Please click your language flag. Bitte Sprachflagge klicken.
  The thread rating is reflecting the best post rating. Have you rated this thread (best post)?

Measuring the response of the I.F. channel in AM receivers

Ernst Erb Jürgen Stichling Bernhard Nagel 
Please click the blue info button to read more about this page.
Forum » Technique, Repair, Restoration, Home construction ** » Alignement of radios: Superhet and TRF » Measuring the response of the I.F. channel in AM receivers
Roberto Guidorzi
Roberto Guidorzi
I  Articles: 27
Schem.: 6
Pict.: 98
14.Apr.12 15:15
Count of Thanks: 16
Reply  |  You aren't logged in. (Guest)   1

After restoring and aligning old superheterodyne receivers it can be of some interest to measure the frequency response of the I.F. channel to verify its alignment and also to document the set performance in view of comparisons with other receivers and also for possible future maintenance. This operation is even more important in case of receivers endowed with a variable selectivity control; this note reports some hints deriving from measures recently performed in a situation of this kind.

The measure of the frequency response of the I.F. channel in AM receivers can be performed either by means of sweep generators or point by point. The first way is a fast real-time procedure very useful during the alignment phase while the second is more time-consuming but potentially more accurate and repeatable. Making reference to the second case, the standard steps are the following:

1) Connect the chassis of the receiver to a good ground. Of course if the receiver has an AC/DC power supply or uses an autotransformer, an isolation transformer must be used to isolate the chassis from the line.

2) Connect the output of the signal generator to the frequency converter tube as shown in the example of Figure 1 (that makes reference to an octode). The capacitor Cg is not critical and can be chosen, for instance, between 500 pF and 10 nF.

Figure 1 – Connection of the signal generator to the conversion stage (example)

3) Block the local oscillator. The fastest way can consist in short-circuiting one of the oscillator coils. Sometimes this step is not mentioned in the procedures but I warmly recommend it. In fact, some asymmetries in the responses that I have observed in a couple of receivers disappeared after blocking the local oscillator because due to intermodulation phenomena.

4) If more bands are available, select MW and turn the tuning capacitor completely open.

5) Turn off the internal modulation in the signal generator and tune it accurately on the I.F. value. If the generator is not endowed with a digital frequency indication, the use of an external frequency meter is recommended.

6) Measure the d.c. voltage at the output of the demodulation circuit (point M in the example reported in Figure 2) and adjust the output level of the generator in order to obtain a reasonable value (e.g. not less than, say, 5 V).

Figure 2 – Demodulator (example)

7) Measure the voltage on the AGC line and connect this line to an external source (battery  or, better, a variable d.c. power supply) in order to impose the AGC voltage independently from the AGC circuit of the receiver. Usually the high value (1-2 Mohms) of the resistances in the AGC circuit allow to avoid any necessity of disconnecting the AGC line from the AGC circuit after its connection to the external voltage source.

It is now possible to make the desired measures at different frequencies. The steps should not be larger than 1 KHz and their range should be not less than 20 KHz. The procedure for a sweep measure would be essentially the same and would require an X-Y scope with the Y channel connected to the demodulator output and the  X channel connected to the deflection signal generated by the sweep generator. If a blanking signal is available it should be connected to the Z axis for canceling the retrace.

What can be obtained is a plot like that reported in blue in Figure 3 that refers to the maximal selectivity condition of a variable selectivity receiver. Note the completely misleading response (red) that would be obtained without locking the AGC voltage.

Figure 3 – Measured response of an I.F. channel (blue plot). The red plot reports the (useless) measure obtained without locking the AGC voltage.

The plot reported in Figure 3 shows a precise alignment of the receiver (the nominal I.F. value was 362.5 KHz) and good symmetry.

This approach is fast (its description requires more time than its application) and, if every step is performed with care, gives reliable and repeatable results. It can be noted, however, that what is actually measured is the response of the I.F. channel plus the demodulator stage and that the non linearity of the diode limits the dynamic range of the measure to slightly more than 20 dB. A direct measure of the I.F. signal at the output of the second I.F. transformer would lead to a more extended range and to a more accurate description of the behavior of the I.F. stage. A measure of this kind, however, can easily lead to misleading results because the load constituted by a scope or RF millivoltmeter probe would be excessive and lead to large errors due to the detuning of the I.F. transformer. It is however possible to perform reliably a measure of this kind as follows:

1) Remove the tube containing the demodulator diode.

2) Connect a capacitor having approximately the same value as the diode capacitance to the secondary winding of the I.F. transformer (Cs in Figure 4). Typical values are in the 2-3 pF range.

Figure 4 – Scope connection to the second IF transformer

3) Connect the scope to the previous capacitor and verify whether the peak of the response has exactly the same value as before. Increase Cs if the peak frequency has increased, reduce its value in the opposite case.

After performing the measures point by point, compare the obtained (normalized) values with those obtained with the standard procedure. What should be obtained is shown in Figure 5 where the plots coincide at the frequencies around the central one and show an increasing difference moving away from this frequency.

Figure 5 – IF frequency measures performed with the standard procedure (red) and by directly measuring the IF channel output (blue). Selectivity control set for minimum bandwidth.

The coincidence of the central measures obtained with the two methods assures that the direct measure has been performed properly. It can be noted that the direct measure leads to a more extended range, limited only by the scope sensitivity and by the disturbances (you have connected the chassis to a good ground, didn’t you?). It can be useful, when this second method is used, to increase the gain of the I.F. channel by acting on the AGC voltage; the plots reported in Figure 5 have been obtained, for instance, by setting -5.8 V on the AGC line for the standard measure and -3.5 V for the direct measure. An excellent agreement between the measures like that shown in Figure 5 can be obtained only by checking with care that the center frequency of the I.F. channel has not changed after connecting the scope; even a variation of 1 pF in Cs can introduce non negligible shifts.

Figure 6 - IF frequency measures performed with the standard procedure (red) and by directly measuring the IF channel output (blue). Selectivity control set for maximal bandwidth.

Figure 6 shows the response of the IF channel of the same receiver with the selectivity control set for maximal bandwidth. In this case the larger bandwidth reduces the gap between the measures at the extremes of the considered frequency interval.

Leonardo Munkeviz JR
Leonardo Munkeviz JR
BR  Articles: 3
Schem.: 1
Pict.: 6
29.Apr.12 07:15

Count of Thanks: 10
Reply  |  You aren't logged in. (Guest)   2

Excellent work,just wanted to add :Selectivity is directly proporcional to the transconductance of FI Valve channel