Redifon R550A

[deliverance]

Hi all. Recently a acquired Redifon R550A and I can find no info on the net. It's similiar to the R551 but there are differences has anyone come across one?

Thanks, Chris.

[Herald1360]

Did Big Al... not know anything about it?

Maybe ms660 can unearth something suitably obscure to help?

[deliverance]

Al bought it years ago from a rally I think and has never touched it .

[Synchrodyne]

The Redifon R550 appears to have been the general purpose version, announced in Wireless World (WW)1969 March:

WW 196903 p.136,137 Redifon R550.pdf


The R551 was the marine version, announced in WW 1970January:

WW 197001 p.41 Redifon R551.pdf


How they differed though is not clear from the limited information provided in WW.

As far as I know both were based upon the R499 as the core (2nd IF onwards). The R499 was a fixed-frequency channelized SSB receiver that had been announced in WW 1968 June:

WW 196806 p.184 Redifon R499, MCU6.pdf


The synthesizer in the R550/R551 appears to have been a little unusual. The 39.4 MHz oscillator feed to the 2nd mixer came not from the synthesizer, as might have been expected, but from a separate crystal oscillator. But because this might drift, the 39.4 MHz was also injected into a drift-cancelling loop that involved the splitting the main VCO combination into two sections. Why it was done this way is beyond me, but one consequence was that the feed to the main VCO main PLL was in the range 3 to 32 MHz rather than the 38 to 68 MHz range of the feed to the 1st mixer. One consequence was that the 2nd conversion had to be supradyne. The apparent reference oscillator for the synthesizer was the 1.4 MHz demodulation carrier insertion oscillator.

Redifon R551 Block Schematic.pdf

With the following R1000 model (1978) Redifon apparently reverted to a more conventional approach. It used the same 38.0 and 1.4 MHz IFs, but the second conversion oscillator feed came from the synthesizer and was infradyne (36.6 MHz).

This group of Redifon receivers, R499, R550 and R551, appears to have been one of the first to use 1.4 MHz as a final IF for SSB processing, although that number, or thereabouts, had seen prior use as a 1st IF. One could say that the R499, single-conversion, bridged between the 1st IF and final IF applications. Subsequently the 1.4 MHz final IF was found in, for example, Racal, Marconi, Eddystone, and Plessey receivers.


Cheers,

[Radio Wrangler]

Originally Posted by Synchrodyne

Why it was done this way is beyond me, but one consequence was that the feed to the main VCO main PLL was in the range 3 to 32 MHz rather than the 38 to 68 MHz range of the feed to the 1st mixer.

It's quite an early synthesiser from the point of view that a lot of developments we take for granted occurred after it.

The use of a mix-down in the main loop was used for two reasons: Firstly, it reduced the maximum frequency that the programmable divider had to work at. This was valuable in the era before fast logic, and the two-modulus prescaler. Secondly, by reducing the N number of that divider, it reduces the multiplication of phase noise vreated in the phase detector, divider and reference.... so long as the mix-down signal is clean.

There is a description of this scheme by the staff of Collins Radio in "Single sideband circuits and systems" book by Sabine, Schoenike et al. I think it also cropped up in the Collins KWM380 amateur transceiver.

One wierdness is that depending on the frequency plan, tuning can be reversed, the highest output corresponding to the lowest N. The issue of arithmetic code conversion wasn't trivial in those days and programmable dividers were usually arranged as BCD machines rather than straight binary. Fun and games with switches in this era.

A programmable divider was often done by loading a number into a counter, then counting downwards. Several counts before hitting zero, a second, faster counter would be triggered, and this would manage the pre-loading of the main counter. This allowed a couple of entire clock periods for the main counter to load and settle before the next counting sequence had to begin. The main counter might be plain TTL, while the preload controller might be a state machine in 74H series.
74H logic was desperate stuff that tried throwing power at the problem of getting more speed. Schottky-clamped logic was still some years off.

Having a crystal oscillator for the second LO, avoids the phase noise of using a PLL with a VCO. Because of the accuracy spec they wanted, they either needed to phase lock the crystal in a very narrow bandwidth loop, or account for its error in the tuning of the main synthesiser.

That WW product announcement page is a classic! The receiver was in the company of the CDU150 scope and the Decca record player arm.

I've never had any of the Redifon receivers, but did once have a GK203 exciter as my main HF transmitter (with a homebrew 10W power amp) so I don't know my way around them.

David

Note, The Collins book is the second one, the earlier one was by Pappenfus, Schoenike et al.

[deliverance]

Thanks for the info Synchrodyne / David there are differences between the two and I need to find more info thanks again .

[Radio Wrangler]

The 550 is 'general purpose' while the 551 is 'marine'

There are some quite tight specs for ship's main receivers, and some of them relate to filter bandwidths and skirt shapes. The phase noise issues I mentioned play into this because the in a measurement of a receiver's selectivity, the width of the filter skirts is increased by the action of phase noise. Some receivers were locked out of ship's main receiver duty because of the use of cheaper filters or filters specified for other services, or due to excessive phase noise in early (and some not so early!) synthesisers.

