An Eddystone 840A

[turretslug]

Re. G4XWDJim and Skywave's points about ergonomics, I wonder if this came about as a result of using the same front panel casting for a wide range of models, from straightforward 840A to quite complex 888 (and "A") and 910 models? I think the linear dial casting first appeared with either 750 or 680X models (perhaps 750?) and a given layout and number of switch holes was allowed for. The number and types of switches/actuators on panel bottom left and bottom right could be varied slightly between models but the main tuning/bandswitch controls and the two potentiometer/small variable capacitor positions at top right and top left were rather more hemmed in by coil-box layout and the narrow chassis each side of it. The 888/A and 910, in particular, are quite precise and ingenious examples of squeezing a bit more in. Al is in a better position to judge, actually having an 888A, but I think that squeezing in the small "fine tune" control between bandswitch and tuning control was quite tight anyway, so the antenna trim (something that wouldn't have gone amiss on the progenitor 750...) ended up being the designer's choice as Cinderella.

A similar scenario looks to have applied in the "later" front panel casting, with the "busy" 830 resorting to coaxial-shaft controls (as did the "early panel" 910). Still a shame that a matching array have knobs couldn't have been used, though.

[G6Tanuki]

The "odd knob" seems to be an Eddystone thing. In the case of the 840A the odd-knob-out is the noise-limiter [which is a chromed lever operating a 2-position rotary switch] - whereas any degree of aesthetics or ergonomic logic would have suggested using a toggle-switch for this function, like the BFO on/off and the "Standby" switches.

Part of me suspects the NL function was grafted in as an "added value" thing rather than having been designed in from the beginning. It uses a semiconductor diode.

[Synchrodyne]

Possibly the use of a semiconductor diode for the noise limiter stemmed from the AC/DC nature of the Eddystone 840 and 840A. The AC-only 740, which was the primary antecedent of the 840, used an EB41 for both the noise limiter and external S-meter drive functions. A UB41 in the 840 would have required disposition of another 19 V in the heater chain, which would have become asymmetric as a result. Not only that, but the facility to drive an external S-meter was a non-starter with an AC/DC receiver. So under the circumstances, a single crystal diode probably looked to be a more attractive solution. Curiously, the 740 required the EB41 despite having two unused EAF42 diodes (RF and BFO). But with “tied” cathodes, the latter were unusable for the NL and S-meter functions.

The 740 disappeared from the Eddystone range quite early on, not even making it to an “A”-suffix revision as far as I know. One wonders whether the majority market was for an AC/DC version, or whether it was just easier (and less costly) for Eddystone to produce – for its agents to sell - just one universal model in this category. (Although some of the “tea-planter export offerings from the setmakers were available in both AC-only and AC/DC versions.)

Cheers,

[G6Tanuki]

Yes I can see the logic of using a semiconductor for the NL: it also removes the possible issue of hum being coupled into the audio chain via heater/cathode capacitance if a valve NL-diode happened to find itself at the "red" end of an AC-powered heater string.

And as you say, it removes the need for an extra valve which would have complicated the heater-chain somewhat.

[G6Tanuki]

I've been doing some more work on the 840A.

As mentioned, it's been "got-at" in the past. I've replaced the orange lawnmower-wire power-lead with a proper 3-core cable (I had loads of old ex-computer 'kettle leads' lying around). Rather than fitting a plug&socket I opted to wire the lead in permanently, running it through an IP67-rated screw-down plastic gland assembly which conveniently fitted one of the holes where the original connector had been. A cable-tie on the inside of the gland provides additional strain-relief.
I've taken the mains-earth to the non-floating case/chassis, and also made sure that the live/neutral wiring is correctly polarised so the floating chassis isn't live.

I'm still using the botched BY127 diode and 'totem-pole' arrangement of dropper-resistors that a previous owner had fitted in place of the original valve rectifier and smoothing-choke. I have a new UY41 on order, and plan to use a small mains-transformer with the secondary winding removed as a replacement choke.

The "been got-at" extends to the audio-stages: 'that capacitor' had been replaced by a 0.68uF plastic-cased component rated at a mere 63-volts. No wonder it hummed. Just to prove a point I hooked up a temporary capacitor and the hum abated somewhat. I have a new-old-stock 0.068uF 400V-rated 'mustard' capacitor waiting to go in when I can find some suitable rubber sleeves for the leads.

Then - RF and AF gain controls to 'max', BFO on, noise-limiter off, a 15-foot length of wire poked in the antenna-socket, power 'on' - and wait... and wait... and wait... what's this? Hey! it receives!! UK and Italian amateurs working each-other with SSB on 7MHz. OK, there's a significant amount of mains-hum frequency-wobble that makes things sound distinctly rough (half-wave rectification and botched smoothing-system are definitely not helping here) but still, it's progress!

