Cossor 500 Short wave weak

[kalee20]

Hi, can anyone help me with this? I have a Cossor 500 'Melody Maker', which I've spent a fair bit of time on (rewinding the mains transformer, refurbishing the cabinet, replacing almost all the fixed capacitors AND resistors!) but I'm not having much luck with short wave.

Basically, the oscillator only runs towards the high frequency end of the scale. Below 14MHz the amplitude drops like a stone, and at around 12MHz it just stops. When running, I can bring the frequency to the correct value with the trimmer capacitor and the coil slug.

The obvious culprit is the 7S7 frequency changer valve, but I have tried an ECH81 using tinned copper wire leads from a B9A valveholder pluged into the B8G socket on the chassis - very little difference - and also a new 7S7, which is no better! Bridging the oscillator anode and grid feed capacitors doesn't help, although if these had lost capacitance I would have expected trouble on the long wave band, which is fine. And the wave change switch has been well checked, cleaned, relevant contacts externally shorted on SW to see if it made any difference.

The only clue is the feed-back coil measures 30 ohms resistance, whereas the Cossor service sheet from Paul's CD, specifies 2.6 ohms. But, I've seen a few instances where this sheet has been out of sync with my own chassis. Further, as the feed-back winding is in series with the MW/LW oscillator coil, a problem here would surely have some effect on the other wavebands.

Has anyone come across this problem? Or at the very least, can anyone confirm whether the SW feedback winding is wound with resistance wire?

Many thanks!

[Steve_P]

C9 might be worth a look. And, clean the wavechange switch thoroughly and align it as to the sheet you have.

Also check the Valveholder. They are prone to all sorts of problems...

Cheers,

Steve P

[kalee20]

Gosh that was quick! Thanks very much.

C9 I've already mentally eliminated after bridging with a modern polystyrene 100pF capacitor. But, if all else fails, I'll replace it. (it has a rather inelegant extended lead, the joint hidden inside sleeving).

The valveholder is a phenolic type, McMurdo, but I'll give it the iso propyl alcohol and toothbrush treatment. (Its already been swept with a paintbrush backed up with vacuum cleaner).

Thanks so far...

[Steve_P]

Don't bridge it - Just swap it out. Easier that way... and if it's got any form of Leakage, that'll still be present with another cap across it!

Cheers,

Steve P

[Sideband]

There was a thread a few weeks ago on a similar set. I think the SW osc coil is wound with resistance wire.

See here:https://www.vintage-radio.net/forum/...ad.php?t=18457


Rich.

[ukcol]

Originally Posted by kalee20

..And the wave change switch has been well checked, cleaned, relevant contacts externally shorted on SW to see if it made any difference.

It may be worth double checking the above.

As you will see from the service sheet, the bottom end of the primary (L8) of the SW osc coil is connected to chassis via a contact on the w/c switch. This same contact short circuits the primary (L11) of the MW/LW coil.

Try temporarily soldering a short piece of wire from the bottom end of L8 to chassis.

This will prove (or otherwise) that that part of the w/c switching is working correctly.

[kalee20]

Hi, thanks everybody. Will be attacking the radio Wednesday!

Ukcol - that makes perfect sense, to short the bottom end of L8 to chassis, but I've already tried it - no good.

Richard Newman - thanks for alerting me to that thread on the Cossor 494, it included the valuable info that the SW feedbck coil typically measures 29.5 ohms which is what I am seeing, (but quite different to the Cossor service sheet). I can understand well why resistance wire may have been used, it would be to clobber any tendency for the feedback winding to resonate with stray capacitance, and mess up the oscillation frequency.

The plan is to clean the valveholder, replace the last 2 capacitors in the circuit (C10 oscillator anode and C9 grid), and see what happens.

I Will post any results - favourable or otherwise!

[Steve_P]

Check the tuning cap itself. Do any of the plates snag up anywhere. One short and no reception.