So, while the difference may be major and affect the whole architecture, I think you'll find it may be down to which filters are fitted and maybe the temperature grades of quartz crystals. It could be a lot of little details buried in the listings of parts.

David

[Synchrodyne]

Information on the R550 and R551 does seem to be quite scarce. Redifon had a history, going back to the R50 of the late 1940s at least, of using the same basic receiver design for both the general-purpose and merchant marine markets, so that would support the notion of the basic similarity between the R550 and R551, with just detail differences. It was normal for professional HF receivers to exist in several variants, the Eddystone EC958 series being an example where much variant information is readily available.

Over time, the role differentiation between the R550 and R551 seems to have become blurred. For example, in 1971, the R550 was advertised thus:




But by 1973, the R551 was being advertised (along with the R499) for multiple roles, not just merchant marine:




Unfortunately, none of that addresses the question as to just how the R550 differed from the R551.

Originally Posted by Radio Wrangler

The use of a mix-down in the main loop was used for two reasons: Firstly, it reduced the maximum frequency that the programmable divider had to work at. This was valuable in the era before fast logic, and the two-modulus prescaler. Secondly, by reducing the N number of that divider, it reduces the multiplication of phase noise vreated in the phase detector, divider and reference.... so long as the mix-down signal is clean.

Thanks very much for that. I had wondered if mix-down to a suitably lower frequency band was the main reason for doing it that way. Now it all falls into place.

Thus the crystal oscillator for the 2nd frequency change provided a clean signal both for its primary purpose and for mix-down. Since for the error correction loop to work, the 2nd frequency change had to be supradyne, the crystal oscillator output itself was too high in frequency for direct use for the mix-down job, but that issue was solved by premixing it with the mixed output from the 100 kHz step synthesizer and the interpolating VFO.


Cheers,

[Synchrodyne]

This excerpt from the Wireless World 1970 June HF receiver survey gives us a clue:

WW 197006 p.310 Redifon.pdf


It implies that the only difference between the R550 and R551 was in frequency coverage. It was 200 kHz to 30 MHz for the R550, and 60 kHz to 30 MHz for the R551. Given that the basic upconversion scheme did not in and of itself limit the lower end of the coverage range, presumably the R551 included inter alia whatever front end and synthesizer changes were required to keep things “clean”, and to meet merchant marine specifications, down to 60 kHz.


Cheers,

[deliverance]

Thanks again for the info Synchrodyne , Today I have been removing modules to check the boards and have found more leaking capacitors especially 5uf and checking the ones that aren't leaking which so far check okay with low esr . The modules are straightforward to remove but getting access to the boards themselves is tricky so I am just taking my time one module at a time I don't want to make mistakes . The power supply is fine the 5v and 20v lines are now fine and has a new bridge rectifier fitted .

[Radio Wrangler]

Allowing lower frequency tuning could involve several approaches:

It might be a simple change to switches/logic to allow those frequencies to be selected.

DC blocking capacitors and coupling transformers may be different to shift low-frequency roll-offs.

A low frequency converter might be needed (EG Racal RA17 family of Wadley sets which have massive spurii below 1MHz)

Even synthesisers can need help... the Racal RA1772 family switches in a crystal filter in the LO path to notch out any phase noise falling on the IF frequency, and by the limited port isolation of real world mixers, lifting the noise floor as you tune towards zero Hertz. You'll also find this trick in the HP3724A instrument.

So, you might find any combination of these things being done in your Redifon. They aren't common sets on the amateur market. Marconi's marine receivers were the same. I think that leasing arrangements common in the marine market meant that a lot of radio gear eventually returned to the manufacturer and were scrapped. They didn't want used ones on the open market threatening their income stream.

I've never seen a 408, 550 or 551 in the flesh.

David

[G8BBZ]

Hi All,
As previously noted, the R550 was the first in the family to be released as the Commercial version of the receiver, closely followed by the R551 with extended LF frequency coverage as the type approved Marine version. Shortly after the launch of the R551 the contract was received for the Royal Navy 643/CJP station, using the R551N version of the Marine receiver.
The only significant difference between the R550 and the R551 was the Aerial Filters module. With production volumes dominated by the R551N requirement, Production decided it was no longer worth making the R550 version of the Aerial Filter module, so all sets were fitted with the extended coverage version of the Aerial Filter.
At this point the R550 designation was dropped and the standard Commercial version became the R551C. There were various other letter-suffix models, generally resulting from different IF crystal filters being fitted.
The synthesiser design was very much dictated by the available technology of the day.,Decade selection below 100KHz steps was ruled out by phase noise considerations, MECL II logic was used in the front end of the divide by N as this was the only technology that could cope with being clocked at 68MHz at the time.
The follow-on receiver to the R550 family was the R500 (I don't know why the number went backwards) This was a single conversion design, IF 1.4MHz, where the emphasis was on a very strong front end using FET amplifiers and digitally controlled, motor tuned RF circuits. The synthesiser is built around the Motorola HEF 4750/4751 CMOS packages and tunes in 100Hz steps selected from a keypad.
The R1000 family reverted to dual conversion with IF's of 38 and 1.4MHz, sub-octave aerial filters and a multi-loop synthesiser with 100Hz steps and continuously variable tuning. The emphasis with this receiver is on full remote control capability.
cheers

[Radio Wrangler]

Thanks Peter, that explains pretty much everything.

