Eddystone 750 wreck

[HamishBoxer]

All comments noted, thank you.

[turretslug]

Are you certain that the circled component is a brown Hunts? Some (rather basic quality) carbon comp resistors have rounded bodies like the rat-dropping Hunts, some small Hunts have squarer ends like carbon comps.... It's not just the current era that gave us component confusion! From its position and if it is a resistor, it could (emphasis) be the EF91 LO's anode load.

The capacitors mentioned in connection with removing the coils are the two 10nF AGC decouplers associated with RF and mixer control grids, in my set they were installed on the Band 1 (12-32MHz) antenna and mixer grid coils and are the ones at top rightmost and middle rightmost in your post #20 picture. Do check though, as there was a bit of shuffling of component positions over the run. Only the two individual coils needed to be lifted out, helped by the Eddystone habit of using solder as glue rather than having a twisted-up mechanical joint, ISTR they were retained by a 4BA screw going into the threaded brass signal common busbar on the coilbox topside- at least no nuts to make you swear on re-assembly.

[ms660]

Originally Posted by PJL

I have ringed one in the attached.

We used to call them rat turds, wouldn't trust them at all, from memory I recall that Murphy used them in some of their TV's.

Lawrence.

[cathoderay57]

Originally Posted by PJL

I have the service manual but is there a component layout drawing?
I am still unsure of the capacitors to replace before powering up as the ones around V1-V3 are difficult to see, I have ringed one in the attached. Looking at the circuit, the caps around V3 are C27 and C28 heater decoupling and listed as 0.0005uF moulded mica.I would also appreciate advice on which coils to remove, do I take out a whole bank?

I did not find a component layout drawing anywhere. It was necessary to trace physical connections and correlate with the circuit diagram. I have tried to help you by giving a drawing identifying all the metal-cased 0.1uF caps. See the attachment to Post#8. You should find it possible to unbolt these where necessary to disconnect the wire and get a probe in to carry out a leakage test. I just used an Avo on the MegOhm range and replaced the ones showing leakage at low voltage. That doesn't guarantee a serviceable component but it worked OK to get the set going. This shouldn't require any coils to be removed. You will see smaller 0.01uF metal cased caps dotted about the chassis. You could test these in the same way if you want but I ended up replacing all of them. Ideally replace the rat turd caps as well but you would have to remove at least one coil to do that. For example, the one you have ringed would require removal of the adjacent coil. You do not have to remove the whole coil bank. Maybe do the easier caps first then power up and see how it goes? Cheers, Jerry

[G6Tanuki]

Those 'rat-droppings' are really a pain-in-the-sphincter: when I rebuilt my bargain-basement Eddystone 840A I overlooked one such capacitor which was part of the detector/AGC-decoupling circuit and buried somewhere deep; it took me ages to work out why the AGC didn't work properly and the RF-Gain control's action varied day-by-day.

The bolt-down decouplers can be similarly problematic, both from the leakage perspective and the issue of how to replicate the physically-robust top connection-thats-used-as-a-tagstrip when replacing them with modern capacitors. I've used short lengths of brass studding and a pair of nuts for the bottom connection and a short bit of Nylon pipe Araldited over the capacitor itself.

[turretslug]

The nice thing about the 2BA stud fixing capacitors (there had to be something) is that once the nut is undone, a sliver of cereal packet etc. can be shoved between it and metalwork for a leakage test. Apart from, that is, where the top tag is angled to directly bear on a terminal- in both of these cases in my set, the joint had broken away with time. Neat idea Eddystone, but not so smart in retrospect.

[PJL]

I lifted the IF cover beneath V4 and found 3x brown Hunts capacitors hidden away (C41, C42, C43 all 0.01uF). Access will require desoldering the can whose pins have been bent over to connect directly to the adjacent IF can. They are only LT and cathode decoupling but...it's not exactly built with servicing in mind and I am beginning to think the optimal solution is to strip the chassis down and replace all suspect caps once and for all.

I would normally check for leakage of decouplers as part of electrolytic reforming by connecting an HT supply to the reservoir capacitor through a gradually reducing series resistor to maintain a low current. Once stable I measure the voltage across the HT feed resistors to work out the leakage current. In this case there are a couple of resistors I need to lift first, R51 and R23, and all I need to do is find the ***** things.

