Released in December 1957, the TR-63 was Sony's first pocket-size transistor radio. It's a 6-transistor superhet design with some interesting design features, including the use of Sony-manufactured NPN transistors in the circuit.

Masaru Ibuka served with the Impe- used in schools and courts. much discussion, Sony's research labo- rial Navy Wartime Research Committee Following Ibuka's visionary 1952 ratory head, Mikato Kikuchi, suggested during World War 2, leaving in 1946 to trip to the USA to sign a licence with dropping Bells' preferred doping agent, join Akio Morita to form Tokyo Tsushin Western Electric, Sony acquired pat- indium, and substituting phosphorus Kogyo Kabushiki Kaisha, "Totsuko". ent rights for the transistor and subse- instead. When that didn't work, Morita Morita, a physics graduate, had served quently began manufacturing portable called for "more doping"! alongside Ibuka in the Research Com- radios in 1955. It soon paid off and Sony were able mittee, and their friendship laid the to produce the transistors used in their foundations for the international pow- Early difficulties first solid-state radios. Their TR-55 erhouse we now know simply as Sony. Sony preferred NPN transistors be- model, released in 1955, is now a rarity Tokyo Tsushin Kogyo's first prod- cause of their better high-frequency and the last one to be listed online some uct, a rice cooker, says a lot about the response but were initially unable to years ago had a price tag of $US1500. company. Japan had suffered massive produce working examples. One can only imagine the energy destruction during World War 2 due to NPN devices exploit the fact that invested by Sony to leap from Ibuka's bombing and people needed utensils to electrons move more quickly than licensing agreement to a marketable cook their staple food, which was rice. holes, ie, they have higher mobility. transistor radio in just three years. It's So a rice cooker that simply used two This is critical in the base region and also possible to imagine their frustra- insulated metal plates ingeniously met it's here that low mobility has the most tion at being pipped at the post by Re- a vital need. That combination of op- effect on high-frequency performance. gency's TR-1 transistor radio (SILICON portunity and ingenuity set the model The problem is that NPN devices were CHIP, April 2013), which was released for Sony's future. Their first radio- more difficult to manufacture using less than six months before. related product, a shortwave converter germanium feedstock. Sony's first "pocket-size" transistor for broadcast-only radios, helped open Knowing that, theoretically, NPN radio, the TR-63, was subsequently Japanese society up to the wider world. transistors were the way to go, Sony released in December 1957. It was, Tape recorders subsequently became saw experiment after experiment fail to however, reputed to be too big for a a major product line and were widely demonstrate useful performance. After standard shirt pocket and the story goes

82 SILICON CHIP siliconchip. corn. au R9 220 R11220 W1 on Vol Cont.

• R16 m 5attery 27 k 1 XS 9V DS L121112 2766 1743 R14 7: IIIMICO C13 C16 200 p R2 t 20 T2 — C3 R17 G.. ToPert 27524 42R Red (5 10 n T R1 5 k D. RIB White Volume aig 42 R

RB C14 C2 it Speaker 3k3 30 8 ohms

NOTES: RF voltages 30 % modulated, 400 Hz, audio 400 Hz, 10 mW output. Resistances in ohms unless k = x1000, all10% unless otherwise. Drawn: lan Charles Batty, 12107/2015) Capacitances: whole numbers/ p = picofarads, n o nanofarads, n = microfarads. DC voltages with VTVM, no signal, thus 0 except max signal thu RFIAudio signal injections Fig.1: the circuit uses six NPN transistors (X1-X6). X1 is the converter stage, X2 & X3 are IF amplifier stages, X4 is an audio pre-driver and X5 & X6 form a push-pull audio output stage.