The case-stripping with NaOH paste seems to be going well too: I've done almost all the outside, including the perforated-metal ventilation panels. By the weekend the inside should be stripped too - then a quick mist with Zinc Chromate primer.

Question: what colour to paint the case though? I'd not be able to easily source paint to match the original and an 'almost-match' with the diecast front panel etc would look silly. I'm thinking that black crackle would look good?

[HamishBoxer]

Black crackle would look just great.As many Eddystones are of course.

[G6Tanuki]

And I 'just happen to have' a couple of black-crackle spray cans in the shed.

Though it's not perfect, I'll be leaving the front diecast assembly the original silvery hammer-finish if only because the prospect of removing it to repaint (and then having a battle of wits to ever get the tuning-drive and dial-pointer to ever work properly again) would take up time far better spent enjoying a couple of pints of Golden Champion.

[G6Tanuki]

I don't usually spend much effort on cosmetics, but a blast of "Mr.Muscle" glass-and-tile cleaner has brought the front of the 840A up rather well. I now need to get some cotton-buds or similar to clean the inner-surface of the tuning-scale glass.

[G6Tanuki]

Further on rebuilding my 840A - the original power-supply choke had been removed and replaced by a gruesome 'totem-pole' arrangement of green dropper-resistors attached to the chassis by a coach-bolt.

I've removed this and replaced it with a proper iron-and-copper choke - fitted under the chassis.

This, along with properly rewiring the UY41 rectifier and fitting a healthy UL41 - seems to have improved matters somewhat.

The next task is to get the waxy-gunge out of all the trimmers in the coil-box and then do a full alignment: right now it's only sensibly-sensitive on band-2 [3.7 - 10MHz].

Outside, the case is finally yielfing-up its paint following multiple treatments with a Sodium Hydroxide paste.

What colour should I repaint it? I'm *not* going to strip and repaint the front-panel, even though it's a bit scabby in places.

[Bazz4CQJ]

Silvery-greys look right, and if you go in to Halfords, there is a huge range of them to choose from among current car colours.

B

[G6Tanuki]

I was thinking either silver-grey, or crackle-black. The earlier louvre-case S640 and 840 had a crackle-black case all over. Though I'm not sure how well crackle-finish would work on the punched-metal ventilation grilles and speaker-grille I've got.

[G6Tanuki]

Another update: I've managed to get said 840A working reasonably well on the first three bands - but it's rejecting all attempts to coax it to work on the highest-frequency band.

Evidence is that the local-oscillator doesn't. The UCH42 looks like it's the one fitted when the radio was manufactured [Electrolytics dated December 66]

The oscillator-coil's windings are good and the wavechange switch is playing sensibly; voltages on the valve electrodes on other bands are sensible.

I've got a NOS Mullard UCH42 on order (only £6).

[VO1TMM Terry]

I have an old 840C, came off a Canadian icebreaker I believe, which may or may not have had DC ship voltage! Does not have that fifth switch position. Not working, it has a sort of isolated chassis cos if the ship was grounded positive the chassis would have been 'hot' negative. I'll be operating AC mains which here is 115-120 volts at 60Hz. Probably with an external isolating trans.

VO1TMM

[G6Tanuki]

It's raining outside so I decided to investigate the lack-of-reception-on-band-1 a bit deeper. I knew that on this band the L.O. wasn't doing any O'ing and I'd planned to buy a new UCH42 frequency-changer because I suspected the triode part had gone a bit down - but before springing £6 to Langrex it seemed wise to do some more checks.

Performing an "AVOscopy" in the L.O. part of the coil-box and comparing resistances of windings across the four coils, band 1 was a lot different to the other bands so I decided to pull the coil out and have a closer look.

A wise move: with the coil out of the box the joint between the metal tag and one end of the feedback winding just didn't look right - I can only assume that they'd had Ray Charles working on the soldering line the day my 840A was built.

Cleaning things up and resoldering, then reassembling - I now have a working band 1 !! It's happily pulling in Russian, German, South American and US stations on 28MHz, though admittedly not quite in the right place on the dial so a bit of alignment's clearly called for.

Geting it working on this band also shows up the presence of some significant issues on the HT line - SSB stations sound like they're gargling, so there's obviously hum getting into the L.O. Operating the RF Gain control also pulls the L.O. sufficiently to shift a SSB station right out of the receiver passband on 28MHz.

Still, being able to hear anything on 28MHz is progress!

Maybe I'll investigate putting in a stabilised HT line?