[turretslug]

R51- big red BTS job under the forward end of the selectivity shaft and parallel with it. R23, likely under the tagstrip half-way along the PSU cap side of the RF box. There's also R47, R49 and R55 (all 100k) also forming HT leakage paths unfortunately. At least with all those dividers in circuit the HT in this set leaks away nice and quickly after switch-off!

[PJL]

Well, it doesn't get better...After considerable effort I have removed one capacitor (they all have wire run through the tag holes and folded over) and tested it. It measures 0.2uF and shows 138uA leakage at 350V (about 2mW), ESR 9 ohms which is better than the replacements. This leaves a dilemma as most could be left if the same but of course they won't be. This one is C79, AGC, so would need to be replaced in any case.

I have also noticed there are some high value 'Mica Moulds' in there too. On a BC348 I have had these spitting boiling wax at me and measuring a short.

The access for servicing is terrible and I am not at all keen on revisiting this so I am coming to the conclusion there is no shortcut and it would be best to replace all the 0.01uF and above capacitors in one shot along with any out of spec resistors.

[cathoderay57]

Well, if you're feeling bold you could try applying power, see if it makes a noise, and doing voltage checks to point to the worst caps. Cheers, Jerry

[PJL]

Firing it up and replacing parts to fix issues as they occur is likely to take far longer in the long run as access to some parts requires stripping out other good or already replaced components. I have checked the major components so it should work.

I am going to strip out the old 0.1uF and make up some brass brackets for the replacements. The 0.01uF seem a real mix of different types, Hunts, TCC metal can, Mica Moulds and rat turds but they will all go. Without leakage, I can then check resistor values in circuit.

If I feel really keen, I will strip the mains transformer, BFO which needs to be opened up anyway, and choke off the power chassis and give it a rub down and coat of grey primer.

[turretslug]

I've a lot of respect for the folk who originally built these close-packed sets, probably on piece-work with less effective lighting than we'd expect nowadays, clumsier and less sophisticated soldering irons and smoky, catch-your-throat fluxes. Eddystones are famous for the neatly formed component leads and that itself must have taken a deal of practice. I suspect that many of the sets that shared the 750's construction style had the left and right side-chassis built up as a near-complete open-L module that was then screwed and wired to the chunky central RF box, but they are still pretty close-packed and the RF boxes are works of tight-knit art.

It all makes for a set that's somewhat daunting to do extensive work on but IMHO it's a worthwhile classic, about half the volume and weight of an AR88, and not a bad looker to boot. I'm sure that the neat and precise mechanical detailing is as worthy a representation of the former Birmingham-area light engineering tradition as much as the more widely-famous names such as BSA, Raleigh etc.

It might be as well to treat it as a succession of overhaul tasks, rather than get bogged down in a marathon all-in-one effort. One time, I removed the 2nd conversion module for a complete R and decoupler C replacement (I didn't want to take it out again!), similarly the BFO module and so on. There are still a few stud-mount 100nFs in place where they're uncritical, unstressed cathode decouplers but the RF amp and first mixer deserve short-lead, low inductance modern ceramics to let them do their best really. Those stud-mount caps doubled as convenient terminal posts originally which asks for a bit of creativity in substitution.

As well as the mains transformer and output transformer, might be worth continuity-checking things like IFTs just so as not to be throwing too much time, money and effort into a losing cause. Good luck with it, I'm sure the others here will agree that it's a pretty good set.

[PJL]

Taken your advice and the IFT's and 2nd oscillator all check out OK, the BFO coil is inaccessible without removing the module.

The temptation is to replace the AGC and audio coupling capacitors and give it a whirl but with the many capacitors with poor leakage/reliability reputation it wouldn't be worth fault diagnosing any issues so I will go for the full overhaul upfront. It will also be much easier to check resistor values and extract the 2nd oscillator and BFO modules.

I have made a prototype stud mount replacement from some thick brass sheet and just need to make 11 more!