that, for its launch, Sony had special The TR-63 of 1957 cracked open the The TR-63 N shirts made with pockets that could US market and launched the new in- was one of 1, take the radios. dustry of consumer microelectronics". the last sets With total exports to the US alone with Sony's Sony's TR-63 of about 100,000, the TR-63 was a true old "lighting bolt" logo. At first glance, Sony's TR-63 is a runaway success. pretty conventional 6-transistor set, The accompanying photo of the TR- with three transistors used in the RF/ 63 shows the red "Conelrad" marks IF section and the other three in the on the dial at 640kHz and 1240kHz, audio amplifier stage. All the transis- as required by US law at that time. So tors were manufactured by Sony and what was "Conelrad"? they are all NPN types. Basically, this acronym stood for onauseum.org and other sites. Any im- As noted above, Sony preferred NPN CONtrol of ELectronic RADiation and portant differences between the R-6C2 transistors because of their better high- was set up in the US in 1951 to pro- and R-6C1 are noted in the following frequency performance. My set was kit- vide emergency radio warnings to the circuit description and on the circuit ted out with the rectangular TO-22 can public during the Cold War. If an alert diagram. transistors, the same style as used by was received, most radio stations were Converter stage X1 uses base injec- Texas Instruments in the Regency TR-1. required to cease transmission, while tion and a cut-plate tuning gang (ie, the Sony's hand-held TR-63 was offered each remaining station was to move oscillator section is smaller than the in lemon, green, red and black. It used to either 640kHz or 1240kHz. They antenna section), so there's no need for a miniature, solid-dielectric "poly- would transmit for several minutes a padder capacitor. This stage follows varicon" for the tuning capacitor and and then go off the air, and another the common practice of fixed bias, so it also required a new battery design station would take over on the same gain control is left for the following that became the iconic "PP9" and set frequency in a "round robin" chain, IF section. the standard for transistor radios. the idea being to confuse enemy air- The first IF transformer (L3) uses As a piece of portable electronics, the craft that might be navigating using a tapped, tuned primary and an un- TR-63 is a winner. It's small enough to radio direction finding. tapped secondary and this feeds the pop into my shirt pocket, something By law, radio sets manufactured be- first IF amplifier stage which is based which couldn't be said for the TR-1 tween 1953 and 1963 had the required on X2. This stage is gain-controlled and other early sets from Raytheon, GE frequencies marked by the triangle-in- by the DC voltage fed back from the and Zenith. It also fits the hand better, circle (CD Mark) symbol of Civil De- demodulator. Unusually, X2's bias is the rounded edges giving it an easier fence, so that the set could be quickly derived from a voltage divider (R6 & feel than many others. tuned to either 640kHz or 1240kHz. R7). While this would usually provide What's more, the TR-63 is a good constant bias and thus constant gain, performer. It's also one of Sony's last Circuit details R6 and R7 have higher values than sets with the old "lighting bolt" logo Several circuit variations exist (de- usual and this allows "relaxed" control that was superseded by the "Roman noted by the circuit board number) and of X2's base voltage. text" logo we're more familiar with. As these are based on either the early pro- Basically, this allows the AGC cir- well, it carries the "Totsuko" stamp on duction R-6C1 sets or the later R-6C2 cuit to control X2's gain but with less the rear cover. version. The circuit shown here (Fig.1) effect than in the circuits commonly But it's not just an elegant personal is based on my R-6C2 and is also the used in other sets. radio. It's described thus in Schiffer's version shown in an H. W. Sams Pho- The second IF amplifier is based on The Portable Radio in American Life: tofact. transistor X3 and this gets its bias from ". . . (Sony) was not first, but its tran- In addition, the schematics for both X2's emitter, so AGC is applied to both sistor radio was the most successful. versions are available on www.radi- IF stages to give effective control. Note

