The Larkspur R210 Receiver from Colin MacKinnon VK2DYM's WEB site
THE R210 H.F. BAND COMMUNICATION RECEIVER.
The R210 communications receiver was developed by the A. T. and E. company as part of the UK "LARKSPUR" military radio communications equipment, although they were offered for commercial sale for use where reliable long distance communications was needed.
For further information on LARKSPUR visit: http://www.wftw.nl/larkspur.html
The article that follows provides the background to the design and manufacture of the R210 and was written by J. A. Knight, A. T. and E. (Bridgnorth) Ltd.
This article describes a hermetically sealed receiver covering the range from 2 to 16 Mc/s. An accurately stabilised oscillator combined with a 52-inch film scale makes it possible to select any 10 kc/s channel without previous ‘netting’. It is powered by an internal 24 V vibrator supply unit, or alternatively from an external a.c. unit.
The R 210 receiver, although originally developed for service use, has many features that make it suitable for general communication use in the h.f. band. It is compact, robust, and accessible. It receives speech modulation and telegraphy (c.w. and c.f.s.). The tuning is simple and so accurate that slow and tiresome netting procedures are eliminated. The set is suitable for operation in any part of the world.
As a result of dividing the total frequency coverage into seven bands, each 10 kc/s channel on the film scale is separated by almost a quarter of an inch. A mechanical law corrector helps to keep the channel spacing over the whole band almost linear.
Great care has been taken in stabilising the oscillator and, with separately compensated bands, the frequency drift compares favourably with that of crystal-controlled oscillators. The specially selected variable tuning capacitor and the sealing of the receiver with dry air also contribute to a thermal drift of less than 0002% per C.
To make the receiver easily accessible for service it is divided into detachable units. The main chassis hinge open, leaving the underside unobstructed but still operational . In the working position the valves and electrolytic capacitors are distributed along outside faces for efficient cooling. The 42 r.f. trimmers are also readily accessible.
The receiver is built into a robust aluminium alloy die-cast panel and case. The panel is 10 in. wide by 7.1 in. high, with a raised edge protecting the controls. The case is 14 in. deep and has cast fins to assist cooling. To avoid possible damage to the scale the dial window is of armoured glass.
The panel controls are as follows:
1. Frequency selector with cranked handle.
2. Dial lock for 1.
3. Frequency band switch, seven positions.
4. Gain control.
5. Beat-frequency pitch control.
6. System switch to select:
Automatic gain control.
7. Dial cursor adjustment.
8. On-off switch (with or without dial light).
9. Noise limiter on-off switch.
The panel also carries the following:
2 Aerial terminal for high-impedance rod
3 Coaxial socket i.f. output for carrier
4 12-way input and output socket. This provides the 24 V input and telephone output. It also feeds H.T. and L.T. for intercom, and isolates these when fed from an external a.c. supply.
5 A fuse in the 24 V input circuit is also panel mounted.
The rear of the case has a screw-in desiccator and dry-air plug for circulating dry air before sealing.
Front panel of R210
The receiver has 15 valves and a vibrator:
1. R.f. amplifier CV131
2. Mixer CV453
3. Oscillator CV850
4. 1st i.f. amplifier CV131
5. 2nd i.f. amplifier CV131
6. 3rd if. amplifier CV131
7. Signal detector and a.g.c. diode CV140
8. Audio amplifier CV850
9. Audio output CV850
10. Double-diode noise limiter CV140
11. Oscillator for 100 kc/s crystal calibrator CV850
12. Double-triode locked 10 kc/s multivibrator CV491
13. Beat-frequency oscillator CV850
14. Cathode-follower for i.f. output CV850
15. Neon stabiliser for oscillator h.t. CV286
16. Synchronous vibrator X331C (now X354C)
The sub-assemblies of the R210 are hinged so that they can be swung out for servicing. The Power Supply and Audio/Calibrator units have been raised up on the left hand side of this illustration whilst the IF module has been swung back on the right hand side. The modules left in place are the RF and Tuning sections on the lower right side. Sorry about the poor illustration, it's the best I have at the moment
Signal/Noise Ratio: A signal of 5.7/µV into an 80-ohm dummy aerial will produce a 20 dB signal/noise ratio.
Noise Factor: Better than 6 dB.
Selectivity: + 3 kc/s not more than - 6 dB + 8.4 kc/s at least - 60 dB.
Cut-off Slope: 10dB per kc/s.
Intermediate Frequency: 460 kc/s.
Output: 200mW with 50-ohm load, 70% modulation.
70 mW with 150-ohm load, 70% modulation.
60 mW with 50-ohm load, 30% modulation.
20 mW with 150-ohm load, 30% modulation.
Negative feedback automatically maintains the correct output in each head-set according to the number, up to three, in use at any given time.
