Fullscreen HTML5Med Res (??MB)HTML4

Vintage Radio Revisting the Zenith Royal 500 AM Transistor Radio By Ian Batty One year after the release of the groundbreaking Regency TR-1 (shown on the left), a major manufacturer entered the market with a new set: Zenith’s Royal 500 – the ‘peak of technology’. Z enith was co-founded in 1918 by two amateur radio operators, Ralph Matthews and Karl Hassel. They adopted their 9ZN call sign, transforming it to “ZN’th”. Joined in 1923 by Eugene F. McDonald, they formalised the name as Zenith, the astronomical term for the highest point overhead. It was a bold move, but the company became famous for high-quality radios and innovation. Zenith offered their first portable radio in 1924, their first mass-produced AC radio in 1926 siliconchip.com.au and pushbutton tuning in 1927. Their self-contained car radios in the 1930s needed no external batteries or generators. Zenith’s purchase of the Heath electronics company in 1979 saw them enter the computer market as Zenith Data Systems. This was a transition from their declining radio business, which they finally left in 1982. Zenith’s slogan, “The quality goes in before the name goes on”, was well-justified. After around 1995, they Australia's electronics magazine were taken over by Korean manufacturer LG and finally filed for bankruptcy in 1999. Design The Royal 500 is larger than the first “trannie”, Regency’s TR-1. However, the TR-1 was forced to be a simplified design due to being first to market and over cost considerations. Released for sale in November 1954, the TR-1 announced a new era in personal radios. November 2024  103 Table 1 – Zenith Royal 500 differences between models Model Date Construction Transistors Types Unilateralisation AGC 7XT40 cct 1 ‘55/56 Handwired All NPN 2N94, 2N94, 2N94, 2N94, 2N35, 2 x 2N35 IF1, IF2 IF1 only 7XT40 cct 2 ‘55/56 Handwired All NPN 2N193, 2N194, 2N216, 2N216, 2N35, 2 x 2N35 IF1, IF2 IF1 only 7XT40Z ‘55/56 Handwired All PNP 121-9, 121-14, 121-10, 121-10, 121-11, 2 x 121-12 IF1, IF2 IF1 only 7XT40Z1 ‘56/57 Handwired NPN, PNP 121-15, 121-16, 121-17, 121-17, 121-18, 2 x 121-19 IF1 only IF1 only 7ZT40 ‘56/57 PCB All NPN 2N193, 2N194, 2N216, 2N216, 2N35, 2 x 2N35 IF1, IF2 IF1 only 7ZT40Z1 ‘56/57 PCB NPN, PNP 121-15, 121-16, 121-17, 121-17, 121-18, 2 x 121-19 IF1 only IF1, IF2 7Z40Z ‘57/58 PCB All PNP 121-9, 121-14, 121-10, 121-10, 121-11, 2 x 121-12 IF1, IF2 IF1 only 7ZT40 revised ‘57/58 PCB All NPN 2N193, 2N194, 2N216, 2N216, 2N35, 2 x 2N35 IF1, IF2 IF1 only By the time the Royal 500 hit the market, we’d had a year to get used to the miniature marvel of the transistor radio. The Royal 500 benefited from that acceptance, and Zenith was a well-known and respected brand at the time. The Royal 500, using the full five stages we now accept as necessary for good performance, was always going to be larger than the TR-1. The use of standard ‘penlight’ (AA) cells also increased its final size. At about 480cc in volume, the Royal 500 is considerably larger than the 305cc TR-1. However, considering the extra components Zenith used, the Royal 500 is genuinely compact. The TR-1’s ergonomic design is sound, with the large tuning dial easily operated by fingertip. The Royal 500, in contrast, demands that you use your finger and thumb to grasp the direct-drive tuning knob – it is doable, but nowhere near as easy, accurate or elegant as the TR-1’s dial. Dr Hugo Holden described a later version of the Royal 500 in May 2018 (siliconchip.au/Article/11076). His version has the tuning knob supplemented by a small coaxial ‘button’ knob. This operates an epicyclic reduction drive, making accurate tuning easier. Device Engineering Council (JEDEC). The review version is built on a metal chassis with point-to-point wiring and a few terminal points. While there’s better access to components than on the PCB-based versions, the assembly is tight, with six capacitors mounted above the chassis, connecting to the underside circuit via small holes in the metal chassis plate. Such an assembly would have been more time-consuming than (for example) the 1957/58 7ZT40 PCB version that followed. PCB construction was retained for all subsequent models. These have resistors mounted on end, with one end lead exposed to help with testing. Hopefully, they’re the ‘active’ ends, not supply or ground. Transistor types The Royal 500 RF/IF section uses grown-junction NPN transistors, while this 7ZT40Z1 version uses PNP transistors in the audio section. They would also be grown-junction types, based on the early release date of this radio. The TO-22 (Transistor Outline 22) style can, shown in Photo 1, was necessitated by the long ‘sliver’ construction