HF RX Cool Wall

[Reelman]

Another suggestion....... FUG10 E10K German receiver.
If you are not familiar then have a look here:

http://www.laud.no/ww2/fug10/

See what I mean?

Peter

[turretslug]

There's an undeniable "engineering confectionary" aspect to the German designs.

I find that the broadly comparable Koln E52 and AR88 (both surely high on the Cool Wall) make for an interesting case study in difference of approach. The E52 has just one type of signal valve throughout (RV12P2000- and arguably compromised in some positions) but laboratory instrument standard of construction based on a backplane and cast sub-module enclosures, with heavy emphasis on modularity and field servicing. The AR88 has just the right valve for each position without compromising but construction is very simple mild steel sheet and hard-wiring- it must have about the simplest fabrications of any professional receiver (with a nod to the CR100), well suited for domestic receiver production lines to take up.

In the end, the AR88 could be churned out in vast numbers, it seems that the obsessive precision of the E52 meant that Telefunken always struggled to make enough. Wouldn't mind one in the collection, though

Perhaps the BC348 is the nearest comparable set to the FuG10- beautifully constructed internally but ugly as sin externally.

[llama]

RA17 was clearly desirable in the '60s, exemplified in my eyes by its use at GB3COV for the queen's inauguration of Coventry Cathedral. Fair to say, if I hadn't been there as a schoolboy helper I'd probably never have heard of it, though!

I'm presuming it was on loan from Racal rather than out of the shack of a local ham.
Graham

[Aub]

Then there's this one..

http://www.google.co.uk/url?sa=t&rct...,d.ZGU&cad=rja

[dave walsh]

Fascinating German set Aub, didn't know that one. High spec like the AR88 but earlier [perhaps... there is some dispute on the 88 origins] and a superior IF frequency. Of course the RA17 was certainly nowhere near the UK surplus market in the sixties Graham, that would have created a sensation!
I think you had a lucky sighting [Act of God] at Coventry, was it part of an Amateur Radio Link?

I'm reminded that I enviously spotted an EA12 at the beginning of the first James Bond film.
I suspect that the E-52 and others might have been John Cleese sets here ie "Don't mention the Receiver" especially given comments in the link re Dutch attitudes!
Dave Walsh

[G4XWDJim]

Originally Posted by Aub

Then there's this one..
http://www.google.co.uk/url?sa=t&rct...,d.ZGU&cad=rja

I had an E52B in the mid 1980s Aub but was persuaded to sell it to someone in Belgium for £60 which seemed a very good price at the time. Wish I'd kept it, they are more than highly desirable these days. All I have now is its photo.

Jim

[turretslug]

Did the Hammarlund SP600 ever feature in significant numbers this side of the pond? It ssems to have been made in large numbers for the US military but, as with the R390A, that's no guarantee that the UK market would have been swimming with them. Whilst the valve-count might seem slightly daunting, the overall architecture could be described as "AR88 successor", the extra valves representing refinement and trimming, rather than the tear-things-up-and-start-again of the RA17. The 30-54 MHz band coverage presumably virtually mandated turret band-changing and anyone who's used a B40 will know the pleasing geared "thunk" of such a set-up.

Not having used one, I can't really criticize, but perhaps a set with low-VHF pretension would have suited a 6AK5 first RF stage (with suitable AGC arrangement) and maybe 6BA7 first mixer after the 6BA6 2nd RF, rather than the 6BA6/6BE6 as designed. As I say, who am I to play Chief Designer to such a respected name, but both these valves were current at the time.

[G6Tanuki]

I'll admit to never having seen a SP600 in-the-flesh so to speak. I'm minded to think that the Racal RA17/RA117 was the UK military's upper-end equivalent radio from the 1950s onwards, and the Eddystone 730/4 (.mod.uk being by far the biggest purchaser thereof) the lower-end option.

[turretslug]

Actually, that seems like a very good comparison- the R390A/RA17 series for the serious monitoring where the precise tuning resolution, stability, and re-setability justified the expense and complexity, the SP600/E730 for more mundane HF capability. The use and maintenance of the latter two would have been thoroughly familiar to both oppoes and techies who were used to the WW2 legacy sets.

[M0SOE_Bruce]

Having owned both an AR88 and the SP600 (still have) it's hard to decide which is better. They both use old school technology with mixer/oscillator and no Wadley loops.

Both are big and heavy. I also still have a SP200 and HQ129, all made by Hammarlund.

One radio that is a sleeper is the CSR5 by Canadian Marconi. Built in WW2 it uses a 9002 (955 acorn) as an oscillator. That valve is good into the VHF range.

