Identifican request - 3 Valve mains superhet

[pwdrive]

I'm surprised at that, 100pF to uF ought to I would have thought.

Lawrence.

[boater sam]

Try wrapping an EF39 in tinfoil and putting an external wire down to pin1 to earth it.
Decrease in volume with rising HT suggests a screen grid problem on the output valve decoupling, How do all the voltages compare with spec?
How and why are you increasing the HT anyway?

[Sparks]

The screen grid of the output valve goes staright to the HT without any dropper or decoupling!    I haven't taken the voltages yet.    As I am currently running the set on 150V HT, that is a lot lower than it is meant to be.     I turned it up to 200 Volts just to see what hapenned.      My HT supply is fully variable via a variac, so turning it up and down is no problem.    The wire insulation is apalling, rotten and crumbly and that is why, at the moment, I am running on 150 Volts.    Here is a close up taken yesterday, and also the tray of parts currently removed, some have been replaced, others have not!    
Bob

[boater sam]

Ugly under chassis, been got at.
If the screen is higher than the anode, its a bit odd. Where does the anode feed come from, off the reservoir cap to the output tranny? That would make it higher than the screen which I would expect is normal.
What ever is that ugly black blobby thing in the top left?
There is a heck of a lot of solder to clean up under there when you start replacing wires.
Nice to see a double tag strip though.

[Sparks]

The anode feed does actually come from somewhere else, so the screen probably is lower than the anode in this case.    I have often been told that the screen V is always less than the anode V, and agree that it should be, but it seldom is!    Most of the circuits I have seen in Radio & TV servicing books, the screen goes straight to the HT+ line, the same as the top of the output transformer.  
That black thing, top left is the 100pF aerial series capacitor.  
In the image, it is me that "got at it!"       I took the bolts out of the double tag strip and bent it back so I could get at the components.     When I put replacements in, I did not bother about doing it neatly, my first priority being to make it work again, which I have done.    When it is working properly, I can set about tidying it up!
Regarding the V3 cathode bypass capacitor.     This is an area where I got it completely wrong through not investigating fully what was happening.     The circuit showed that it was 25uF, and I only got a few pF.    At that time, I was unaware that one of the leads to my capacitance meter had become open circuit inside the insulation, and the reading was, of course, between the broken ends in series with the 25uF in the set.    Once I fixed the lead,  and tested the capacitor again, it showed 63uF.     I removed it, and found that 50uF was written on the casing, so someone had replaced it in the past.    I put a 22uF in place of it, and no real change in the performance.
But here is something for those more knowledgeable that myself, and I have not investigated it further, because we are going out shortly.    I had an EF39 in the V2 position that had no metalised coating at all.    I happened to touch the top cap, and immediately, the volume increased, the tone improved and the set became 100% stable!    No whistles or screeches.   I tuned across the whole band with finger on top cap, and the performance was absolutely perfect.    Remove finger, and whistling and shrieking again!
Bob

[boater sam]

I've had that before, top cap being the grid, I think the valve is oscillating and you dampen it with the finger.
Check all the caps around EF39 and the resistor feeding the grid. If all is good try a small pf value cap to ground on the grid. But these valves are in my experience prone to spurious oscillation, some more than others. If it is oscillating that would account for the volume going down with higher HT, the valve is being driven harder at the expense of the intended signal.
I'm assuming that the regeneration on the 1st stage is not the cause?

[Terry]

As the EF39 can work properly, as you've proved, unless you've got a glut of them it would be worth using a metallic paint   to replace the screen - I'm sure someone can suggest something suitable. Clean the earth wire first to ensure a good contact. If necessary extend it with a bit of thin wire, loop it round the glass at the bottom, trapping it in the crack between the glass and the base  then, while it's pulled as tight as possible, run some solder over it.

If it snaps off or isn't present, get a length of very thin tinned copper wire (or use one wire from a piece of stranded wire) and wind it round the earth pin, attaching it with a very small amount of solder then run it up the side of the base and loop it round the glass at the bottom, trapping it in the crack between the glass and the base  then, while it's pulled as tight as possible, run some solder over it as before.    

It wasn't uncommon to see this trick done on many radios in the past.                                                                                                        

[Crackle]

As a quick test you can wrap the valve in aluminium foil and then earth that. (obviously cutting it away from the top cap)
Not very pretty but it should help stop the instability.

Mike

[Sparks]

At the moment, I am not too bothered about it, because with 150 Volts HT, it is OK. I am concentrating more on replacing the wiring and searching for more faulty components, after which the instability may clear up of its own accord. I can't imagine that all the EF39s that I have tried have lost their metailising connections, although I haven't actually checked them with the meter. Currently taking a rest from it.
Bob

[Sparks]

I just had another look at it. I discovered when I put the meter prod on the anode of the EF39 (IF Amp), it suddenly became stable with excellent tone and volume across the band. Just the meter lead alone has the same effect. Very puzzling. All the components in that area are OK and the valve metalising is intact and connected. The anode V was 96 and also the screen V was 96. I disconnected the screen grid of the EF39 (which shared the dropper and decoupling capacitor with the frequency changer and fitted a separate resistor and capacitor for the IF amp. It did not make the slightest bit of difference. Perfectly stable with a one-foot long wire connected to G2, shrieking and howling without it!
Bob