Mains Hum - Ekco A274

[ms660]

Me too, it is what it is, any fresh scratches etc will usually stand out less with a clean, a light touch up and some polish, same for woodworm holes, unless the cabinet is hanging or big chunks of veneer are missing I would do the minimum.

Lawrence.

[Synchrodyne]

Originally Posted by cjwhoishe

I still think there is something wrong and after looking at the circuit I have a couple of questions.
1. Is it normal to apply AGC to only 1 IF stage(the last one) could the first stage not get over-driven?
2. The input RF amplifier (V1A configured as grounded grid) does not look normal for an ECC85 application could this be my cause of problem in that the ECC85 is 'picky' about working in this mode?

1. Regarding AGC and FM receivers, one finds so many variations that it is hard to generalize. Look at 12 different receiver/tuner circuits and you might well come up with 13 different answers! It would seem to be desirable to control where in the signal path that limiting occurs, and this seems to have been more of an issue with solid state circuitry. Limiting in the front end – more likely in the mixer than the RF amplifier – could lead to spurious response problems. For FM, the bellwether parameter is repeat spot suppression. Delayed RF amplifier AGC was sometimes used to prevent mixer overload. In an IF strip with distributed selectivity, early limiting may reduce selectivity, so AGC applied early in the IF stage might be used to control that aspect. AGC was also used to provide a more constant IF signal level to ratio demodulators. And “fast” (unfiltered) AGC was sometimes used to assist with AM rejection. Thus, the stages to which AGC bias was applied, and also the point from which it was derived varied according to its purpose(s). When looking at the AGC circuits of valved FM receivers and tuners, some puzzling is often required to figure out – approximately at least - the original intent.

In the case of the Ekco A274, I’d infer that the designers reckoned that in typical domestic use, with indoor or modest outdoor aerials, it was unlikely to be fed with an aerial signal of sufficient magnitude to cause any early overload problems, so that preventive AGC to that purpose was not needed. (And if overload was encountered when it was fed by a “large” aerial array, that could be cured quite simply through the use of a standard in-line attenuator at the aerial input socket.) With a two-stage IF strip and a ratio demodulator, Ekco could have configured the second stage as a combined amplifier/limiter, with limiting delayed until the signal level was sufficiently high, in which case no AGC would have been needed. But instead it opted to have the second IF stage as a full amplifier, with AGC bias from the ratio demodulator applied to the suppressor grid as a means of keeping the demodulator input reasonably constant. I don’t know, but possibly doing it this way allowed the second IF stage to have slightly higher gain. Also, although suppressor grid AGC had a lower slope (dB/volt), there was plenty of bias available from the ratio demodulator. Also, at high gain reductions, it likely avoided the problem with control grid AGC that the valve was biased back to the curvy part pf its characteristics, so getting into modulation rise distortion.

In complete contrast for example, the Pye FenMan II had AGC bias (obtained from the limiter grid) applied to the FM RF stage and the 1st and 2nd IF stages. It seems that the designers wanted to be sure that no limiting occurred until the limiter valve itself. This model shared its RF circuitry with the HFT111 hi-fi tuner, which was probably thought more likely to be fed with a big signal from a large aerial array, hence the need to control early overload.

2. The grounded grid configuration was the norm for the RF amplifier section of the ECC85, and this was the circuit described in Mullard “Valves, Tubes & Circuits” #19. It was also used by some makers in the neutralized grounded cathode mode, and also in the mixed mode.

Cheers,

[ukcol]

Originally Posted by cjwhoishe

...The input RF amplifier (V1A configured as grounded grid) does not look normal for an ECC85 application ......

Just to add to what Chris and Synchrodyne had to say about this -

Configuring the first half (the RF amplifier) of the ECC85 as grounded grid is standard practice.

A triode has an advantage over a pentode from the point of view of signal to noise ratio; in a pentode the flow of electrons from the cathode splits, one path going to the screen grid and the other going to the anode, it is this splitting of the electron path that generates a certain amount of noise.

Having chosen a triode over a pentode to take advantage of its noise performance the designer is faced with another problem. There is a small amount of capacitance between the anode and control grid that, at 100mHz, is significant. This capacitance provides a negative feedback path that reduces the gain of the stage, this is the famous "Miller Effect". The fact that the valve amplifies this fed back signal makes the value of capacitance "look" much larger.

By using the valve in grounded grid configuration this feedback cannot happen and the low noise performance of the triode can be exploited without suffering its disadvantages at high frequencies.