Audio Home

FM tuner alignment.

to be completed

If you, like me, own an old V/T FM tuner, you are eventually going to face the problem of needing to replace one or more V/Ts and with that the problem of realigning the beast. If you're lucky you live near someone who understands this arcane procedure, owns the appropriate instruments and is willing to do it for a reasonable fee. If not you can either live with a malfunctioning tuner or try re-aligning it yourself. Having recently been through this procedure I thought it might be worth sharing the results. The following refers specifically to the Leak Trough Line 3 mono simply because that's what I own, listen to and have just realigned.

So why do V/T FM tuners need realignment? These old tuners use a series of tuned circuits which incorporate the characteristics of the V/Ts. This means that as the V/T parameters change with age or if one or more of them are replaced it becomes necessary to re-tune those circuits so that they are all work together – “in alignment”.

A general block diagram of an FM tuner is shown in Fig 1. The input, RF amplifier and oscillator stages are tuned to provide best sensitivity and selectivity. The oscillator, mixer and intermediate frequency stages are tuned to provide best gain and linearity. Yes that's right the oscillator needs to be tuned to both the input frequency and the intermediate frequency – this is due to “heterodyne” action. Briefly, heterodyne is a system invented by Major Armstrong which mixes an RF signal a specific frequency (known as the intermediate frequency or IF) above the broadcast frequency with the broadcast signal, derives the difference frequency from a mixer and uses this as the basis for all further amplification and processing. This avoids the difficulty of having several independently operating tuned circuits which must all change frequency together if a different signal is tuned. With the heterodyne system the difference frequency will be constant as long as the oscillator tracks the tuning of the input accurately. This requires an incredibly stable oscillator and this was a huge problem with RF coils in V/T gear as the heat generated by the V/Ts is enough to cause the coils to expand and the frequency to drift. A stability level of 200ppm was considered adequate. The “Trough Line” after which the Leak tuner is named is H.J Leak's genius solution to this problem - instead of using an RF coil he used a stable coaxial inductance tuning mechanism which gave this tuner a stability of 30ppm – an excellent figure for its day.

The problem of alignment then becomes fairly simple – to ensure that the IF stages and the limiter / discriminator are centred on the intermediate frequency. The value of the IF must be more than half of the difference between the lowest and highest tuned frequencies in the RF stage. Since the FM band is about 20MHz wide this means that the IF must be greater than 10 MHz, preferably a fair bit greater. Different tuners use different values of IF – 10.7 MHz is the most common, the Leak Trough Line uses 12.5 MHz. Now as the name suggests FM broadcasts modulate the “carrier” frequency with the signal which means that the actual signal frequency varies either side of the carrier. This modulation is carried through to the intermediate frequency so the IF stages need to pass a band of frequencies centred on the nominal IF. For best linearity the IF stages should be capable of passing signals well above and below the actual modulation values in use – at least +/-100kHz. To check this we need to be able to swing the modulation to a level which is beyond the pass band of the IF stages, so +/- 200kHz or more is a good idea. We therefore need a generator which can sweep across a range of frequencies between say 12.3 MHz and 12.7 MHz.

This has always been the difficulty with FM tuner alignment – RF sweep generators are fairly complex and expensive. Fortunately, with a bit of ingenuity we can use a common or garden function generator as long as it will generate a sine wave to around 13Mhz. My favoured candidate for this is the trusty Hewlett Packard HP3312A function generator. These can be found from second hand test gear suppliers or on Ebay for $250 or less. This is quite possibly more than your tuner is worth but a good function generator is a terrifically useful piece of test gear in its own right. Of course if your tuner uses the standard 10.7 MHz IF you only need to get to around 11MHz.

Next we need some way to measure what's going on. Again fortunately we can use a piece of gear which has numerous other uses- an oscilloscope. To use the method I will outline you will need to display two oscilloscope traces which are synchronized to the sweep rate of the generator. This means using either a dual beam oscilloscope with external time base synchronization output, or a 3 (or more) channel oscilloscope with the third channel as trigger synch on the sweep generator. My scope is a Tek 2465 which lacks an external time base synch output but has four channels so I used the second method. A word of warning here – the modulator output on the HP3312A is more than the channel 3 input on the Tek 2465 can handle so I inserted a 10:1 attenuator in line. Set the 3312A up on sinewave output and sweep mode, with the sweep range to 100Hz.Set the oscilloscope timebase to 1ms, trigger using Ch 3 and you should get this:

The general procedure with aligning FM tuners is to work from the last stage backwards – you need to make sure the 2 nd IF and discriminator are aligned before doing the 1 st IF and you need to make sure both 1st and 2nd IF are aligned before doing the mixer and RF stages. As mentioned above, the IF stages need to have a passband either side of the IF and in an ideal world this would have vertical sides and a perfectly horizontal top. This would require an infinite number of tuned circuits with infintite q all placed infinitely close together within the passband. Nobody's worked out how to do that yet, so we make do with two tuned circuits which are as close to symmetrically disposed either side of the IF as possible with a small centre dip. The IF response curve we are looking for is as shown in figure 2.

The limiter / discriminator should respond linearly to the modulation of the IF so in an ideal world it will be a perfectly straight line centred exactly on the IF. Again in the real world we can't have that so we go for the best linearity possible. The curve shape required is as shown in figure 3.

We can overlay the two curves and check that they are centred on the same frequency, which is why we wanted to use a dual trace oscilloscope – it makes the whole deal much easier . Fig 4

Now armed with all this information we can proceed to the alignment. First we need to know where in the tuner to apply signal and where to look for the IF response curves and the discriminator response curve. The general rule is that the input should go the grid of the V/T immediately before the transformer being tuned and the response check on the AGC of tuning indication. For information specific to your tuner you may need to consult its specifications or work from the circuit diagram. I am heavily indebted to Steve Spicer, general all round Leak guru at Kiewa Valley Audio ( http://home.mira.net/~kiewavly/TLalign.html ) who provided the original alignment notes.

Click photo for larger image.

If it is still not legible you can open this file: Troughline.asc but it will only run in SwitcherCad, which is available free from Linear Technology .

For our purposes we can check the discriminator output at the point where the audio signal is taken out, which is where R37 attaches to the tag board next to V5 / T4

We can check the output of the IF stages anywhere after the second IF, the junction of R25 and C36 is convenient

First up put a signal into the 2nd IF stage and adjust the discriminator and the 2nd IF. A convenient insertion point for the second IF is green wire between T2 and pin 2 of V4

Set the turn the AFC on the tuner off and set the output voltage of the function generator so that the tuning indicator has about a 1 cm gap between the lit bands. We now adjust the discriminator transformer (T4) slugs to give the discriminator curve shown above. Once that is established, adjust the 2nd IF(T3) to give the curve shown.

Once that is established, move the generator output back to the Ist IF coil, a convenient point is the wire between L5 and pin 2 of V3

Check the discriminator curve again, then check the IF curve and this time adjust the slugs on the Ist IF transformer(T2). Once his is brought to optimal shape we go back to the beginning and do it all again. I susally find that by the third go through the cahnges are so small that everything stays in synch.

The IF and discriminator sections are now aligned, so we can proceed back to the front end. I simply attach the generator output to the grid of V2 (to be completed).