Heathkit - WARC & Other Bands

Home

I am often asked if the HW/SB line units can be converted for other HF / low VHF bands and my usual answer is ‘yes’ with some caveats, based on my own experiences with adding 160m to an SB101 and an SB401 and completely rebuilding an SB310 for all of the HF ham bands.


1. One option, if you must for whatever reason, is to modify your (only?) HW/SB unit for the required bands by losing some existing bands - see elsewhere on this site about adding 160m as an additional band. However, my usual recommendation is to obtain another SB/HW101 and modify that for the additional bands, leaving the original unchanged. The most common needs are Top Band and the WARC bands. In a standard unit the switch wafer construction only connects the 200pF preselector tuning capacitor on 80m so this has to become the new Top Band position - unless you want to re-engineer the switch wafers and that is not simple.


My preference for the other bands is to use the coils from the next band up and add capacitance to move the resonant frequency down to the required new band, for example use the 28MHz coils and add some fixed capacitance to move the resonance down to 24.9MHz for 12m and so on. The taps on the standard PA coil will need moving to optimise the PA loading and the best starting point is to move them halfway down to the next band tap.


It should be possible to modify the new HW/SB101 for 160m 30m, 17, 12m and use the remaining four band switch positions for 6m but care would be needed to ensure that sufficient RF filtering was provided on 6m to minimise spurious responses on both receive and transmit as Heathkit did not any form of balanced mixer for the RF conversion stages. Also bear in mind that the band oscillator is a parallel resonant type and crystals should be specified accordingly.


Modifying the switch wafers requires a vast amount of patience and access to a switch kit or similar collection of spare switch fittings - definitely not recommended for a beginner.


You would need to have two sets of loudspeaker, microphone and key or find some way of switching one set between the two units.

 

2. Another possible solution is to modify an HW/SB101 for use as a second front end for an existing unit with the signal interconnections done at the tunable IF. This means that a lot of the internal circuitry in the new unit will no longer be required and can be removed, making room for a new internal power unit. A number of the front panel controls will no longer be required and there will be a technical issue to resolve relating to how the two units will be interconnected - not impossible but needs some careful thought as the impedances in the standard circuits are high and would need to be matched to and from 50ohms for the coaxial interconnections. The wideband couplers couldn be opened up and a capacitive tap created to match to 50ohms but realignment would be required.


3. Alternatively the required circuitry (copied from the original) could be built inside an SB600 loudspeaker box with a new front panel containing the required controls - bandswitch, preselector, tune/load, meter switch and meter. Control lines would need to be connected for Tx/Rx, AGC, ALC etc and some method determined for switching between the original front and and the new front end. This option would require an additional external power supply.


4. New band crystals will be required as follows:




5. I think that it would be difficult to modify for the 5Mhz band as the internal VFO covers 5.5 - 5.0MHz and would be heard inband with a significant risk of a transmitted spurious response.


Additionally, conversion to LF (137KHZ or 470-500KHz) would be difficult because the existing variable capacitors would be too small for the correct LC ratio and PA matching. Perhaps an LF only version with larger tuning capacitors and a restricted coverage VFO or a transverter in an SB-600 case?


6. I hesitate to suggest the use of a transverter in front of a standard HW/SB101 because the additional conversion process and gain would significantly increase the spurious responses and degrade the overall dynamic range. However, the increased VLF RF selectivity may make this option viable.


7. Shown below is a simplified functional block diagram of the SB/HW101 transceiver front end which is a bi-directional crystal controlled converter translating the required band frequency to and from the tunable IF. Components Cn1, Cn2 and Cn3 are for PA neutralising. The diagram shows how the various tuned circuits are shared between the transmitter and receiver.


NOTE: This is not a full schematic and significant components including DC isolating capacitors have been omitted in the interests of simplicity. The driver stage neutralising is not shown.

Lower Band Edge MHz

Crystal Frequency MHz

Lower Band Edge MHz

Crystal Frequency MHz

1.50

10.395

24.80

33.6950

1.80

10.6950

50

58.895

10.0

18.8950

50.5

59.395

18.0

26.8950

51

59.895

24.50

33.3950

51.5

60.395

Additional Capacitance

Original Band

New Band

15pF

28MHz

24.5MHz

21pF

21MHz

18MHz

90pF

14MHz

10MHz

Top Band: New tuned circuits will be required for the RF, mixer and band oscillator circuits. The RF and mixer circuits should use fixed tuning capacitors around 330-470pF plus the 200pF preselector tuning capacitor and an appropriate value of inductance. The band oscillator can use the original 80m band oscillator coil with additional capacitance to lower its resonant frequency from 12.395Mhz to 10.395/10.695Mhz as required.


The following table shows suggested values of additional capacitance to move the resonant frequency of the original RF and Mixer band coils down to the nearest lower WARC band. The formulae to calculate the new values is located on the Useful Formulae page.