Elenco Radio Build: Part II

October 18, 2024 · by Budi Isdiyanto

While testing the radio’s audio amplifier in the last part, it was definitely annoying having to hearing the test tones and annoying for my family too. Offset Voltage recommended making a “dummy load” for the amplifier via the earphone plug with resistors attached to emulate a speaker’s 8Ω impedance, though my understanding is that a speaker’s impedance rating is a nominal value and in reality the impedance can change a lot depending on the frequencies going through it. Nevertheless it makes sense to have some load resistance for the amp to work against so why not choose 8Ω.

For the dummy load I used three 1/4 Watt resistors in parallel in order to handle the roughly 700 mW of max power that the amplifier can output. Using two 22Ω and one 27Ω resistors I was able to get a combined resistance of 8.6Ω.

The Dummy Load

Creating the dummy load was not so easy.

Using heat shrink to bind the three resistors together was the easy part and it helped making soldering the earphone wires to the resistor leads a lot more manageable.

The tricky part was dealing with the very fine earphone wires. I don’t know what kind of wires they are but when you strip the insulation they look like they are bare but are still coated in some kind of insulation. I used a flame to burn it off, as well as a blade to lightly scrape the surface, and this did the trick.

Next was soldering them to their respective ends of the resistors. Fiddly work but not too hard. However the wires are so fragile I felt it would be smart to use epoxy to protect them from mechanical wear. The epoxy along with a final heat shrink over the whole thing made for a very solid device. Very happy with how it turned out.

Next up was to assemble the alternative amplifier that uses discrete transistors to replace the LM386.

As an aside: I found that there was a version of the kit that only used discrete components and has since been discontinued: The AM/FM 108TK Rev-L

Wonder why it changed? Maybe educators wanted the option to use an LM386 to shorten build times for their students?

Transistor Amplifier

The transistor amplifier is on its own board and simple to assemble:

Figured the soldering for this board would be much easier than the main board as the traces and pads are much smaller, as you can see:

Nevertheless it was still challenging. The long header pins ended up cockeyed due to the heat melting their plastic holders. Not terrible but took a lot of bending and tweaking to get them aligned enough to all go together into the IC socket on the main board:

The other thing to note is that when the amplifier is installed it covers the earphone jack and even without my modification it would be impossible to have the earphone plugged in at the same time, so for testing I’m going to have to listen to the annoying tones again. When done with this section, the LM386 is going back in and the dummy load will be used for the rest of the build.

Fixing the Stand

Another suggestion from Offset Voltage was to use some tape on the supplied stand to stop the radio from slipping out so easily.

I used three layers of gaffer tape. The tape doesn’t need to go down far into the slot:

Just creating a small lip on the inside will keep the board from slipping out:

Testing the New Amplifier

The tests are the same as for the LM386, except for an additional transistor bias test and a DC gain test. The instructions for the DC gain test were not clear that you effectively need two power sources; the regular 9V as before and an adjustable power source that gets feed into test point 16 via a 1 meg resistor. The instructions seem to imply only one power source.

Here are the results for the transistor amp, along with the results for the LM386 for comparision:

Metric	Expected	Measured	LM368 Results
Idle Current	<10 mA	6.8 mA	4.4 mA
DC Gain	47.4	41.66	N/A
AC Gain	100-180	122	166
AC Bandwidth	>30 Khz	62 Khz	52 Khz
Clip Voltage	~	5.04 V	6.7 V
Max Power Out	> 200 mW	398 mW	714 mW
Max Source Current	~	109 mA	143 mA
Source Voltage	9 V	8.99 V	9 V
Max Source Power	~	0.98 W	1.29 W
Efficiency	~ 50 %	40 %	55 %

You can see the LM386 is quite a bit more efficient.

One interesting thing you can do with the transistor amp is bypass its biasing diode and view the effect of crossover on the output:

And doing the tests again with the biasing diode removed resulted in an efficiency of 45% compared to 40% with it in.

Next up is to put the LM386 back in and start on the AM detector and automatic gain control section… with the dummy load, yay!