IF alignment with AFC

[Jan Zodiac]

Hi all.

I am now ready to do an alignment on the FM section of an old Bofa tuner. I made a thread on it some time ago regarding which valve to use in the front end.

I have attached a pdf file of the schematic, and mine has the AFC circuit fitted (shown below the main schematic).

The problem is that AFC is in function all the time, and I believe this will be problematic when attempting to do an alignment. The AFC function is not working as a very strong pull in of the station. I think it was applied just to keep the tuner from drifting off station. At least it did not strike me as pulling strongly, but since someone before me had tampered with almost everything inside this tuner, a lot can be completely misaligned.

The circuit seems different from what I have seen on other period radios from that time. Normally a 4VDC reference voltage is applied to the varicap diode, and then using the voltage from the demodulator on the other end of the diode to regulate from.

In this tuner, the "reference voltage" moves in the opposite direction (opposite polarity) of the voltage from the demodulator. In this way I believe it is supposed to regulate with a greater effect.

I think the varicap diode is a BA101, but since is was launched in '63 I could be wrong about that. The tuner was developped in '58-'59, so either another earlier type of diode was used, or the AFC option was not introduced before late in production. This tuner might have been produced until mid '60s when Bofa went into collaboration with Philips, and eventually shut down.

My question is, how can I shut off the AFC circuit so I can make an alignment of the IF strip?
My guess would be to ground the connection coming from the demodulator to match zero volts when a station is on spot.
Then connect a 4VDC supply on the other connection, overriding the regulation happening at this point.

After that, make a complete alignment. Then apply an unmodulated signal to the tuner and disconnect the 4V supply (keeping the other side od the varicap diode grounded). Adjust the trimmer to maintain zero volts coming from the demodulator, and then hoping it stays stable when I remove the connection to ground on the other side of the varicap...

Am I right about that, or is there something else I should take into consideration? All other radios I have aligned, had a switch to kill the AFC function, but not this one...

Regards,

Jan

[simpsons]

Is there a reason why you think the set needs realignment?

When looking at service manuals for FM and sometimes AM IF alignment, the objective is to optimise the symmetrical pattern of the response which is difficult if just peaking the coils.

Perhaps I have been lucky but, both Grundig and Philips valve radios I have seem fine having been built in the 1960's.

Chris

[Radio Wrangler]

AFC has to be slow. If it was quick-reacting, it would start to track the lower frequency components of the frequency modulation, and will strip-off the bass portions of the audio.

The answer then is to sweep fast enough for AFC reaction to be insignificant.

Sweeping too fast causes the highest-Q filters in the receiver - the IF filters to be too slow to react fully and you get a distorted picture.

Between these two extremes there is enough space to allow an IF response curve or discriminator to be potted quite well enough.

Speed up your scanning until you see your picture shift rightwards if you are using a sawtooth left-to-right scan and see the peak reduce in amplitude, then slow from this speed until a bit of further slowing makes no difference.

With AFC, you won't get a good idea of just where the discriminator centre frequency happens to be. You can move the centre of youe sweeper around a bit to explore and find the limitations, then find the midpoint. Ideally you want to defeat the AGC for this and have a marker generator to put a pip on your trace at a known frequency.

If your sweeper can do 10.7MHz, you can inject at the IF and the AFC becomes moot.

David

[Jan Zodiac]

Hi Chris,

Yes, the alignment is off. On station (max. on tuning indicator) the demodulator is off, causing distortion. The AFC tries to pull it to zero volts on the demod. which then again causes the signal strength to drop. I suspect someone at some point has been "at it" with this set.

I use a sweeper/marker for alignment, to make sure the waveform is symmetrical as you mention (both IF curve and demod. curve).

The problem is, that the AFC circuit constantly tries to compensate, and I would like to kill its function while aligning.

I agree with Philips sets, probably others too, often still being spot on after decades.

Jan

[Mr 1936]

Hi

That is an unusual circuit

I think the simplest method (and probably the one used by the factory) is to inject with a sweeper at 10.7 MHz. The AFC will be in circuit but will have no effect on an IF-only alignment.

The LO alignment can then be done separately to get the tuning scale correct, then the RF stages to peak the sensitivity on a weak signal.

[Jan Zodiac]

Yes, since I can't adjust speed of the sweep (sweeps at mains frequency), I will have to inject into the IF strip directly. The sweeper/marker can do 10,7MHz so no problem in that. The problem is to place a metal can over the tuner valve, because of the mechanical construction of the set.

I will have to move the complete front end to slide a can over the valve. In fact it has to be moved just to replace the valve, risking to have the dial string slipping of the wheel. The whole set is a bit odd when it comes to mechanical construction.
That's why I thought it would be easier/faster just to sweep from the antenna input.

It will of course be possible to sweep into the IF directly after the front end, but that way I will not get the first 10,7MHz band filter into the signal path. That filter is in the front end casing.


Jan