[turretslug]

I like it! A similar approach would be good for the multi-unit "bathtubs" found in old US gear in particular where they get used as anchor-points for wires and other components. Perfectly functional if not as authentic-looking as re-stuffing but I'm wary of handling the insides of old capacitors from the era when asbestos, leaded petrol and PCB oils were touted as wonders of the age and schoolkids were handed gobs of mercury to play with in science lessons....

When overhauling the 2nd conversion module, modern capacitors were small enough that I was able to accomodate C51, 100nF V4 hexode screens bypass and C55, V4 triode HT feed decoupler inside the module and earthed to the same point on V4 brass support limb as C41, 42, 43- this struck me as more prudent from the RF circulating current point of view than their original relatively far away placement on the IF/AF chassis. It'll cut down the number of stud cap replacements needed to 9 too!

[PJL]

I am slowly working my way through the IF unit and have extracted the 2nd oscillator unit and so far the resistors seem to be OK.

My question relates to R6 (12 ohms) located in the oscillator unit on the mixer grid. I am wondering if this is exploiting the inductance of the resistor as 12 ohms would seem to have little impact on the circuit?

[turretslug]

I assume that it's there as a parasitic stopper, there's a similar one in series with the RF amp's grid- quite a few other Eddystones have these 12 ohm resistors in similar circuit positions. A good question as to whether they were actually found to be necessary, or whether the designer included them as "good practice" and a good deal could be had on 12 ohm ones at the time. When I see multi-band front ends with lots of switchery, interconnections and devices capable of activity into the VHF region, I often think that it's a stroke of luck that the lot doesn't take off! Maybe the designer had been bitten in the past and it stuck.

[turretslug]

Just for the record, one of the 10nF AGC decoupling capacitors mentioned in posts #3 and 5, replaced by one of the "usual" Vishay 1813 yellows. Someone more dextrous than me with a slim but powerful iron might have been able to leave the coil in question in place, but removing it also gives better (but still by no means easy!) access to the other components around the bandswitch wafer/valve socket. The screening cover in my early poduction set has something of a cottage industry look to it, a reminder that Eddystone was never a huge concern, making it impressive that they made such a range of receivers and other products for so long.

[G6Tanuki]

Be aware that the orange trimmer-capacitor in one of your photos is a potentail failure-point: they depend on a silver-plating on a ceramic substrate being linked to the spindle of the adjustment, and half a century of ageing really compromises this continuity!

Trust me - I rebuilt a WWII-era 'RF26" unit that used similar trimmers and could never get its alignment to stay consistent for more than a few days: in the end I ripped out the trimmers amd replaced them with Philips 'Beehives'. Not original I know, but far more reliable.

[turretslug]

It adds to my suspicion that a fair few surplus components found their way into Eddystone's stock-rooms that this set (and the similar date-code example overhauled in one of Gerry O'Hara's 750 articles) features a couple of the "orange trimmers", the one shown is the 12-32MHz band oscillator trimmer, the other is part of the three-point tracking used on the MW range. I wondered if they'd used one of these on that highest HF band to give a degree of tempco compensation, or if it just happened to have a more suitable swing there than the familiar air-spaced ones elsewhere.

I'm wary of that style of ceramic trimmer generally as it seems that even the tiniest bit of grit or corrosion or just neglect-induced stickiness can result in the rotating "saucer" fracturing. Which at least solves the problem of whether to replace it or not....

[PJL]

Unsurprisingly, as it was kicking around at a carboot without a base plate, one or two of the trimmers have bent vanes but that will wait till I have the set running.

I have now cleaned and recapped the IF side and replaced about 4 resistors, one because it would not give a stable reading. The ally plate has arrived so I need to fit that then I will strip off the front panel, dismantle the PSU and BFO, give the steel chassis parts a clean and spray with grey primer, and recap. The rust on the tuning capacitor cover plate cleaned off easily and it is already re-sprayed. I checked the leakage of a Dubiiier 'moulded mica' 0.01uF and it was only about 1uA at 350V unlike my experiences with the MicaMold versions, but it was replaced anyway as it serves as the audio coupling capacitor.

Next major hurdle is taking the front panel apart as the pointer is jammed solid and it needs the dial cord replaced with the pike wire I bought.