Put it fully open, use the Compressed Air to blow out all the muck, and look as you close it. If it goes off at one place, or gives any loud crackles, look very closely...

Cheers,

Steve P

[kalee20]

Well, I have cleaned the valveholder, both sides, with RS computer solvent and toothbrush, blown dry with compressed air (see photos). I've swept between the variable condenser with a paint brush dampened with solvent, and blown dry. I've replaced C9 (the oscillator grid capacitor), which was 50pF (replaced with 100pF in accordance with the service sheet value). I powered up, hoping that the extra capacitance did not cause squegging.

Almost no difference! MW and LW still work fine - I think this puts the tuning capacitor above suspicion anyway. SW still gives up at lower frequencies, although it does seem to oscillate now down to 11MHz on the dial - as this is approached, the amplitude diminishes gracefully and reaches zero at this setting. (I am checking amplitude using a 'scope connected to the oscillator grid using a x10 probe, but if I disconnect the probe, hold it near the coil, and turn up the scope sensitivity I can tell that the behaviour is almost exactly the same).

I tried again shorting the 'earthy' end of the SW coil feedback winding direct to the chassis, using a 3" lead, to bypass the wavechange switch and associated wiring. All this does is increase the amplitude of oscillation a bit. It doesn't make any difference to the non-functioning at lower frequencies.

I'm really wondering if there is a shorted turn in the coil, although if there is, I'd expect it to kill the SW oscillator completely! Does anyone have any experience of this? Or anything else to try? (My next effort will be C10, the anode coupling capacitor, but I've run out of 100pF capacitors till I get a few more). After that, I'll be running out of ideas!

[yestertech]

Just a long-shot - Have you checked that the heater volts are OK at the holder pins ? (Long story involving an EKCO... best avoided ! ) Also, I assume that the HT volts are in spec. ( as this is not specifically referred to ? ) And I know it's obvious, but could any of the replaced components around the 7S7 be incorrect value or miswired and have any of their relative positions changed ?
Andy

[PJL]

A little odd it works at the HF end and LW - how about leakage in the waveband switch? best way to check this would be to disconnect the SW coil and grid capacitor from the switch and connect direct.

You could also try a bit of extra HF decoupling on the HT - say a 0.1uF
...Peter

[ukcol]

There was yet another thread about the SW osc coil in the Cossor 500 that you may or may not have seen.

https://www.vintage-radio.net/forum/...read.php?t=686

The thread talks of it being common for the SW osc coil to go O/C resulting in a lose of all wavebands.

Perhaps before it fails completely, due to (say) corrosion, it may become inefficient giving the symptoms Kalee20 describes.

[kalee20]

Hello everybody, well, it's not the LT or HT, I'm measuring 210V on the frequency changer hexode anode (the servuice sheet says 208V) and as this is connected by a low resistance DC path (the IFT primary) to HT+ then I'm happy. The LT is also correct at 6.4V, and although I couldn't probe the valve pins themselves directly (to eliminate poor valveholder contact), the cathodes look heathily alight. I used a Variac to apply an overvoltage for a few few moments and even at 244V DC the oscillator still gives up (it does run down to 9.5MHz however at the higher voltage).

Setsappeal - the symptoms were present before I started wholesale replacing of components, so miswiring isn't a possibility. However, checking through after your suggestion, I find the capacitor I changed on 1st August is the RF coupling to the hexode grid, not the oscillator capacitor to the triode grid, so it's back to work with the soldering iron! (and of course, I've put the wrong value in here. Doh!)

PJL - I'm hoping I've not got a switch problem, it's had the air line, the paint brush, the switch cleaner, the switch lubricant, but I'll disconnect the switch to totally discount it. Thanks! I tried extra decoupling but no effect (it actually slightly diminished the amplitude of oscillation at the HF end).

Ukcol - thanks again for highlighting this thread, yes I had read it, I suppose it's possible the thin-wire feedback winding has a shorted turn due to enamel degradation. I'm hoping not - if it's resistance wire (as it almost certainly is), it could be a pain to source.