of grown-junction devices. This is a clue to the type of construction, as alloyed-junction types are commonly enclosed in cylindrical cases, Releases RadioMuseum lists eight 7X/7Z models in the original case, manufactured from 1955 to 1958. The product line continued until 1965, using the same “Owl Eye” form. There were four initial releases for the Royal 500: 7ZT40 and 7ZT40Z1, and the second production 7ZT40/7ZT40Z1 – see Table 1. The set I’m reviewing is a 7ZT40Z1. The 121-series transistors used are Zenith’s own part numbers; 2N-­series codes were registered with the electronics industry’s Joint Electron 104 Silicon Chip Photo 1 (inset, highlighted in yellow): one of the TO-22 PNP transistors used in the 7ZT40Z1 version of the Zenith Royal 500. Photo 2: the complete set for the Zenith Royal 500 with the radio, case, instruction manuals and listening earbud. Australia's electronics magazine siliconchip.com.au such as the OC44/45 and the improved AF116/117 series. The transistors are socketed, so they are easily removed for testing or replacement. The sockets do not conform precisely to TO-22 spacings, as shown by the bending of the transistor leads to fit the socket in Photo 1. The 7ZT40-R2 version of the Royal 500 uses NPN types throughout, with the audio section’s 2N35s also using grown-junction technology. Circuit details The original circuit diagram is well laid out, with one oddity: while capacitors are numbered (C1, C2 etc), resistors are not. In Fig.1, I have kept Zenith’s capacitor numbering to prevent confusion and have added resistor and transistor numbering for clarity. Signal pickup is via a rectangular ferrite rod antenna. It’s tuned by C1a, the antenna section of the gang, with a separate secondary winding on the rod supplying the signal to the converter. There are no external antenna/Earth connections. The converter stage uses separate excitation. NPN local oscillator (LO) transistor Q2 injects the LO signal to the base of NPN converter transistor Q1. Separate excitation allows the designers to apply automatic gain control (AGC) to the converter. Circuit measurements show that the converter has just 0.03V (30mV) of standing bias, so this transistor is very much in class-B operation. With full AGC, Q1’s base voltage goes to zero. While it might seem that this would cut the transistor off completely, its emitter voltage in these conditions is about +0.12V, showing that it is still drawing current. This is explained by the applied LO signal of about 700mV peak-to-peak. This means that, even with the base bias at zero, Q2 still swings between being in cutoff and conduction by the alternating negative- and positive-­ going parts of the LO signal. The LO itself also works with a slight DC bias of only about 0.1V. This measurement obscures the oscillator’s action; the base signal is some 0.5V peak-to-peak, confirming it operates mainly in class B. The oscillator’s collector connects to pin 6 of LO autotransformer T6. Its pin 4 ‘cold’ tap goes to the positive supply. This allows the transformer to provide phase reversal, making this a modified Hartley circuit. It was here that I discovered an error in the original circuit diagram. Their diagram shows pin 1 of T6 going to the end of the winding, while pin 2 connects to the next tap along. However, on testing my set, I found more signal (700mV peak-to-peak) at pin 2 than at pin 1 (500mV peak-to-peak). That means that pin 2 actually connects to the end of the winding and pin 1 to the tap, as shown on my corrected circuit diagram. If you need to test the LO section of one of these sets, you should carefully check whether Fig.1: the redrawn circuit diagram for the Zenith Royal 500 (version 7XT40Z1). The circled numbers are voltage readings taken at various points in the circuit. your set is like mine or matches the original circuit, with the internal connections to pins 1 & 2 of T6 swapped compared to mine. The LO circuit is tuned by the tuning gang’s C1d connection to T6’s pin 3. The Royal 500 uses a cut plate design, so there is no padder capacitor. Both the incoming signal and the LO signal are applied to the converter base. I found that this prevented direct measurements at Q1’s base at any frequency other than the IF. There is a workaround, which I will describe in the performance section below. The converter feeds the untapped, tuned primary of the first intermediate frequency (IF) transformer, T1. T1’s untuned secondary feeds the 455kHz IF signal to the first IF amplifier transistor, Q3 (NPN). This transistor is stabilised by unilateralisation network R10/C7, which compensates for the high collector-base feedback of early devices. Emitter resistor