Collins S/Line Influence on Heathkit's SB-Series

By Chris Codella, W2PA

July 2008

When designing the SB Series of ham equipment in the mid 1960s, the engineers at Heathkit took inspiration from the Collins S/Line, first introduced eight years earlier.  While there may not be documentation to support this assertion – I haven’t seen any myself – even a cursory comparison of the physical and electrical designs leads clearly to that conclusion.  It must have been an interesting and challenging process to take the Collins design, widely considered the best of its generation, and adapt it to design a kit with many of the same features at a greatly reduced price that more hams could afford, by using less expensive (and therefore less robust, lower quality) mechanical parts and electrical components, and simplified construction methods. 

Physical and Mechanical Design

The most obvious place to start comparing designs is with the cabinets.  They are both wrap-around designs with rounded corners and a piano-hinged top lid.  Both chassis slide into the cabinet front and are fastened to the cabinet by screws that also hold the feet on.  In the S/Line there is one additional screw halfway between the two rear feet.  While the S/Line cabinet lid is flat, fits into an opening entirely on top of the cabinet and has a hinge you can just see on the top rear edge, the SB lid curves around to meet the sides, with the hinge hidden from view on the cabinet rear side.  The SB cabinet is all one piece (counting the hinge and top), whereas the S/line cabinet has a separate “trim ring” that frames the front panel and is held onto the cabinet by screws on top and the front feet on bottom. 

The two paint schemes differ in both color and texture.  The Collins paint is medium gray, has no texture and seems to be quite hard.  The trim ring is a slightly lighter gray.  The front panel is a dark gray with white lettering and has a bumpy texture to the panel itself (not the paint) except for around the controls where the white markings are printed on smooth, untextured islands.  The Heathkit paint is textured all over.  The cabinet is a light gray tinted slightly green that contrasts with the dark green front panel.  The front panel texture is finer grained than the cabinet allowing the control markings to be printed without the need for smooth islands like those on the Collins.

The S/Line front panel, shown edge-on in the photo below, consists of two aluminum plates: a rear one that is welded to the chassis and a front decorative textured one that is glued to the rear one (a bit of glue that has oozed out can be seen to the left).

The Heathkit front panel is a single aluminum plate held to the chassis by the lower row of control nuts and screws near the top that fasten to vertical chassis panels that slope down to the rear, just like on the Collins chassis.

Heathkit’s two sizes of plastic green SB knobs (large and small) are reminiscent of the S/Line with their large knurls and spun aluminum inserts – but the Collins knobs are black bakelite.  Collins added a third kind – the pointed selector knob – and Heathkit adds skirts to both kinds of its knobs.  Collins has concentric full sized knobs, such as those on the 75S-3 AF/RF gains and thumb knobs such as the loading control on the transmitters.  Heathkit only adopted the thumb knob.  The knurled Collins control knob appears both with and without a pointer, and, when the pointer is present, it sometimes is attached to the knob and sometimes moves independently, as with the preselector.  It also comes in two sizes, regular, for controls, and large, for the 30L-1 tuning capacitors. Lastly, the 32S-3 transmitter has a small knob for spotting. Thus, Collins has at least 10 different kinds of knob variations where Heath got it down to only 3 (4 if you count the two kinds of large knob: the one with the wide skirt for main tuning and the one with a narrow skirt with a pointer that is used on amplifiers).

Tuning knobs:

Control knobs:

The meter on the S/Line is metal, bakelite and glass, and is heavier than Heathkit’s mostly plastic meter.

Both mount in large round holes on the front panel and fasten with threaded bolts that go through separate holes.  The Collins meter is illuminated from the back and is painted white inside the body to reflect the light onto the front surface of the white opaque meter face.  Heathkit relied on the clear plastic of the meter body to conduct light shining from the back through translucent white markings on a black meter face.  Incidentally, the same company that made the meters for Heath also made them for Galaxy’s transceivers of the 1960s: Ideal Precision Meters, Inc., who still makes the same style panel meter today!

The main tuning dial on the SB-series borrows heavily from the S/Line design and its basic operation is copied directly.  The main tuning knob drives a shaft that turns a pair of washers behind the panel that pinch the inner rim of a large wheel that has kHz markings near its outer edge that show through a window in the escutcheon.  Since the knob shaft drives the inside edge of the kHz dial, the knob and dial turn in the same direction, and provide a gearing down of the turning rate.  The hub of the kHz dial is, in turn, mounted to a multi-turn VFO shaft which protrudes through a separate hole in the front panel from the knob shaft and ends nearly flush with the front panel escutcheon surface.  The VFO shaft has a slotted end to allow it to be turned through the front panel separately from the tuning knob shaft.  It covers 100 kHz per revolution in both versions.  But Collins and Heathkit have different ways of keeping track of 100 kHz segments.