Calibration Check: A built-in crystal calibrator gives a calibration check at 100-kc/s and also at l0-kc/s points. The cursor is adjustable to correct dial readings.
Automatic Gain Control: Output variation not more than 6 dB when input is varied from 10 µV to 100 mV.
Modules of the R210. Top
left is the IF unit, lower left the RF and tuning unit, whilst on the
right are the Audio and Crystal Calibrator and Power Supply Units.
The heart of a communication receiver is the r.f. unit. This combined with the dial presentation that is its main link with the operator, will be considered in detail.
During the early stages of development a double superhet using seven crystals and a second oscillator with a 1 Mc/s swing was considered. It was found difficult to design this for the space available and at the same time reduce the spurious responses to a minimum.
Development then turned to a stable free-running oscillator covering 2.5 to 16.5 Mc/s. Investigation into the temperature coefficient of various variable capacitors showed that one commercially available type gave a small and consistent positive drift with increase in temperature of approximately 5 to 10 parts per million.
It is interesting to note that this is a straight-line capacitance unit with the rotor plates forming a true semicircle about the driving shaft. There is no unbalanced overhang of the plates and no bonding strip, which in some capacitors is of dissimilar material. The capacitor is all brass with a simple ball race at each end, and it has the added advantage of being quite small. To keep the rotor bearings as near to the oscillator as possible it was decided to use a single gang for the oscillator, coupled to a separate two-gang for the aerial and mixer section.
By splitting the frequency coverage into seven bands it was possible to use a reasonable amount of parallel capacitance in addition to the trimmers and switch 'strays'. Providing this capacitance has a consistent temperature coefficient it adds to the stability, by swamping valve and other capacitances.
To obtain consistent results all the compensating capacitors had to be in one unit of say -30 parts per million per 0C. This gave better results than a large mica capacitor of +5 to +50 parts per million shunted with a small capacitance of large negative temperature coefficient, say -750 parts per million.
This receiver uses –150 + 20 parts per million per C. capacitors to compensate the three low-frequency coils. The four high-frequency coils have in addition –30 + 20 parts per million per 0C. capacitors. On the h.f. band, which in practice appeared to be the least difficult to compensate, the parallel capacitor is 180 pF.
The third important factor is the oscillator coil itself. Not only had it to be a stable inductance, but also a number of coils had to fit neatly round the waveband switch, with trimmers in accessible positions. The coil-former required tags to rigidly support both the wires of its own turns and those to the switch. The capacitance trimmer also had to be supported.
In many receivers the r.f. coils are designed to screw on a metal dividing screen, with their tags passing through it. Since it is quite unnecessary to screen the coil from its own switching circuit this only adds to the stray capacitance. Any movement of the wires near to this screen will cause extra uncontrolled drift.
The coil-former has an extended base, which carries both the tags and capacitance trimmer. Two tapped holes in the side of this base mount it so that the axis of the coil is parallel to the screen. The capacitance trimmer mounts through a hole in the base, and solders into a raised earth tag underneath.
It is possible to mount seven of these formers round an Oak type of switch with short and rigid connections. The variable tuning capacitor mounts between the valve and the h.f. band coil.
The coil-former is moulded in Nylon-loaded Bakelite, and has the anchor ring that has been incorporated in coil-formers made at Bridgnorth for the past ten years. This consists of a washer punched in high-grade Bakelite sheet, with a projecting key that locates in a slot moulded into the former. By passing wires through holes in this ring it is possible to secure it at any point on the former and obtain a good winding tension with no fear of it slipping when stoved for drying and varnishing. The four holes in this ring are always in line with the tags moulded in the former so that the wires go directly from the tapping point over the ring to the tags. This keeps the coupling and stray capacitance consistent. A thin version of the ring is used for tapping to the middle of a coil, and another (normal) one for terminating it.
The coil-former moulding tool has three inserts. This allows for the outside to be threaded at a suitable pitch to receive the different gauges of wire and give the Q required at various frequencies.
Bands 6 and 7 are threaded at 16 turns per inch and have a positive temperature coefficient of 12 parts per million per 0C. Bands 4 and 5 are threaded at 32 turns per inch and have a positive temperature coefficient of 26 parts per million per 0C. Bands 1, 2, and 3 use a plain former and have a positive temperature coefficient of 55 parts per million per 0C.
All the coils are adjusted with a 6 mm threaded dust-core. They have concentric-type trimmers.
The aerial, mixer, and oscillator sections are each mounted on a common metal pressing, which also forms the valve platform. The individual sections are wired before being fitted with gang mounting bracket and screens. These three right-angle cross braces make the unit into a very rigid box section.
The aerial coils have auto windings wound on the grid end of the tuned winding to give additional voltage step-up to overcome first circuit valve noise.