R11, bypassed by capacitor C5, provides DC stabilisation. Q3’s collector feeds second IF transformer T2’s untapped, tuned primary. T2’s untuned secondary feeds second IF amplifier transistor Q4 (NPN), which lacks unilateralisation. This stage is biased from the emitter circuit of Q3. DC stabilisation is provided by emitter resistor R15, bypassed by capacitor C9. Q4’s collector feeds the untapped, tuned primary of third IF transformer T3 and T3’s secondary feeds demodulator/AGC diode D1. The original diagrams for all versions except the initial 7XT40Z have D1’s anode and cathode reversed. This error is confirmed by theory, inspection and circuit action. As they showed it, it would have reverse bias applied rather than the weak forward bias universally applied in such circuits. It would not demodulate, nor would it generate a suitable AGC voltage. D1 is loaded by 5kW volume control potentiometer VR1, while 50nF capacitor C11 forms a low-pass filter with its resistance to remove the IF component of the signal. D1’s output also feeds 4.7kW resistor R17, which conveys D1’s DC output to the AGC line. The AGC line is biased weakly positive by 47kW resistor R9. This provides a slight forward bias for D1, improving its sensitivity, plus a standing bias for the converter and the first IF amplifier transistor. Automatic gain control (AGC) The review set is labelled as 7XT40Z1. According to the circuit diagram, that version applies AGC to the Volume Control 2nd IF Oscillator 1st IF Audio output stage 3rd IFT 2nd IFT 1st Audio Osc. Coil 1st IFT Output Converter Output Demod Driver Transformer Output Transformer Photos 3 & 4: annotated top and underside views of the Zenith Royal 500’s chassis. 106 Silicon Chip first IF stage alone. However, this set’s AGC circuit applies control to both IF stages. It appears to be an undocumented factory variation. The AGC line is bypassed for audio by 3μF capacitor C16. This is generally frowned on, as electrolytics perform poorly at intermediate and radio frequencies. The Regency TR-1 I tested in April 2013 showed RF instability due to such a capacitor having aged (siliconchip.au/Article/3761). The AGC voltage is applied to the converter stage via 1kW series resistor R6, which also isolates the LO signal from the AGC circuit. The first IF transistor (Q3) has the AGC voltage applied via 2.2kW decoupling resistor R7. Recall that the second IF transistor (Q4) is biased from Q3’s 470W emitter resistor (R11) via 2.2kW resistor R13. As the AGC circuit reduces the bias on Q3 (also reducing its emitter current), the voltage across Q3’s emitter resistor, R11, will fall. Full AGC action brings Q3’s bias close to cutoff, with a bias of about 0.22V, so its emitter voltage will fall to only about 0.05V (50mV), implying a collector current of 100μA. A drop of only 0.05V across R11 would reduce Q4’s available bias to zero, but 47kW resistor R14 from the positive supply rail ensures that Q4’s bias never goes below the cutoff threshold. In effect, AGC is applied to the converter and both IF amplifiers in this radio. See the voltage annotations on the circuit for the actual operating values. Australia's electronics magazine The audio section follows the design that had become standard at about this time. Like many other circuits, it uses PNP transistors with a positive battery supply. This sees the emitters fed from the positive supply and collectors going (via their loads) to ground. PNP driver transistor Q5 uses fixed combination bias: R18 & R19 form the divider, while R20 is the emitter resistor, bypassed by 50μF capacitor C13. Q5 feeds driver transformer T4, which has a split secondary that provides anti-phase drive to the Class-B PNP output transistor pair, Q6/Q7. 1nF capacitor C14 is wired across T4’s primary. This looks like it would provide a top-cut function, but its low value means it will have no effect until about 15kHz. It’s most likely there to siliconchip.com.au filter out any remaining IF signal that C11 did not remove. The output pair gets about 150mV of forward bias from divider R22/R23. This bias network is not compensated for temperature or changing battery voltage. The lack of temperature compensation makes it inadvisable to run the output stage at full sinewave power for any length of time. The two emitters share a common 10W resistor, R24, which provides some local feedback and helps compensate for mismatches in Q6/Q7. They drive output transformer T5 which, in turn, drives the 15W internal speaker. 