Decca VHF FM radion

[djrm]

Yes the main unit looks like the photos in the links, It has the Decca pickup fitted. Nearly all the caps have been changed before I got involved to new yellow poly ones. (part of the problem was some wiring errors made in the process) The screen resistors are a bit on the high side, there is no potential divider. I'll change the screen resistors and see what happens, and then re-check the others too at the same time. One of the valves wasn't even plugged in properly which didn't help! I've read people remove the spring clips - might be worthwhile insurance if I take a valve out.

Here is picture of the underside, not a pretty sight. David.

   

[Mike Watterson]

Not too bad.

Some sockets don't have the clip which anyway can rust and seize resulting in valve damage on insertion or removal. Really the clips only needed (if at all) for car radios and transportable sets, oh and a few battery sets, but most common was the DK40 and the other valves would be B7G anyway. The DK40 replaced by DK92, USA used 1L6. The DK91 and DK96 weren't much good for SW.

Take off the spring clips!

Most resistors in a valve set can be +/- 25% or more and it won't make much difference. Cathode resistors (especially) and some screen grid resistors are the exception, needing better tolerance.

Sounds like you are on the home stretch. You have a castle, hall, mansion etc to put the Stereo Decola in? I'd have to remove the settee and sit on the hearth if we had one.

I like to change caps one at a time and cut leaving a stub of wire.

[djrm]

I would like to fix the thing before doing a blanket capacitor change. But it was like this when I got it, I'm just trying to fix it. I'm repairing the Decola for somebody else, I appreciate it for what it is but I don't like it.

There are no cathode resistors, the screen resistors were marked 82k but read around 110k, changing them has raised the screen voltage from 75 to 95 volts, there is about 5 volts on the detector capacitor now.

I'm going to have a exploratory go at IF alignment next to see if it helps any. The volume is just about acceptable now. I'll measure the output level tonight.

[Mike Watterson]

OK, 95 V on g2 is good.

So maybe someone twiddled.

Not far to to go now. If you even get to 10V on the detector, you could be content!

The amps seem practically like guitar amps. It seems designed to play organ music rather than being really HiFi. Strange for 1959-1962 ish, especially the VHF tuner which is like 1954.

I'm glad you are making incremental progress.

Do you know how to align FM IFs and the Ratio detector? Hint, without a wobbulator you just use the generator at 10.7 MHz with no modulation and measure voltage on the ratio detector and after main IF at 10.7 MHz, adjust generator +/- to see that ratio detector is correct.
Then put AM modulation as a final check. It should only give DC level and no audio.

Then adjust the VHF mixer/osc to be on frequency.

There should be detailed alignment procedure for that kind of detector somewhere. But in any case, you don't need AM or FM modulation on the generator to align FM IF or detector or RF.

The magic eye will fully close on a strong local station, even though very faded.

[djrm]

Thanks for the instructions Mike. My chapman has a detailed alignment manual following a similar procedure.

I started to try adjusting this Decca but the cores are stiff and I don't want to crack them. They are locked with very fragile elastic. I'm not sure what to try next to get the output level up. It's at about 200mv pp at presen t using my outdoor aerial.

I have a faster scope available now, (Tek 453) the idea is to check the gain of the if stages. David

[Mike Watterson]

Note that even a 10:1 scope probe can detune even AM IFTs, never mind 10.7 MHz.

The 75V changed to 95V on g2 with resistor change suggests the EF41s are OK.

Best idea is to unplug ECC85 and use 10pF and 100k in series to ECC85 mixer/osc section anode and measure ratio detector volts on the cap there as you slowly sweep from 10.1 to 11.3 MHz (or wider) and plot graph. Either by hand or entering into spreadsheet. Use as low a drive level as will give a reading at 10.1 MHz.

If the cores are so stiff then it's unlikely they have drifted or been twiddled. At least it goes, so you don't want to damage a core. Sometimes I'll let the cores soak in WD40, but that only much helps with hard wax.

[Mike Watterson]

Have you got a known good EB91?
If not, you can unplug it and poke in a pair of 1N4148, suitably orientated, to eliminate that as a possibility. Though they rarely give trouble. Or a pair of 1N60, but the greater sensitivity isn't needed and the max PIV isn't really good enough.

[djrm]

Hi Mike,
I'll check the diodes by substituting something, better to be sure one way or the other.
I got a bit sidetracked fixing four faults on the tek 'scope I borrowed. It has a really bright clear display -I love it!
David.

[djrm]

1st attempt at a response plot, this one ino grid if first IF amplifier.
Next time into anode of mixer. step ~50kHz, input level ~20mV
x=MHz, y=DC mV on detector capacitor
   

[Mike Watterson]

Only one peak, so it's unlikely the IFTs need adjusted.