Bruce M0SOE

[Synchrodyne]

Originally Posted by turretslug

Not having used one, I can't really criticize, but perhaps a set with low-VHF pretension would have suited a 6AK5 first RF stage (with suitable AGC arrangement) and maybe 6BA7 first mixer after the 6BA6 2nd RF, rather than the 6BA6/6BE6 as designed. As I say, who am I to play Chief Designer to such a respected name, but both these valves were current at the time.

Possibly the benefits at 54 MHz were not great enough to justify using a relatively expensive valve (6AK5) or a relatively scarce valve (6BA7)?

The 6BA6, 6BE6 and 6AU6 were allegedly designed with FM in mind, but this could well have been in respect of the 44 to 50 MHz band. The 6BA7 I think sought improvement over the 6BE6 at 100 MHz, although the underlying problem was that a pentagrid was out-of-place as a VHF mixer.

There were options between the 6BA6 and the 6AK5, namely the 6BJ6 and 6AG5. The 6BJ6 was fairly new and was probably somewhat better than the 6BA6 at VHF, but the 6AG5 was well-established as a military valve with a cost-performance point a bit below that of the esoteric 6AK5.

As you say, the 6AK5, being sharp cutoff (and the 6AG5 too) would have required specific agc arrangements. At least on the face of it, they would require reduced agc magnitude, and arranged so that the 6BA6s in the agc system reached the bottom of their curves before the 6AK5 (6AG5) did. But on the other hand, I think that Eddystone managed to hang both 6AK5 and 6BA6 types on to the same agc line without any proportioning. And Marconi had an HF receivers or two that mixed a Z77 (EF91) 1st RF stage with a W77 (EF92) 2nd RF stage, although right now I do not recall what the agc arrangements were.

Cheers,

[turretslug]

Had it existed on the market in greater numbers, Eddystone's 910 model (apparently aimed at the marine market and heavily developing the basic architecture of the 750) would surely qualify for the Cool Wall with its product detector, smooth main tuning, low-rate incremental tuning and Eddystone mechanical build quality. It featured 6AK5 as both RF amp and second mixer, 6BE6 as first mixer and 6BA6 as 85kHz IF amp, all fed in direct parallel (other than filter resistors) from the AGC line with no potting-down. Possibly, Eddystone arranged for stage gain and cathode/screen feed resistors to give an overall even and effective AGC characteristic but it seems surprising that even the basic AGC diode load wasn't arranged as a favourable potential divider between sharp cutoff and vari-mu types.

[G6Tanuki]

Interesting.... must admit, if I had been tasked with designing a general-purpose HF receiver in the 1950s I'd probably have gone for 2 RF-stages ahead of the mixer.

First stage would have been something like a 6AK5. Deliberately run at relatively-low gain, with the antenna-coupling and interstage-coupling tuned circuits optimised for selectivity rather than transformer-step-up gain. Seriously-delayed AGC applied.

Then with the front-end noise-figure essentially having been established, something like a 6BA6 and then an ECH81. Normal, non-delayed AGC applied to the 6BA6, to the point where it could actually serve as an attenuator.

If you want decent dynamic range, applying AGC to any kind of mixer is an enduring sin which not even the Pope can give you absolution from.

I'm reminded of several US amateur-band receivers in the 1950s that had two AGC loops - one a rather-wideband one which was designed to throttle-back the front-end gain when faced with spectacular local signals within the front-end passband [to pre-empt crossmodulation] and the second being the more-usual AGC that responded to signal-levels that made it down the main IF selectivity chain.

[Synchrodyne]

Originally Posted by turretslug

Had it existed on the market in greater numbers, Eddystone's 910 model (apparently aimed at the marine market and heavily developing the basic architecture of the 750) would surely qualify for the Cool Wall with its product detector, smooth main tuning, low-rate incremental tuning and Eddystone mechanical build quality. It featured 6AK5 as both RF amp and second mixer, 6BE6 as first mixer and 6BA6 as 85kHz IF amp, all fed in direct parallel (other than filter resistors) from the AGC line with no potting-down. Possibly, Eddystone arranged for stage gain and cathode/screen feed resistors to give an overall even and effective AGC characteristic but it seems surprising that even the basic AGC diode load wasn't arranged as a favourable potential divider between sharp cutoff and vari-mu types.

It would be interesting to see the agc curves for this kind of valve mix. Diagrams and graphs such as Mullard provided for valve TV receiver agc systems would be very illuminating, but such seem to be scarce for radio receivers. Eddystone also connected the ECC189, 6AK5, ECH81 and 6BA6 to the same agc line in the 830/7. The ECC189 was a true remote cutoff cascode valve, essentially an ECC89 with altered pinout (see post #270 here: https://www.vintage-radio.net/forum/...=38395&page=14). As a TV valve though it was probably normally used with agc voltages down to around -12 V, not as low as would be likely in an HF receiver with 6BA6s to control.