Plan is to replace the grid capacitor (the right one this time), then isolate the switch! But that's got to wait till Wednesday evening now!

Thanks everybody for help and advice.

[kalee20]

Well, I have isolated the wavechange switch; the SW oscillator coil (tuned and feedback windings) are soldered through short leads to chassis, the 7S7 oscillator triode is connected to the coil through new capacitors (100pF both anode and grid), the variable capacitor is connected to the coil directly through a short lead soldered at both ends. I've even swapped the sections so I'm using what was the RF tuning section, just in case there was a fault in the other section. Oh yes, and Im using my new 7S7.

Guess what? It works down to 9.5MHz (marginally better than before), then gives up. As before, amplitude drops drastically as the frequency is reduced.

I'm really wondering if the unit ever worked. Could it have escaped from the factory as a non-runner? Have there been any instances of this?

[Steve_P]

Oh heck! I'm short of ideas here. Rewind the coil L7 and then do the IF/RF alignment as per the book/trader sheet.

Then if it doesn't work, come back....

Cheers,

Steve P

[MikeBeith]

What are you using to check the amplitude of the oscillator output? From looking at the circuit and the few components in the SW osc. path it points to some strange happening within the coil itself. Another thing I would be inclined to try would be an SW osc coil (almost any with seperate pri/sec eg denco etc etc) temporarily wired in place of the existing one. I would expect amplitude to be higher at the LF end of a SW osc. Are you able to measure the inductance of your osc. coil?

Mike

[kalee20]

Thanks for replies folks!

Mike - I'm checking the amplitude using a 'scope probe held close to the coil, and turning up the scope sensitivity. Not making a direct connection, so as not to add extra loading to the circuit. (However, if I do clip the probe to the 'hot' end of the grid coil, I get the same result.)

When running, the oscillator is easily settable at the at the correct frequency, using either the trimmer capacitor or the core slug position. So I've not bothered to measure inductance. What you say about the coil is what I had feared!

Steve - that's what I think I'm going to have to do, rewind the coil. The grid winding looks OK, especially as it's fairly chunky wire with spaced turns, not much to go wrong here. It's the feedback winding that I am worried about, as it looks like it's wound with random overlapping turns, thus shorted turns is a real possibility. And, with a resistance of 27 ohms, it's very likely wound with resistance wire (as I mentioned in an earlier post in this thread). But having exhausted most of the other possibilities people have kindly suggested, it's got to be the next suspect!

Only thing that bothers me is whether I can find enamelled resistance wire. I don't want to use copper with a series resistor, because the idea of using resistance wire is to distribute the resistance along the coil itself, and curb any tendency to resonate with self-capacitance.

When I've done this, I'll post up the result! Meanwhile, thanks to both of you.

[Station X]

I have some enamelled resistance wire. 24.37 ohms per metre. Any good?

The Scientific Wire Company (google for their website) have all sorts of resistance wire.

[Station X]

Forgot to say. The bare wire is 0.16 mm dia (6 thou approx) with the enamel it's about 0.2 mm dia (8 thou approx).

[kalee20]

Graham Station X - thanks for this offer! I've just taken out the SW oscillator coil. It's covered in wax and varnish, and I'm not stripping it until I've taken a few photos. However, I've unsoldered the ends of the windings from the solder tags in order to mike them up.

What I can see, is the tuning coil is 7.5 turns o.48mm tinned coper wire. The former is of moulded phenolic, with a helical groove to hold the turns spaced apart.

The feedback coil is very thin wire, 0.07mm, and it appears that it is wound between the turns of the main winding, in a second groove. I'm really hoping this is NOT the case, because I'm pinning my hopes on there being shorted turns in the feedback winding, yet if the coil is of this construction, I can't see how there can be!

I'll be stripping the coil for a re-wind. New wire for the main coil is not a problem - the feedback winding may take a phone call to the Scientific Wire Company. This radio has now become a challenge! Watch this space for photos!