Collins used an ingenious scheme that allows the same kHz dial to rotate twice with different numbers showing through from behind for each rotation.  The toothed outer edge of the clear kHz dial turns an idler gear attached to the inside of the front panel.   

S/Line Dial Mechanism Detail - Inside View

The idler, in turn, turns the toothed outer edge of the numbers dial mounted behind the kHz dial but free to rotate around the VFO shaft independently of the kHz dial (except for its being coupled to it via the idler gear).  Since the two dials have a different number of teeth but mesh with the same idler they turn at slightly different rates.  As this happens, the kHz dial masks a different set of numbers on each rotation – either 0 to 100, or 100 to 200.  The dial on the S/Line turns only twice, covering 200 kHz for each band switch position.

Heathkit, in an equally ingenious mechanism, uses a small slide rule dial at the top of the escutcheon to keep track of the 100’s of kHz.  A follower arm linked to the 100 kHz marker is driven by a small peg that rides inside a spiral groove on the rear of the main kHz dial.  As the dial revolves, the groove moves the peg further from the center of the dial, which, in turn, makes the 100’s marker move from left to right in the upper window.  The Heathkit dial rotates 5 times covering a 500 kHz wide segment for each band switch position. 

Heathkit seemed to be particularly proud of this dial mechanism.  Their 1967 catalog notes that it provides:

 "... tuning with bandspread equivalent to 10 feet per megahertz ... truly linear tuning over the entire receiver coverage ...  the good solid feeling of tuning a non-backlash vernier dial mechanism ... plus the ultimate in tuning repeatability that means real satisfaction in on-the-air performance."  

They may have thought this was a little over the top and trimmed it back to a single bulleted feature item in the next release of the catalog. 

Moving inside, the biggest difference, noticeable immediately, is that the S/Line has all point-to-point wiring and the SB-series uses PC boards, mostly to make the kit less costly and easier to build.  Both manufacturers use a color-coded wiring harness. 

Both the Collins and Heathkit transmitter final amplifier stages are isolated in an RF cage.  The S/Line has a safety interlock on the final cage that grounds the HV if the cover is removed; the SB-series has a red “Danger” sticker.

Most interconnection ports on both series use phono (RCA) jacks and cables.  Many hams scoff at the use of a phono jack for the RF connectors on the SB-series; but Heathkit was only copying the practice from Collins, who only used UHF and N connectors on their high power units.  On the S/Line transmitter and receiver, they are phono jacks.

Both Collins and Heathkit use separate power supplies for the transmitters and transceivers that connect using cables with 11-pin round, keyed connectors (similar to large octal tube bases and sockets) and could be mounted inside a speaker cabinet.  Collins had at least three other accessories in which their power supply could be mounted.  The chassis-mounted connector is male-type, so the pins stick out the back of the chassis.  However, Collins made the extra effort to use a recessed connector so these pins are protected.

Circuit Design and Electrical Construction

SB-300,301 and the 75S-1,2,3 receivers

Let’s start on the front panel. Aside from the converter selector switch, the SB-300 controls exactly reflect those on the 75S-1: AF and RF gains, function switch, AGC speed selection, mode, band, and preselector tune. The SB-300 completely lacks the rejection tuning of the later 75S-3 designs. Only the preselector controls look different - the one on the SB-300 doesn't have the band markings present on the 75S – a hint of one major design difference.

The two receivers are quite similar in their basic overall design. Both are double conversion, using a band switched, crystal controlled heterodyne oscillator to produce the first IF. In the SB-300 IF stages are at 8.395 MHz and 3.395 MHz, and in the 75S they are at 2.955 MHz and 455 kHz. In both receivers the VFO (LMO) is injected into the second mixer to produce the second IF where a crystal filter determines the receiver bandwidth – different filters are selected by the mode switch. Both receivers also use a crystal controlled BFO with a triode product detector on CW and SSB, and switch crystals with the mode switch to keep the carrier frequency constant when changing sidebands. Although the design influence is evident, many differences emerge as you look one level deeper.

To make the kit easier to build and less expensive, Heathkit sacrificed the general coverage capability of the 75S, designing the SB-300 to cover only five HF ham bands of the 1960s (80, 40, 20, 15, and 10 meters). In the Collins, a set of three inductors, one for each side of the RF amplifier and one for the heterodyne oscillator, have their tuning slugs linked together in a mechanical rack so they move in and out together controlled by a geared-down control shaft (and band-switched trimmer capacitors are used for alignment).  This scheme enables continuous coverage of virtually the entire HF spectrum (about 3 to 30 MHz) and is common across all S/Line units.  By using a ham-band-only design, Heathkit could instead use band-switched inductors (with their slugs used for alignment only) and a set of variable capacitors on a common control shaft for tuning. This accounts for the difference in the front panel preselector markings. On each band the SB-300 preselector control tunes near the middle of its range, whereas the 75S preselector tunes at a different spot for each band across its entire continuous coverage range.