Talking of Eddystone and the 6AK5, where, if at all, would its 880 series fit in this HF receiver hierarchy? This was its most complex valve receiver, and to some extent it may have competed with the RA17 for government business, such as with the Diplomatic Wireless Service. It used the older Collins-type topology, but then the Wadley Loop was I think unique to Racal until the late 1960s at least. The Collins layout had life in it yet; for example the Marconi HR120 of 1960 was apparently of that form. I have never seen any mention that Eddystone offered a rebroadcast quality ISB adaptor for the 880, whereas Racal did for the RA17. (The Racal rebroadcast combination would I think easily pass the BBC World Service Play of the Week listening test that I mentioned in this thread: https://www.vintage-radio.net/forum/...=107786&page=2, post #34.)

Originally Posted by G6Tanuki

I'm reminded of several US amateur-band receivers in the 1950s that had two AGC loops - one a rather-wideband one which was designed to throttle-back the front-end gain when faced with spectacular local signals within the front-end passband [to pre-empt crossmodulation] and the second being the more-usual AGC that responded to signal-levels that made it down the main IF selectivity chain.

I think that approach became more commonplace in the solid sate era, when it was much easier to provide adequate DC gain for the separate RF agc loop. Even the relatively simple Eddystone EB35 Mk III did it that way, more-or-less. It had zero RF stages, the input stage being a dual-gate mosfet mixer with signal and agc on gate 1 and mixer on gate 2. Wideband RF agc also spread into consumer equipment. Some FM tunerheads had their own RF agc loop, and the National LM1865 FM IC (perhaps the next big step in functionality after the CA3089 and its derivatives) had both wide and narrow RF agc. Later AM receiver ICs, such as the LA1135, also provided wideband RF agc. And in TV receivers, the Plessey SL1431/2 IF pre-SAWF amplifier ICs provided for wideband RF agc.

On the other hand, back in the valve days, even having the RF agc delayed as compared with IF agc was not too common; the Marconi Atalanta comes to mind as an example where it was done, though.

Cheers,

[G6Tanuki]

Originally Posted by Synchrodyne

On the other hand, back in the valve days, even having the RF agc delayed as compared with IF agc was not too common; the Marconi Atalanta comes to mind as an example where it was done, though

I think that delaying ther AGC to the first RF amp was generally done as a way to improve the noise-figure back in the days of relatively-noisy first RF valves: it has the disadvantage that when faced with really strong signals you can still get too much gain ahead of the mixer.

Some radios I know of (Heathkit ones in the 1960s) used a non-variable-mu valve as the first RF-amp - and applied AGC to it! A significant improvement in strong-signal performance could be obtained by either replacing said valve with a variable-mu one, or removing the first RF-amp from the AGC line.

I also recall reading in some 1960s technical journal about receiver-designers who were using varistors and suchlike in the AGC line to tailor the AGC voltages applied to individual valves in various non-linear ways.

[turretslug]

Originally Posted by Synchrodyne

Talking of Eddystone and the 6AK5, where, if at all, would its 880 series fit in this HF receiver hierarchy?

It's courtesy that Radio Wrangler- the thread originator- should hold the rule-card and call the shots here, but on the assumption of similar weighting to the motoring Cool Wall, then the primary consideration should be excellent and highly respected all-round performance in its intended role but with additional factors such as innovation, association, character and style all contributing subsidiary credit or debit points in the overall pecking-order. Thus, the RA17 scores on innovation, the R1155 on association, the B40 on character. Finally, style- well, very nearly absent from the dour and business-like world of the comms receiver but Hallicrafters did try with several of their sets and the BRT400 with its plentiful chroming and impressive, heavily illuminated scale-strips looks like a cathartic shedding of wartime austerity. Also style-wise, the AR88 could at least be described as possessing purposeful gravitas and the 880 fits here too.

So, the 880 was intended to offer top-drawer capabilities in all standard criteria, with particular emphasis on low radiation, high stability and excellent resolution/accuracy and with little regard to cost. Lots of bells and whistles and ticked boxes (though I can hear a faint whisper of "camel" looking over the front-end arrangements). That must leave it jostling keenly at the front.

It was designed and made to exacting mechanical standards by a long established British company with a small workforce in "traditional" premises- surely there's an Aston Martin comparison to be drawn. Credit.

Foreign Office/Diplomatic Service association (not to mention a hint of "spook"). Credit.

Rarity- like the R390A, that may have inflated its esteem, and stopped too many people discovering its flaws. Credit (until the truth comes out).

Based on an architecture that had been around for maybe 15 years and possibly regarded as eclipsed (though Rolls-Royce sold happily on the "tried and tested" feeling for many years)- debatable, but not really "cool". Slight debit.