siliconchip.com.au JANUARY 2016 83 which is shunted by a top-cut capacitor Output Output in both versions. T2 in turn drives a 3.5- inch (89mm) internal speaker via an Bias earphone socket. The earphone socket FAD, disconnects the loudspeaker when an Diode AY-0-VAt. 1 earphone is plugged in. MALLORY : R OR EQUiV LEN Initial tests 1st Converter This was another easy set when it Audio 1 came to restoration, at least as far as Earphone its appearance was concerned. A good clean and a light polish were all that OSC', Socket were needed to restore it to near-new condition. A quick check of the ear- Coil phone socket revealed that it was OK and I gave the volume control a light spray of contact cleaner to ensure trou- ble-free operation. I then applied power and checked This view inside the unit shows the PCB from the component side. The parts are the supply current. This was as ex- packed close together, although individual components are still easy to access. pected and there was some noise from the set when the volume control was also that the AGC return from the de- and also to the emitter of X4, the audio operated. This was then followed by modulator is fed to X2's emitter, again driver stage. The AGC control voltage wild oscillation on all stations and an unusual configuration. Commonly, itself is derived from D1 's anode and then silence. the AGC return is to ground, which is series-fed back through the third IF Replacing Cl and C2 (both 30µF means that the IF amplifier's emitter re- transformer's secondary to X2's base electrolytics) cured the oscillation at sistor forms a negative feedback circuit (both versions). those times when the set was working. for the AGC control voltage. This helps Unfortunately, there were still times to "soften" the very strong "sharp cut- Audio amplifier when it refused to work. off" AGC action that would otherwise The recovered audio from demodu- I soon discovered that when the set occur if the control voltage were sim- lator D1 is fed to transistor X4 via the stopped working, X2's base and emit- ply applied between base and emitter. volume control and capacitor C3. This ter voltages were way too low. The In operation, the R-6C2 version of audio driver uses combination bias. bias circuit itself checked out OK and the TR-63 applies a moderate amount The R-6C1 circuit (unusually) connects the problem turned out to be a faulty of AGC to both IF stages and directly X4's emitter to X2's emitter, so that X4's transistor. applies AGC between X2's base and emitter voltage varies somewhat with In operation, X2's internal collector emitter. By contrast, the 6C1 uses a AGC action. The R-6C2 circuit omits connection was going intermittently conventional series bias circuit for X2 this connection, giving a constant volt- open circuit. Normally, the bias circuit but still has the AGC voltage applied age on X4's emitter. supplies the base current and multi- directly between base and emitter. X4 feeds driver transformer T1's pri- plying this by the transistor's current Transistor X2's collector feeds the mary. The R-6C1's circuit shunts this gain produces the emitter current and tapped, tuned primary of the second winding with a treble-cut capacitor but thus the intended voltage across the IF transformer (L4). As shown, L4's the R-6C2 omits this component. Trans- emitter resistor. centre tap connects directly to the sup- former T1's centre-tapped secondary However, if the collector connec- ply rail, while the top of the primary then drives a push-pull Class-B out- tion goes open circuit, the base-emitter connects to neutralising capacitor C10 put stage based on transistors X5 & X6. junction behaves as a simple forward- (2pF). L4's untapped, untuned second- This stage uses bias diode D2, which biased diode. If that happens, the emit- ary feeds the base of the second IF am- is described as a "varistor". ter resistor is pretty much shunted plifier (X3). In reality, this diode is the collector- across the bottom resistor in the bias As mentioned, X3 in the R-6C2 ver- base junction of a transistor. It's used network, resulting in very low base and sion gets its bias from X2's emitter. This here as a temperature-sensitive bias emitter voltages. means that the AGC controls both IF supply that matches the base-emitter The most common causes for open stages. By contrast, the R-6C1 version characteristics of the output transis- collector connections are open-circuit simply uses fixed voltage-divider bias tors. It basically provides thermal loads (especially inductors and trans- for X3 and so its resistance to overload compensation for the push-pull output formers), bad solder joints and bad isn't as good. stage and is a mark of good design by socket connections. In this case, the The third IF transformer (L5) feeds Sony. Any number of other manufac- set came good after a few sharp taps on demodulator Dl. In the R-6C2 circuit, turers were still struggling with less- transistor X2, indicating that its collec- the AGC return is via R14 and volume effective fixed/adjustable bias schemes tor was going open circuit inside the control R1 to X2's emitter (and X3's or complex thermistor-compensated