These Redifon receivers do seem to be somewhat rare. I've seen a few of the exciters (GK203N and SD1) but never a receiver. I even had an HF125 power amplifier.... I think I still have a spare set of boards for it in the attic.

MECLII was the fastest logic available at the time. It was very power-hungry, ran hot and was a bit iffy on temperature compensation of logic levels. Decent temperature compensation only really came along with MECL 10k. One MECL II/III part did live on long past the rest of its family, the MC1648. Motorola kept shifting it onto newer and newer processes and the Ft's went ever upwards. The later production of the parts would oscillate at the resonances of PCB tracks rather than the intended tank components. With its limited voltage swing and too-direct coupling tot he tank, in-circuit Q was limited and the phase noise pretty bad. Avoid!

I came on the synthesiser scene at HP in the era when TTL with an ECL 2-modulus prescaler was just becoming the best choice.

David

[Radio Wrangler]

Redifon R500

Yes I remember it coming out. There was a lot made of it in the trade press at the time. It was supposed to be fairly good on performance despite being designed principally for low cost. Underlying the articles was the feeling that this was very much an attempt to save the company.

They made much of the preselector allowing single down conversion to 1.4MHz and saving the need for the hardware of a high IF. The crystal filters in the IF were also arranged serially. (widest first, of course) and a switched tap along them chosen. This meant that the wider filters contributed to the skirt selectivity of the narrower ones, making them a bit cheaper to build.

I wasn't much impressed by their choice of synthesiser.

The user interface meant that it was not a receiver you'd want to search with.

David

[majoconz]

I have a R499 here that is in mainly good condition and would work well if I could find some crystals. Back in the mid-70's they were used in the NZ HF aero circuits for SSB reception and we were setting them up for use at various stations around the South Pacific. Somewhere I have a pic of me instructing an apprentice on the setting-up and checking to specs. Good receiver, bombproof front ends - I have a set of filters too which are often missing. It works quite well with a synthesized local oscillator but needs more oomph!

[Jon_G4MDC]

Weren't HEF4750&1 Philips designs?

[deliverance]

Very interesting more to digest I will be speaking to Peter later on thanks for info .

[Synchrodyne]

Originally Posted by G8BBZ

The follow-on receiver to the R550 family was the R500 (I don't know why the number went backwards) This was a single conversion design, IF 1.4MHz, where the emphasis was on a very strong front end using FET amplifiers and digitally controlled, motor tuned RF circuits. The synthesiser is built around the Motorola HEF 4750/4751 CMOS packages and tunes in 100Hz steps selected from a keypad.

Possibly the R500 was so-named because it was to some extent at least an extension of the R499 single-conversion receiver concept. That is, it was a tunable version of what was originally a fixed-frequency receiver, giving the flexibility to be set to any desired frequency within its range, but not intended for search work. I’d guess that its cost was somewhere between that of the R499 and the general-purpose receivers with manual tuning search capability, such as the R1000. There doesn’t seem to be all that much about the R500 on the web; this item, in “Electronics & Power” journal, came from a quick search:

Electronics & Power 198104 p.277 Racal R500.pdf

The R1000 was described in Wireless World 1978 May:

WW 197805 p.91 Redifon R1000.pdf


Cheers,

[G8BBZ]

Hi,
As you wrote in an earlier post, the R499 was a member of the R550 family. The RF amplifier was essentially the same design, though tuned rather than broadband as in the R550, the IF amplifier was the same design, though a chassis mounted pcb rather than a plug-in module as in the R550.
The R499 can be used with an external LO source simply by feeding the LO signal directly into the mixer - this was a factory option. Handy in these days of sky high crystal prices.
The rumour about the R500 being designed to "save the company" seems a bit odd. If anything, it was the contract for 643/CJP which turned the company's fortunes around, as many thousand of these systems were made for navies of numerous nations. So successful was this product that Rediffusion - Redifon's parent company - decided they wanted to be more closely associated with the success and renamed the company to Rediffusion Radio Systems. The company traded under this name until reverting to Redifon at the later management buyout.
The R500 was definately not an operator's receiver. Keypad entry - and a nasty keypad at that - coupled with a dim LCD display makes for a very poor search receiver. The optional full remote control capability and the availability of a control terminal based on the DEC VT103 LSI-11 minicomputer gives a better idea of the intended application for this family of equipment, which was part of the Navy ICS25 Integrated Communications System and also flew in Nimrod aircraft.

[Radio Wrangler]

Originally Posted by Jon_G4MDC

Weren't HEF4750&1 Philips designs?

Indeed, and I think Mike Underhill was involved at some point.

David