Superhet Alignment on Shortwave

[JHGibson]

Superhet alignment on medium and long wave is straightforward, but on shortwave you have the problem of the mixer signal trimmer adjustments pulling the local oscillator off its correct setting.
The standard approach is to rock the tuning while adjusting the signal trimmer for a peak, which is somewhat of an art.

A more concrete adjustment can be made by connecting a RF noise generator to the receiver and then peaking the mixer signal trimmer for max output - no rocking the tuning and there is no LO pulling at all.

RF noise generators are not very common but they are simple to build, just a 6.3 V zener and a 2 transistor amplifier. You can find a suitable circuit in the Measurements section of a ARRL Handbook, incorporated in an Antenna Noise Bridge. Just build the noise generator part.
John.

[Skywave]

Reference: for those of us who have access to 'The Radio Designer's Handbook', 4th. ed., Langford-Smith, this topic is covered extensively on pages 969 and 990. Far too much detail provided there for even a precis to made here! But does make very interesting reading.

Al. / Skywave.

[JHGibson]

In a pinch, (I was in my pyjamas, it was pouring with rain, and my noise gen was in the garage at the end of the garden!), for a noise generator, I used a table lamp with a dimmer control placed near the receiver aerial, and peaked the signal trimmer on the audio racket generated.
John.

[Ian - G4JQT]

If you have a SW radio with digital readout - even a cheap one - this can be a great help in aligning a superhet, particularly if you haven’t done it before.

Tune the digital radio to the alignment frequency PLUS the IF of the set to be aligned. (The local oscillator frequency is usually, but not always, above the tuned frequency on SW. On the rare occasions it is below, then subtract the IF. It’s never below the tuned frequency on MW or LW – think about it ;-)

Set the pointer of the radio being aligned to the alignment frequency and adjust the trimmer or inductor until the local oscillator is heard on the nearby digital radio. This way you can’t inadvertently tune to the image signal – which is easily done on many sets above about 10 MHz.

e.g. Alignment frequency is 15.00 MHz.
Set pointer to 15.00 MHz (20 metres)
IF is 470 kHz
Tune digital radio to 15.47 MHz
Adjust local oscillator capacitor or inductor until it is heard at 15.47 MHz.
Adjust aerial/mixer trimmers for best sensitivity at 15.00 MHz on set being aligned

The digital radio can be used to accurately set the frequency of an RF oscillator which is useful in peaking up the aerial/mixer stages at the calibration points if you are not using the noise generator method described above. (That sounds interesting, I'll give that a go). A DC-coupled oscilloscope or high impedance voltmeter across the AGC line helps get this spot on.

You cay buy a cheep digital set for under £10. The frequency accuracy is often not as good as one costing £30 or more, but sufficiently accurate for most analogue tuning scales. It’s a very handy piece of test gear if you don’t have a frequency counter and in some ways more useful as you don’t have to actually connect it to an oscillator to get a reading.

Regards,

Ian

[G8HQP Dave]

Are people talking at cross-purposes here? The OP was talking about pulling of the LO as the RF trimmer is adjusted. If this is happening, listening for the LO on a nearby set will merely confirm that pulling is taking place. The whole point of a noise generator is that it generates a signal at whatever frequency the receiver is actually tuned to, even if this changes when it shouldn't. A sensitive receiver could use band noise instead. An RF sig-gen will only give a signal at the frequency the receiver is supposed to be tuned to.

[Skywave]

Originally Posted by G8HQP Dave

The whole point of a noise generator is that it generates a signal at whatever frequency the receiver is actually tuned to.

That's not correct. A noise generator generates signals that occupy a very wide band-width. The receiver simply filters out those freqs. that lie within the receiver's much narrower band-width.

Al. / Skywave.

[MichaelR]

Which is the same thing isn't it ?

[Skywave]

Strictly speaking, no.

G8HQP's Post implies that the noise generator is adjusted to produce a signal at a freq. that the receiver has been previously tuned to. It's the receiver under test that is 'tuned' to the noise generator, not the other way round.

The difference is subtle, but significant.

Al.

[MichaelR]

Ok, I did not read his original post that way. "In Effect" is what he implied to me.

Happy New Year Al

Mike

[Skywave]

OK, Mike - I'll think it's time we left this now since it is a side issue to the main Thread. I think that those of us who have used wide-band noise generators know their appropriate use.

And a Happy New Year to you, too.

Al.

[G8HQP Dave]

Originally Posted by Skywave

That's not correct. A noise generator generates signals that occupy a very wide band-width. The receiver simply filters out those freqs. that lie within the receiver's much narrower band-width.

Al. / Skywave.

I think my statement is true, although not the whole truth, and perhaps slightly tongue-in-cheek. In addition to generating a signal at whatever frequency the receiver is actually tuned to, a noise generator also generates signals almost everywhere else although the only signal we are interested in is the first one. The point is that you don't have to adjust it, unlike a sig gen.