can (probably between the collector base). Alternatively, in the R-6C1, the bias circuits. lead wire and the germanium slice). AGC return goes to the emitter of X2 X5 & X6 drive output transformer T2 I needed a replacement transistor 84 SILICON CHIP siliconchip.com.au and a search through my trusty junk- box soon yielded a Sony 2SC73 (a germanium NPN). This transistor has a bandwidth (Ft) of 8MHz as opposed to the 2T524's 2.5MHz, so I expected to get more gain with the new transis- tor. This was subsequently proven to be correct. During my initial tests, I found that the audio section needed many tens of millivolts to produce an output, so electrolytics C3 & C5 were replaced. This immediately brought the audio gain up to expectations. As an aside, using electrolytic ca- pacitors for IF/RF bypassing is now considered poor design and as noted above, the set's initial violent oscilla- tion problems were cured by replacing Cl & C2. Electrolytics exhibit consider- Several parts are also mounted on the underside of the PCB, as shown in this able series resistance and inductance, photo. The old TR-63 was easy to restore to full working order. restricting their effectiveness to audio frequencies. Common practice would low-battery performance. With a 4.5V The Totsuko now be to shunt C1 & C2 with disc supply, the set begins to clip at only "stamp" is ceramic capacitors to ensure effective 5mW and its THD at 4mW is around moulded into RF bypassing. 7%, with little sign of crossover dis- the TR-63's rear tortion. Admittedly, 5mW isn't much cover. How good is it? but the set is still working perfectly Looking at the circuit and its build when the battery is down to 4.5V. Of quality, the TR-63 appears to be a well- course, it's only delivering one-quarter engineered set but how well does it the power output at half supply but it's perfectly adequate on the bench in perform? To find out, I decided to make its low-battery performance is excel- my 130m2 shed. some basic sensitivity and distortion lent and anticipates sets such as the measurements. Pye Jetliner. Would I buy another? As shown on the circuit, a 10pF ca- Because the set begins to clip at Would I buy another one? The an- pacitor is connected to the top of the around 20mW output, the following swer is "yes" if an R-6C1 version be- ferrite rod. As this set uses base injec- RF/IF measurements have been taken came available as I'd be interested to tion for the local oscillator, connect- at 10mW output. The RF bandwidth is compare it's AGC action against my ing my signal generator to X1's base ±3.7kHz at -3dB and ±55 kHz at -60dB. current R-6C2 version. stopped it dead for broadcast-band The AGC is quite effective and limits Finally, is it possible to "hot up" signals. I was able to get IF sensitivity the output to an increase of just 6dB an old set with better-performing RF/ readings but no RF readings. in response to a 26dB signal increase. converter and IF transistors? Sure but The simplest way around this was The received signal performance that's not the point. Repair necessities to connect the generator via the 10pF is quite good, though with poor S/N aside, these are old radios and it's best capacitor. This gives reliable results but ratios. At full gain, for 10mW output, to keep them in original condition. it doesn't give the actual signal voltage my modified TR-63 needs 2001.1V/m required at the converter's base connec- at 600kHz and 1100//m at 1400kHz. Further Reading tion, as would usually be specified. The However, both these figures result in (1) Many online sites describe the set did, however, respond correctly to an S/N ratio of only about 5dB. TR-63. For a thorough description, a direct IF injection, so I've given this The set's early AGC detracts from see James J. Butters' fine site at: http:// result as it's a better guide to the set's the 20dB S/N ratio figures, so I've www.jamesbutters.com/sonytr63.htm sensitivity and will help in diagnosing opted for 15dB. This demands an in- (2) For a tear-down and description: low-gain faults. put of around 700µV/m at 600kHz https://www.ifixit.com/Teardown/ Dealing with the audio stage first, the and 500µV/m at 1400kHz. In this set, Sony+TR-63+Transistor+Radio+Tea TR-63 goes into clipping at 20mW, with however, I had replaced transistor X2 rdown/1219 a THD of 8.5%. At 10mW, its distortion with a higher-performing substitute (3) A photo catalog is at: https://www. is 6% and the -3dB frequency response (as mentioned), so you can expect an flickr.com/photos/transistor_radios/ from volume control to speaker is unmodified TR-63 to have around half sets/72157603555111543/ 290Hz - 5.9kHz, with a peak at around the above sensitivity figures. (4) Ernst Erb's Radio Museum: http:// 1.3kHz. From antenna to speaker, it's With only 20mW of audio output at www.radiomuseum.org/r/sony_tr63_ 290-3000Hz. clipping, is it good enough? The an- tr_63_tr_63.html (6C1, 6C2) and http:// The diode biasing circuit used in swer is that while you'd need to use www.radiomuseum.org/r/sony_tran- the output stage contributes to the the plug-in earphone at the football, sistorLtr_63a.html (6C1) SC siliconchip.com.au JANUARY 2016 85