Famously/notoriously very heavy and notably deep- so not easy to accommodate in the amateur shack. Debit.

It would be interesting to hear from an actual owner/user!

[G0HZU_JMR]

When I was a spotty student I had to study the design/workings of various military radio receivers and transceivers and one of the nicest ones I actually got to use was called a Marconi Nebula.

At the time I'd only ever used my own homebrew SW receivers for cw/ssb and to me this receiver was awesome to use in comparison. It's not in the same class as some of the radios already mentioned but I sneaked into the labs on many a lunch break to tune around with this radio because there were various HF antenna feeds available there.

I had to google it to make sure of the name and apparently it was a rebadged Eddystone receiver. definitely a cool receiver for me (back in the 1980s)

[turretslug]

Originally Posted by G6Tanuki

If you want decent dynamic range, applying AGC to any kind of mixer is an enduring sin which not even the Pope can give you absolution from.

One of my criteria for qualifying as a "proper" communications receiver has long been that mixers should be left to get on with being mixers, under optimum operating conditions, without being press-ganged into the gain-control requirement part of the spec. Needing to do so implies not spending money on enough IF stages- the lower consequent gain per stage means less likelihood of response shift with AGC and more tuned circuits for improved (particularly skirt) selectivity to boot. No need to go to the extreme of the original R390's six 455kHz stages!

[Synchrodyne]

It is interesting though that the commonly used mixer valves, the 6BE6 and ECH81, were both of the remote cutoff type, even if often used in HF receivers without agc. One imagines that this was determined by domestic receiver needs, whose production numbers were such that they completely swamped non-domestic products.

Eddystone seems to have been guilty of applying agc to mixer stages, for example in the 940. In the 830/7 and EA12, agc was applied to both mixers in each case. The 830/7 had a 6AK5 (sharp cutoff) 1st mixer as well, connected to the same agc line as the remote cutoff valves – breaking all of the rules...But the EA12 seems to have been well-regarded, so maybe it is deserving of Tannhȁuser-like absolution.

One wonders, though, whether in these cases the application of agc to the mixers was done primarily to help achieve the desired agc range, or to protect the mixer itself from overload (in exchange for which the other ills of gain-controlled mixers were tolerated). In a general sense I think that the cross-modulation level vs. gain reduction curves for agc’d valved head upwards at first, then drop down towards, but not quite as low as the starting level at mid-gain reductions, then climb again as gain is further reduced. So as long as input signal levels were controlled so as not to cross the dip in the curve, the use of agc would seem to improve large signal handling capability. This might have been more beneficial where there was only one RF stage, as in the 830/7 and EA12, and so a smaller RF agc range.

Paradoxically in view of the conventional wisdom, the release of a television frequency changer with a remote cutoff pentode mixer section (Mazda 30C17) was heralded as an advance that would allow improvements to the agc and cross-modulation performance of the whole receiver.

Cheers,

[Synchrodyne]

Originally Posted by turretslug

So, the 880 was intended to offer top-drawer capabilities in all standard criteria, with particular emphasis on low radiation, high stability and excellent resolution/accuracy and with little regard to cost. Lots of bells and whistles and ticked boxes (though I can hear a faint whisper of "camel" looking over the front-end arrangements.

One hump or two? Or maybe a camel train?

I suppose it can be seen as a two-band, single-conversion receiver with one “RF” stage, and thus three-gang (including the VFO), preceded by a crystal-controlled converter that had two RF stages, and thus was three-gang (no VFO to be counted). But the two were ganged together so that the RF tuning of the “converter” tracked the tuning of the “single-conversion receiver”. And even then there were image-rejection filters to help lift this parameter.

The “converter” RF section was originally conventional, with a pair of agc’d 6BA6. At the 880/2 iteration, the 1st RF stage was changed to an ECC189 cascode. I don’t know whether Eddystone was the first to do this, but it may well count as an innovation. The 1st IF stage (tuned), effectively the “RF” stage of the “single-conversion receiver” was another agc’d 6BA6, so conventional. But the 1st and 2nd mixers were both 6AK5 with oscillator injection at the cathode. Low noise, no doubt, but did it matter that much given what was ahead of them? Or was the mixer choice related to the need to really minimize oscillator radiation? Did the fact that the cathode was a relatively low impedance injection point have something to do with this?

Short of a Wadley Loop, I wonder how else might Eddystone have configured a receiver to meet the stability and frequency readout requirements. I am not sure if the pre-mixer technique (as used by Drake in its R4-series) was extant back then. And even so, it would seem to have the problem that pre-mixing itself created unwanted products (mirror images) that needed to be rendered harmless both inside and outside of the receiver. Did the 880 have a lower level of oscillator radiation than the RA17? That might have been a plus (for the DWS and the like) that justified the use of an older topology/architecture.

Cheers,