Now let’s compare one stage at a time. Tube types appear in parentheses separated by a comma, with the SB-300 first.

Received signals from the antenna are first fed into a triode RF amplifier (6BZ6, 6DC6), tuned by the preselector. The first mixer (6AU6 tetrode, 6EA8 triode) uses the signal from the heterodyne oscillator (6AB4 triode, 6EA8 tetrode) and feeds an interstage transformer into the second mixer (6AU6 tetrode, 6EA8 tetrode), where the VFO (LMO-6CB6, VFO-6AU6) signal is injected.

Heathkit used a single inductor winding whereas Collins used two inductively coupled windings in all interstage transformers.

Both VFOs use a fixed capacitor and variable inductor to achieve linear tuning. The Heathkit LMO was factory built to match components for thermal characteristics to improve stability.

The crystal filters come next in the path, and in the 75S-3 these are followed by a two-triode Q-multiplier (12AX7) for rejection tuning. When AM is selected a different IF transformer having a 5kHz bandwidth is used and no crystal filter is in the path (although the 75S-3B and C added an extra filter socket that could be wired into the AM path if desired).

Both receivers use a two-stage IF amplifier (two 6BA6 tetrodes in both), and the S-meter indicates cathode current in the second stage. A transformer then provides the signal to the product detector (6AS11 triode, 6EA8 triode) and BFO (6AS11 triode and tetrode amp, 6EA8 tetrode). In the 75S-3, a tunable BFO is also supplied (6DC6 tetrode) so that the CW pitch could be varied. AM is detected using a single silicon diode in the SB-300 and in the 75S-3 a diode connection to the first audio amplifier (6AT6) is used. AGC voltage in the SB-300 is derived using two silicon diodes while in the 75S, a second diode connection to the first audio amplifier is used.

The detectors drive a two stage audio amplifier in both receivers (6HF8 triode and tetrode, 6AT6 triode and 6BF5 tetrode) followed by a tapped audio transformer for driving both speaker and headphones.

SB-400,401 and the 32S-1,2,3 transmitters

As with the receivers, the overall basic design is quite similar and uses the same mixing scheme in reverse order.

In both transmitters the carrier is generated by a crystal oscillator in which the frequency is changed according to the mode switch setting. It’s called a “carrier generator” in the SB-400 and uses a 6AV11 triode with another one as cathode follower, and in the 32S-3 it’s called a BFO and uses a 6U8A tetrode. The carrier is injected into the balanced modulator for SSB in both transmitters. But to generate a CW signal, the SB-400 simply unbalances the balanced modulator through the CW Level control, whereas in the 32S-3 the carrier is instead injected directly into the first mixer. (The 32S-1, however, creates the carrier entirely in the balanced modulator using the sidetone generator, which leads to problems with spurious signals being generated.  This is probably why the offset is much higher in the 32S-1 - an attempt to get rid of some of them.)

Next in the chain comes a transformer feeding an IF amplifier (6AU6, 6CB6) followed by a crystal bandpass filter at the first IF which in turn drives the first mixer (6EW8 tetrode, dual 12AT7 triodes) where the LMO or VFO signal is injected. The second IF signal is coupled through a bandpass transformer to the second mixer (6EW8, dual 12AT7) where the heterodyne oscillator signal is injected to produce the operating frequency. In the SB-400 the signal feeds directly into the driver stage but in the 32S-3 it gets amplified first by an additional 6AH6 tetrode. Both transmitters use a 6CL6 tetrode as a driver stage followed by a dual 6146 power amplifier.

The ALC signal is derived from two diodes in both transmitters and controls the gain of the IF amplifier stage, and, in the 32S-3, also controls the additional pre-driver RF amplifier.

In the SB-400 the CW sidetone is generated by a 6J11 oscillator and 6D10 amplifier and is simply fed into the receiver audio and the VOX circuit. The tone frequency is controlled to closely (but not precisely) match the CW offset in transmit which is provided by a separate CW carrier oscillator crystal. Collins used a different scheme. In the 32S-3 the sidetone, generated by a 6U8A, is fed into the transmitter audio chain just before the audio gain control in addition to the receiver audio and VOX circuits. In this way, the CW offset and the sidetone are always exactly the same, since they are both determined by the same oscillator. In the 32S-3 the spotting signal can be varied with a front panel control independent of the sidetone level. In SSB mode, the tone generator is disabled.