100nF capacitor C15 does have a top-cut effect. Photo 5: a close-up of the front panel controls on the Zenith Royal 500. The left control handles volume and power, while the right is for tuning, audio stages (preamplifier, driver, output). The Royal 500 uses the more familiar two IF stages and two audio Cleaning up the set stages (driver and output). I was offered this set at the HRSA’s As noted earlier, designs applying RadioFest in September 2023. It was both the LO and signal to the converter complete, including the original ear- base do not allow a signal generator, phone, leather and cloth carry cases, with its low output impedance, to the original handbook, and even a spe- inject a signal directly into the base. I cial (unused) label allowing the owner was able to measure its sensitivity for to ‘personalise’ the set. a direct 455kHz input but not for anyCollectors will appreciate the rarity thing in the broadcast band. I solved of getting any old radio complete with this problem by adding a 470W series all accessories, so thank you to the own- resistor between my generator and the ers who kept this set complete as pur- converter base. chased! It also had a receipt for repair Comparing my direct 455kHz injecwork at Truscott’s, dated 14/11/2001. tion at 8μV and the modified input at The set was pretty much undisturbed, 110μV, I have an attenuation of 13.75 apart from a professional recap. times. Assuming that ratio holds, senThe case showed signs of wear, sitivity at the converter base is about mostly affecting the front gold-­ 135μV/13.75, ie, 10μV at 455kHz, and coloured ZENITH branding, the rear about 6μV at 1260kHz. These values set name and the “tubeless – 7 transis- are roughly comparable to the sensitivtors” moulding. It came with batteries ities of other five-stage sets I’ve tested. and worked at first switch-on. Its sensitivity is superior to the The volume control and tuning were T-2500; 3WV Horsham rocked in noisy, so I applied contact cleaner with pretty much at local station volume. good results. The alignment seemed I tried getting some Hobart and SydOK but I went over it to be sure. I ney stations during the day, but either found that Zenith’s suggestion of using it could not pick them up or adjacent 535kHz for the low end did not give the Melbourne stations blanked them best results, so I aligned it at 600kHz. out. I did manage to pick up 3BT BalZenith also specified aligning the top larat and 3EL Maryborough, while end at only 1260kHz, and I followed 3CS Colac treated me to some vintage their recommendation. Fleetwood Mac! With the set working well, I put it The Royal 500’s service sheets on the test bench for evaluation. give a sensitivity of “approximately 500μV/m for 50 milliwatts output”. How good is it? On test, sensitivity at 600kHz was For sensitivity, it’s up there with the 120μV/m at 600kHz and 115μV/m best of the day. It’s much smaller than at 1260kHz. The signal-plus-noiseRaytheon’s ‘picnic set’ T-2500, which to-noise (S+N:N) ratios were 11dB in it rivals in all but sound quality. both cases. These were for a modulaBoth sets use five active stages and tion frequency of 1kHz; sensitivities a separately-excited converter stage. for 400Hz were, unusually, worse by The T-2500 uses only one IF amplifier some 2dB. but makes up for that by having three For the standard 20dB S+N:N siliconchip.com.au Australia's electronics magazine ratio, sensitivities were 325μV/m and 400μV/m, confirming Zenith’s original specifications. RF bandwidth was ±2.5kHz at -3dB and ±30.5kHz at -60dB. The AGC was effective, with a +40dB change of input needed to give a +6dB output change. I was not able to force it into overload. The audio response was perhaps adequate, given the small speaker. From the volume control to the speaker, it reached -3dB at about 320Hz and 3.9kHz; from antenna to speaker, it was around 260~2700Hz. Distortion at 50mW was 6.4% and the output started to clip at 100mW. The output sinewave was visibly asymmetric at low volume, indicating a mismatch between the two output transistors. With no feedback in the audio section, this radio does depend on output transistor matching for best performance. Special handling It’s well-built, but be aware that the ferrite antenna bar is fixed to the chassis by a semi-flexible clamp. Mine was still intact, but I’d be careful about applying too much stress. Purchasing advice The Royal 500 was released in several colours. For me, the ‘black brick’ design is most appealing. It’s striking, but not as ‘shouty’ as the cherry red version. If you don’t have one, consider this fine example of radio technology. It’s a ‘proper trannie’ with all the design features and performance you’d expect, and it runs on ordinary AA cells. For more information, see: • https://w.wiki/8$Z4 • Search www.radiomuseum.org/ for Zenith Royal 500 SC November 2024  107