #116
Radio Wrangler
Moderator
 
Radio Wrangler's Avatar
 
Join Date: Mar 2012
Location: Fife, Scotland, UK.
Posts: 18,578
Default Re: ISB Receivers

It was a time of fast progress in receiver architecture, but the immense number of models Marconi did in the sixties and early seventies makes it look a miracle that they ever turned a profit considering the probable shipment numbers per model versus development costs. No wonder they were scared of Racal with there more carefully managed portfolio.

In the H2900 block, the 79.8 to 81.8MHz filter is a helical design.

The 81.8 to 109.8MHz first LO is one of a switched bank of free-running low phase noise crystal oscillators in a massive block of aluminium to slow temperature changes.

The 49.8 to 51.8 MHz LO is from the PLL of the synthesiser. The selected first LO crystal oscillator signal is used as a mixed-in offset in this loop, so that the error between the crystal's actual frequency and the exact frequency it should have been is accounted for in the offset of the second LO (Shades of Wadley!)

The RF amplifiers in the thing use 2N4391 series JFETs. These are thought of as low RDSon analogue switches with high IDss, but this is where I discovered that these large area JFETs, when heat-sunk and run at high quiescent current made fairly high dynamic range RF amplifiers.

The big disappointment at the circuit design level are the mixers.

The AGC is as it has to be for a single RF structure doing ISB, and it pushes greater demands onto the RF section. I begin to wonder how much more expensive two independent receivers would have been once the go narrow as soon as possible ethos had taken over non-ISB receivers?

I think I have a couple of Marconi's 68.6 narrow first IF filters from when Marconi went down this route. 70MHz LO to convert down to 1.4MHz IF, and you can divide 70 down to make the CIO with filters properly offset for USB and LSB.

These structures lock-out the trick of using LO2 to implement the fine resolution steps of the synthesiser. LO1 has to do everything, and there goes the low phase noise crystal oscillators. But it's not so dire. The H2900 LO1 may have been wonderfully clean, but the wide first IF planted obligations for equal cleanliness on the second LO.

Marconi's synth loop for the H2900 did not use programmable dividers, it used an older technology, rate multipliers, with some effort to try to break up structural spurs when set to cardinal points in the range.

David
__________________
Can't afford the volcanic island yet, but the plans for my monorail and the goons' uniforms are done
Radio Wrangler is offline   Reply With Quote
Old 23rd Nov 2021, 7:35 am   #117
Synchrodyne
Nonode
 
Join Date: Jan 2009
Location: Mt. Maunganui, New Zealand
Posts: 2,700
Default Re: ISB Receivers

Quote:
Originally Posted by Radio Wrangler View Post
It was a time of fast progress in receiver architecture, but the immense number of models Marconi did in the sixties and early seventies makes it look a miracle that they ever turned a profit considering the probable shipment numbers per model versus development costs. No wonder they were scared of Racal with there more carefully managed portfolio.
I suspect that Marconi (MWT) made its money selling and supporting equipment for the middle/upper end of the point-to-point market, with the HR92/93 (1952), the MST (1964) and the H2540 (1976), in that sequence treating the H2900 as a concept receiver rather than the genuine heir to the MST. At the middle level, the HR21 family was essentially based upon the HR92. The Hydrus of 1968 looks to have been a unique approach to the lower cost end of the point-to-point market. But MWT did not seem to make a big effort in the general-purpose market, offering the CR150 series, then the NS702 (essentially the Atalanta marine receiver), the H2301 (Eddystone 880), followed by the N2050 (the Apollo marine receiver). I could well be wrong here, but it looks as if Marconi was saying “if you want us to supply some GP receivers to round out your complement in a point-to-point station that we are equipping, then we do have some”. On the other hand, if GP receivers were your main concern, then Racal would look probably more attractive – which is why it sold around 20 000 of its RA17 model.

By the time that the H2540 was released, it might have been difficult to distinguish between that and a fully up-optioned RA1772, with ISB, AFC and so on. But then Marconi was selling complete point-to-point stations of which the H2540 was just one part. I imagine that quite a few stations had both the H2540 and the RA1772, the latter as an “operator’s receiver”. I think that the Eddystone EC958 was popular in that category, as well.


Cheers,
Synchrodyne is offline   Reply With Quote
Old 24th Nov 2021, 9:10 pm   #118
John KC0G
Pentode
 
Join Date: Mar 2009
Location: Minneapolis, Minnesota, USA.
Posts: 206
Default Re: ISB Receivers

Re post #116 and the large number of Marconi receivers, this reminds of when I first read the article "Survey of Communications Receivers" by J. V. Beard in Point-to-Point Communication, August 1970, pp 110-120. This article has been mentioned before in this thread. See: https://themarconifamily.pbworks.com...and%20Journals

I had a feeling that Marconi had deigned and manufactured a bunch of different receivers to meet a bunch of different specifications. As such they were tailored solutions produced by an engineering company. The Racal approach with the RA1772 seems to have been different, ie they made a common platform which met most needs and could take options for special needs. Others may disagree?

@Synchrodyne, re. post #117. I have seen the number of 20,000 before as the number of Racal RA17 receivers which were built. Do you have any sources for this data?