The Bandswitched RF,
Oscillator and Mixer sections.
To take full advantage of a wideband stable oscillator it is necessary to have a drive and dial unit with high resetting accuracy. To achieve this the film unit was designed.
Five of the bands have 230 l0-kc/s channels for a rotation of the variable capacitor of approximately l550. This represents about 40' of rotation per channel.
Since each 10 kc/s channel had to be accurately set on the dial it was decided to separate each one by 0.2 in. This required a film 52 in. long, allowing a few inches for start and finish. To find sufficient room for seven bands a perforated film 70 mm wide was used.
The film is contained on two spools, which are spring loaded against each other. This is achieved by a coil spring mounted in the spool and connecting it with its spindle. The two spindles are coupled with gears. The film also passes over a sprocket that is driven by the tuning knob through a bevel gear. The drive to the gang shaft is via a pair of sprung anti-backlash gears and a worm wheel and sprung pinion. The two main shafts rotate on single balls held in spring tension by their adjustable ball cups. The film is prevented from travelling too far by a 14-turn tab washer stop, which brings the overload clutch into operation.
The film is printed in black and passes over a black backplate. The band indicator passes between these, and consists of a narrow white strip moving up and down as the waveband switch is turned. Only the required band is visible whether the dial lamp is on or off.
Besides the main tuning and band-changing knobs the film unit also has a dial lock and cursor adjustment (for accurate adjustment against the crystal calibrator). The four shafts carry washers to seal them to bush and panel.
At the rear of the unit a law corrector couples it with the main tuning capacitor. The connector consists of two offset spindles. A projecting pin from an arm on one spindle moves up and down a slot in spring-loaded plates on the other. This gives a variable ratio from one end of the dial to the other. At the high-frequency end the speed of gang capacitor rotation is reduced by a ratio of 10 to 13. At the low-frequency end it is increased by 10 to 7. This helps to produce a straight-line frequency scale, while at the same time retaining the stability advantages of a straight-line capacitance gang.
This unit consists of two L-shaped chassis mounted back to back between end plates. Besides being a detachable unit it will also hinge open for maintenance or repair while still remaining operative. The one chassis carries the three i.f. amplifiers with their associated transformers. The second chassis contains the final i.f. transformer with double diode beat-frequency oscillator and coil, cathode follower for i.f. output (c.f.s.) and the double-diode noise limiter. A separate tag panel is mounted on the base of the unit for terminating the main cable-form. This can easily be disconnected by a series of soldered link connections to the i.f. tag panel.
Electrically the i.f. unit is conventional. The four transformers are tuned with 850-pF capacitors in order to reduce the effect of valve capacitance to a minimum.
The underside of the IF
module. Two L shaped chassis are mounted back to back.
The audio unit is mounted with angle strips to the power-supply unit, from which it is easily detachable. Besides the two audio amplifiers it also contains the crystal calibrator, oscillator and multivibrator. The two ferrite inductors tune the 1000 c/s filter and the 10 kc/s crystal-locked multivibrator.
The system switch is arranged so that extra attenuation is introduced on the 100 kc/s calibrate position. The calibration signal on both this and the 10 kc/s position has had its level carefully chosen to silence the noise, but at the same time not to cause excessive second-channel whistles.
The aerial inputs are switched over to earth in the calibrate position. The 100 kc/s crystal has a trimmer capacitor to adjust it accurately to frequency.
By removing the protection plate on the rear of the unit the tag panel folds down to make components more accessible.
The underside of the Audio module and Crystal Calibrator (on the left) and Power Supply module (right).
INTERNAL DC POWER SUPPLY UNIT.
This unit is built into a box-type chassis with a separate filter box built into one corner to carry the vibrator. This permits most of the vibrator's spurious frequencies to be removed before reaching the transformer. Very suitable filters with low d.c. resistance can be made by passing ferrite beads over the connecting wires. Their losses at high frequencies are used to good advantage.
The component panel in the filter box may be removed on long leads for maintenance.
The X 354C vibrator is rated at 80 watts, and as the total load of this receiver is about 40 watts it was decided to use indirectly-heated valves. These are supplied from the vibrator transformer as in a standard a.c. receiver. This makes it possible to regulate against varying battery voltages by the usual tapped primary.
The C11 transmitter, which is designed to work with this receiver, contains a voltage sensitive relay. This operates mid-way between 21 and 29 volts, and energises a slave relay in the receiver that changes over the input on the power transformer.* This automatically selects the correct tap according to the state of the battery.
* Ammendment to Colin's artice. The voltage control relay is in the SUTR, not the C11.
We wish to acknowledge and thank S.R.D.E., Christchurch, UK, for whom this set was developed, for permitting the publication of this article.
Please note, this article is copyrighted.
and text reproduced here with
kind permission of Ray Robinson VK2NO