I have been recently contacted by a friend which asked for help, in order to repair a car radio installed in his Lancia Fulvia sedan, a memento of his father, which he was restoring.
Going into details, the car radio is an Autovox RA 2011 which dates back from 1967, a rather interesting model since not only entirely solid state (this details is not really adding any extra value to the set, since in that period the tube technology was being entirely replaced by the transistor and the integration of complete electronic circuits on silicon boards), but especially for its automatic electronic tuning of the station.
Indeed, the advertisement of the radio was proudly claiming that the radio is equipped with a fully electronic tuning system.
To better understand what this definition means, it is beneficial to recall an older article of mine, where I described another radio with automatic tuning (a German Saba Meersburg-Automatic 8, to be precise): radio with automatic research of the stations were already existing at the end of the ’30s, with control circuits which were of course electronic, but mechanically driven, that is the tuning devices were operated by an electrical motor, normally having a common axis with the tuning knob. The varicap diode (for its detailed definition, please see the wikipedia page) allowed the switch to a fully electronic tuning: the varicap is a device that changes its capacity (and therefore the
resonance of the tuning circuits) according to the voltage which is applied at its terminals. This allowed to stop using variable capacitors, enabling the development of smaller, more economic and simpler circuits: manual tuning could be done with a potentiometer which was similar to the one used for volume, or with an automatic circuit that had only to adjust a DC voltage.
Original advertising brochure was also adding that “the principle of electronic computers is applied to your
car radio FOR THE FIRST TIME IN THE WORLD BY AUTOVOX. Three
logical integrated micro-circuits selects the stations and implement a completely electronic automatic tuning.”
Well, when I read this sentence I hadn’t yet opened the radio, and was rather skeptical and doubtful: from one side I was thinking that the radio was more likely to be produced in some later years, from another side I was persuaded that it was a sheer marketing sentence and the radio had some traditional circuit, or maybe the radio had some analogical integrated circuit (such as a couple of transistors embedded in a single package), or even that the radio had really some logic IC but embedding modules from outside Europe (USA or Japan)… Well, my country’s current situation deceived me, Italy was really in the forefront of engineering at that time!
When I opened the radio, together with the multitude of components packed in the slender interior space (this, at first glance, discourages from starting any reparation), I could easily see the three metallic integrated circuits with a TO100 round case , two with 10 pins and one with 8, marked “SGS AUTOVOX 4, 5 e 6”. It was evident that everything was “made in Italy”!
Furthermore, but not very positive aspect, they were “custom” IC made by SGS(Società Generale Semiconduttori,
the supplier of semiconductors for the well-known Olivetti computers) specifically for Autovox: no chance to find a spare part if one is faulty (unless I scrap a similar set for parts)…
Finally, an even less positive aspect, I had no schematic of the radio and I didn’t know the functioning or the exact purpose of these ICs.
Only later I will discover that the operation and design of these integrated circuits was a trade secret, jealously guarded by Autovox: even the service manual (which I’ve received, sadly, only at the end of the restoration) pass over the topic, not even mentioning the content nor the operation of the automatic tuning.
Driven by curiosity during the restoration, I did many speculation about the operation of the automatic tuning and the content of these enigmatic custom ICs (some of these conjectures proved to be true, when I was later able to bring the radio to be back in working order): for instance, I believed that those were circuits for industrial usage (logic gates, etc.) which were re-branded for Autovox. SGS was a supplier for the Olivetti computers, and was also producing under licence manufacturing granted by Fairchild, which – as far as I know – was among the first to use that kind of metallic case for its own integrated circuits.
In order to state my other deductions which came into my mind during the several days of restoration of this set, I have to briefly describe the working of the radio: the automatic tuning is operated pressing the knob on the right (which, if pulled, will switch the tuning as manual). When this knob is pressed, the logic of the radio (one of the mysterious integrated circuits) generates a voltage ramp which is applied to the varicaps and, increasing, moves the tuning from the lower end of the frequencies to the upper end. When getting close to a station, a FM radio detector provides the sequence of triggering signal to stop the voltage ramp and therefore the tuning search: the radio is tuned to the station.
Another pressure on the knob will release the tuning and will restart the ramp. Once arrived at the end of the dial (where the maximum voltage of the ramp and, thus, of the varicap), the logic makes the voltage drop to the minimum value and the scan restart from scratch, back to the lower end of the dial.
From these notes we may understand a couple of facts: the operation is different from the tuning device in the Saba Automatic radios, since there is no automatic frequency control here. When a station is tuned, the device will no longer keep searching for it: this is intentional in a car radio, since the radio would start searching for a new station even in case of a temporary poor reception of a station. Moreover, automatic search is done in a single direction, from the lower to upper frequencies. This is possible thanks to the dial pointer, which is not really a pointer but a… drum wrapped with a black and red spiral which looks like a movable pointer when it rotates; the drum is moved not by a motor, but by a voltmeter that is moving according to the different varicap voltage. Simple but ingenious solution.
