Elenco Radio Build: Part III

 October 18, 2024 · by Budi Isdiyanto

The next part of the build is to assemble and test the AM detector and automatic gain control circuits:

AM Detector

The “detector” is surprisingly simple, it’s just a 5 KHz low pass filter with a diode to chop off the positive side of the IF signal.

The IF (Intermediate Frequency) is a copy of the original transmitted signal but using a fixed carrier frequency of 455 KHz rather than the original carrier frequency of the station. The signal is created in the initial stage of the receiver so that the rest of the pipeline (i.e. the amplifiers and bandpass filters) can be tuned to this fixed frequency band, i.e. 455 KHz +/- 5 KHz.

Testing the detector is simple enough: Pass in a 1 KHz modulated 455 KHz carrier, using a signal generator, and check that the output to the audio amplifier is a 1 KHz tone. Then increase the 1 KHz signal until you get -3dB attenuation of the output to find the bandwidth of the filter. It should be greater than 5 KHz, otherwise you’ll be losing part of the audio.

In my case the bandwidth was only 4.4 KHz as the installed C38 capacitor was actually closer to 20 nF than the 10 nF it was rated at. Replacing it with one that was around 11 nF got the bandwidth up to 6 KHz exactly.

Automatic Gain Control

Another simple and ingenious circuit is the AGC which is a feed back loop from the AM detector output to the input of the first signal amplifier via a really low pass filter. The two things to note here is that the low pass filter creates a DC “averaging” of the detector’s output signal, and that the output signal, due to the orientation of the diode D4, is negative. This means that if the audio’s overall output level increases then the averaged negative voltage from the AGC also increases on the first amplifier’s input (the base of Q8) which will attenuate the gain of Q8. This negative feedback loop ensures that the audio output from the AM detector will stay at a constant level even if the signal received from the radio station changes in strength.

Lead Solder

In Offset Voltage’s videos he states that the PCB will not take to lead based solder. However I decided to try it and it worked fine. The large traces and pads still made achieving good flow hard but being able to work at 650F rather than 750F is a big plus. I found I still needed to use a lot of solder on each joint and not really happy with them still. Might try more heat in future.

Second Amplifier Stage and Filter

Next was to assemble and test the second amplifier circuit:

The output of the second amplifier goes through T8, which is a bandpass filter, and then on to the AM detector. The manual describes how to test the gain of the transistor and how to tune T8 for 455 KHz, and to test its bandwidth.

Here I got a some interesting results. The gain came out at 256 which is high but I guess transistors can be between 100 to 300. The bandwidth of T8, tuned to 455 KHz went from 413 KHz to 459 KhHz which seems very lopsided. In the manual it notes that the bandwidth should be around 445 KHz to 465 KHz for a single stage.

It also states that the input capacitance of the oscilloscope must be 12 pF or less. Mine are rated at 17pF so not sure if that is messing things up. It might also have to do with the transformer itself. Maybe the ones in the kit are not the highest quality?

First Amplifier Stage and Filter

Next was to assemble and test the first amplifier circuit:

This is identical to the second amplifier but with T7 as the bandpass filter and the base is adjusted by the AGC output.

Here I got a gain of 400 for Q8 which seems crazy high.

The AGC is disabled during testing. When the AGC is enabled again it should cause a 4 Vpp signal to drop to 0.8 Vpp. In my case it dropped to 0.2 Vpp which, judging from the extreme gains of the amplifiers, is probably correct. Still, at this point I’m a little nervous that there might be a problem with the build… Is so much gain going to increase noise?

Oscillator, Mixer and Antenna

Final assembly is the oscillator/mixer circuit, with antenna:

And large tuning capacitor C1 that has the dial on it:

From watching Offset Voltage’s videos I have a fairly good understanding of how this section of the radio works, though it’s not entirely clear to me yet how the “oscillations” of the oscillator happen. Nevertheless Q7 oscillates and its frequency is determined by combined inductance of L5 and capacitance of C1. C1 is also used with the antenna coil to create a bandpass filter that filters out all frequencies except the one we select with the dial.

The trick here is to tune the tuning capacitor C1, using two trimmers on its back and L5, to not only to reject all the frequencies except the one selected on the dial, but also to set the oscillator to oscillate at exactly 455 KHz above that selected frequency.

So if you turn the dial to 1000 KHz then the frequencies coming from the antenna to Q7’s base would only be in the range 995 KHz to 1005 KHz, i.e. we are only selecting the frequencies that belong to the station broadcasting at 1000 KHz. However, Q7 is also oscillating at 455 KHz above 1000 KHz at 1455 KHz and so the station frequencies are getting mixed mixed together with the 1455 KHz singal.

When you mix two frequencies together you get the original two frequencies plus two new frequencies; the sum of the two originals and the difference of the originals. So on the collector of Q7 we now have:

• The original station signal (which is the 5 KHz audio signal @ 1000 KHz)
• The oscillator signal (which is 1455 KHZ)
• The sum of two (which is the 5 KHz audio signal @ 2455 KHz)
• The difference of the two (which is the 5 KHz audio signal @ 455 KHz)

The neat thing here is that regardless of the frequency of the station there is always a 455 KHz version of the signal at the collector of Q7, i.e. the difference of the original signal and the oscillator.

T6 is another bandpass filter that only passes the audio signal @ 455 KHz on to the amplifiers. Pretty neat!

The manual details how to achieve all this tuning which is definitely easier if you have a signal generator that can do AM modulation.

I was able to tune the oscillator however it became clear that my amplifier transistors have way too much gain, to the point where the radio is unusable. Turning up the volume control just a small amount causes distortion and extreme loudness. Need to fix this problem first before I can complete the tuning and have a functioning AM radio. So close….