73 John KC0G/M0KCY
John KC0G is offline   Reply With Quote
Old 25th Nov 2021, 1:31 am   #119
Synchrodyne
Nonode
 
Join Date: Jan 2009
Location: Mt. Maunganui, New Zealand
Posts: 2,700
Default Re: ISB Receivers

Quote:
Originally Posted by John KC0G View Post
@Synchrodyne, re. post #117. I have seen the number of 20,000 before as the number of Racal RA17 receivers which were built. Do you have any sources for this data?
Just one, namely this article:

Company Profile – Growing with Racal
John P. Wilson
The Radio & Electronic Engineer; 1980 October; p.499ff

It is findable on the web, although I did not record where. It came up during a “lateral search” session, rather than as the result of a specific search. Anyway, please PM me if you want a .pdf.

Here is the pertinent part:

“The company's first identifiable Racal product was a high-frequency communications receiver, the RA17, designed in the mid-50s around the triple-conversion drift-cancelling technique originated by Dr. Trevor Wadley, the technique eventually becoming known as the Wadley Loop. The idea had been canvassed round the established companies and rejected as being `too difficult' to achieve in practice. Racal engineers undertook the project and with assistance from Admiralty engineers in the final stage of design solved the outstanding problem of internally generated spurious signals. The RA17 was the most successful general-purpose h.f. receiver of its generation with some 20,000 units sold throughout the world.

“Two other notable achievements of the period were the first Racal digital counter/timer, and an unpublicized contract from the Ministry of Defence for the development and production of experimental single-sideband equipment, this providing Racal with early experience on a major new technique.

“Building on the success of the RA17 the company soon developed a range of transmitters and, more important, became engaged in the development of frequency synthesizers.

“In the 1960s Racal, in common with its competitors, experienced the trauma of the transition from electronic tubes to solid-state with varying results. A transistorized version of the RA17 receiver, for example, was only partially successful with the advantages of size, weight and heat dissipation offset by a degraded electrical performance. The later RA1770 series of h.f. receivers introduced in 1973 was a worthy successor and became another world leader, winning the Queen's Award for Technological Achievement.”


Cheers,
Synchrodyne is offline   Reply With Quote
Old 25th Nov 2021, 1:34 am   #120
Synchrodyne
Nonode
 
Join Date: Jan 2009
Location: Mt. Maunganui, New Zealand
Posts: 2,700
Default Re: ISB Receivers

Quote:
Originally Posted by John KC0G View Post
Re post #116 and the large number of Marconi receivers, this reminds of when I first read the article "Survey of Communications Receivers" by J. V. Beard in Point-to-Point Communication, August 1970, pp 110-120. This article has been mentioned before in this thread. See: https://themarconifamily.pbworks.com...and%20Journals

I had a feeling that Marconi had deigned and manufactured a bunch of different receivers to meet a bunch of different specifications. As such they were tailored solutions produced by an engineering company. The Racal approach with the RA1772 seems to have been different, ie they made a common platform which met most needs and could take options for special needs. Others may disagree?
I think that that is a good assessment of the situation. The naval receivers in particular seem to have been precisely tailored to the users’ needs. The point-to-point receivers may have been developed to some extent based upon MWT’s assessment of user needs derived from its long experience in the field. The Eddystone EC958 was more in the nature of an expandable common platform, including ISB capability. Development was started around the time that Marconi acquired Eddystone, so just how much of it was legacy Eddystone and how much was MWT input is hard to say. That the design encompassed mercantile marine main receiver requirements may have been the result of MWT (and Mimco via MWT) influence, as previous Eddystone general-purpose HF receivers did not. Mimco adopted the EC958/5 as its Nebula in 1971, as a companion to the Apollo (MWT N2050). Before that MWT had modified the EC958 for higher stability with a frequency-counting system for the interpolation oscillator as the H2310 Argo (EC958/4, I think) for naval use. The ISB versions were the EC958/9 (very obscure) and EC958/12), both I think derivatives of the 2nd iteration EC958/7. The EC958/12 at least followed the “built-in adaptor” pattern, with a separate ISB section taking the 100 kHz final IF from just after the final signal mixer. The USB, LSB and carrier channels each had their own AGC loop. (RF AGC on the EC958 was anyway a separate, albeit narrowband, loop.)

As said before, the H2900 demonstrated that a top-level point-to-point ISB receiver could be condensed into a single-box. Then the H2540 was the convergence point where the specialist point-to-point and general purpose types of receiver were effectively combined.

The timeline looks something like:

1952 HR92 – about the pinnacle of valve point-to-point ISB receivers.
1964 MST – change to synthesized LO and to solid state.
1970 H2900 – single box.
1974 Racal RA1770 series – a new standard in GP receiver performance and operability, also configurable for ISB.
1976 Marconi H2540 – top level point-to-point and GP combination, with remote controllability.


Whether Marconi had planned the H2540 before the RA1770 series made its appearance is unknown, but I suspect that it did take notice of what Racal had done.


Cheers,
Synchrodyne is offline   Reply With Quote
Reply

Thread Tools



All times are GMT. The time now is 5:12 pm.


All information and advice on this forum is subject to the WARNING AND DISCLAIMER located at https://www.vintage-radio.net/rules.html.
Failure to heed this warning may result in death or serious injury to yourself and/or others.


Powered by vBulletin®
Copyright ©2000 - 2021, vBulletin Solutions, Inc.
Copyright ©2002 - 2021, Paul Stenning.