Long time and patients would have been required to do a complete reverse engineering of the content of the integrated circuits from the restoration of the functionality of the radio. I believe to have already spent too much time and efforts on this radio, mainly because I couldn’t find the schematics, and I wanted to avoid any risk of damage after its reparation for further tests… that being stated, I am quite sure that these ICs contains at least some operational circuits (one of which dedicated to the generation of the voltage ramp by charging a capacitor at constant current)
and, above all, a “sample and hold” circuit, serving the purpose (surely not trivial for the time) to memorize the voltage reached by the ramp when the FM radio detector has triggered a station (quite advanced technology for the period, since this required the usage of field-effect technology (FET)).
In this link found on the Internet, we can see an example of a package with 8 pin of a sample-and-hold circuit, which could make acceptable the hypothesis of having identified at least one of the “mysterious” integrated circuits.
Regarding the reparation, my friend warned me before sending me the radio, that he brought it to a local repairer who gave up after several days, affirming that the radio was very old, a big number of electrolytic capacitor were faulty and he didn’t want to take the risk of further damaging the radio while attempting to replace them.
After having brought the radio on my work bench and powering it up at 14 volt, the radio had only a strong constant noise in any wave bands and at any position of the volume knob. I also realized that nothing could be heard when injecting a signal on the live pin of the pot. With further inspection, I saw that an capacitor was leaking electrolyte, and indeed its measurement showed it had lost nearly 100% of its nominal capacitance value. Since the other electrolytic capacitors were the same type, I decided to replace them all, at the price of many working hours and with the patience of a saint…
When the recapping was finished,
– noise was still there, even with the volume was at the minimum,
– the radio was finally able to receive some station in the standard broadcast band, but the volume was still very low: to be more precise, it was louder when the radio was powered up between 12 and 13 volt, whereas the volume was lower and lower as the voltage was being increased to 14V (which is the normal voltage of the car when the engine is running);
– automatic tuning was still not working at all;
After several further attempts (and swearing), I could finally understand and solve the faults relating to the constant noise and the strange behavior when the voltage was varying. These problems were also connected to the defects in the automatic tuning.
Shortly, the preamplifier transistor, just below the volume potentiometer, is enabled or disabled by a muting circuit (coming from one of the three custom ICs), which should set the volume to zero when the automatic tuning is working or enable the pot when the radio is operating normally.
This muting circuit seemed to be faulty since there were no voltage. In these conditions, the transistor SHOULD be off and should not be amplifying. However, the transistor found in this set was getting somehow self-polarized on its own and was amplifying in any case, and the amplification was maximum at 11 volt, whereas the volume was decreasing towards zero when the voltage was increasing towards 14.
The same transistor was also the origin of the noise, but I had to spend some more time investigating before I could realize it. This kind of noise is usually associated to carbon resistors, and also the symptoms suggested this fault (connecting to earth ground the base of the transistor with a capacitor, the noise was disappearing whereas I was expecting it to remain), so I replaced all the resistors related to this circuit. But finally I luckily found an equivalent germanium transistor and, after having put it in place, the radio was (correctly) mute!
In order to bypass the problem of the muting circuit, I temporarily disconnected and I polarized the transistor with a resistor: I was finally able to make the radio working well, at least with manual tuning!
The bad news was that the problem with the muting circuit, connected to the now well-known custom IC, could mean that I had reached an impasse and that my friend had to make do with a radio with a manual tuning only!
But no! After some more time following the connections on the printed board and looking for faults, I realized that the integrated circuits were not powered up or, when powered, in a wrong way (I still had no technical documentation at this stage…) since operation at a voltage lower than 3 volt was inconceivable for circuits of those years.
Well, one of the supplies was short due to a shorted 4.3 volt zener diode.. After replacing this part, I powered the radio with a lot of trepidation… and everything was working properly, including the automatic tuning!!!
I’ve done a short video which shows how the automatic tuning works:
By a twist of fate, the day after having solved the problem, I’ve received the envelope from the Italian Antique Radio Association (AIRE) with the electric diagrams of this car radio and the similar models Concorde RA3016/RA3006… they were however useful for readjusting some alignments and do an even better job.
After this restoration, I decided to have some days of break, away from electronics (at least the “home” electronics), since during normal office hours it is my work) since the whole experience has been very demanding.
I thank my colleague Dario Tirel and my friend Nicola Giampietro for the tips they gave during this complex restoration, and the website “Viva Lancia” for the review.