Turning the Hammarlund HQ145 into a Sweet, Sensitive, and Stable Ham Receiver

In this article we will look at a few of the big deficiencies of the Hammarlund HQ145 general coverage shortwave receiver, and put in some fixes.  CW/SSB reception with AGC and S-Meter, stability issues, bandspread, and sensitivity on the high band will be addressed.

Once modified, the HQ145 turns out to be a very nice ham and general coverage receiver all the way up through 10 M.  It’s like my wife: attractive, capable, and a pleasure to be with.  Or, it could be said Sweet, Sensitive, and Stable!

 

CW/SSB OPERATION WITH AGC AND S-METER
One of the inelegancies of this era’s typical receiver was the lack of a proper product detector with AGC functionality – when switched to CW/SSB, the AGC would be disabled and the user had to adjust RF and AF gains to hear an undistorted signal.  In addition, the S-Meter was disabled.

One way to overcome this limitation with minimal re-work of the receiver’s circuitry is to implement an audio-derived AGC.  Most circuits found on the web seem to suffer from stability and “pumping” problems.  The solution was to design and employ an AGC filter along the lines of the “lead-lag” filter often used in PLL loops.

The circuit below provides full functionality of the receiver’s switches, controls, and S-Meter;  with carrier-derived AGC when in REC/AM mode or audio-derived AGC when in CW/SSB mode.  The “lead-lag” filter is a standard fast-attack slow-decay design as set by the 0.1 uF capacitor and associated 27K and 10M resistors.

Audio is taken from V6’s cathode output, is amplified, full-wave rectified, filtered, and then buffered by the op-amp circuitry.  Calibration of “Gain” and “Offset” are required for proper operation, and is accomplished  with a signal  generator or with on-the-air signals.

Note that at S5, we remove the S-Meter wiring from the switch, and connect the two wires together permanently.  This is what we want since the S-Meter now works properly for ALL modes.  We now use the freed up switch contacts to over-ride the new audio-derived AGC, and instead use the existing carrier-derived AGC (ALC) when not in CW/SSB mode.

Measured voltage values of AGC (ALC) and detected audio for the new audio-derived AGC (in CW/SSB mode) are shown below.  These voltages are very close to the values as produced by the original ALC circuit when in REC/AM mode:

S-Meter Reading	V AGC (Volts)	V Audio p.p. (Volts)
S0	0.24	0.02
S1	0	0.3
S5	-1.09	1.4
S9	-2.14	2.7
S9 + 20 dB	-2.92	3.7
S9 + 40 dB	-4.02	4.5
S9 + 60 dB	-5.35	6.0

All of this circuitry is hidden under the aluminum cover on the inside rear chassis wall.

 

STABILITY ISSUES
Two main sources of frequency instability are addressed here – One is “warm-up” drift, which is mitigated with a simple always-on filament supply for the VFO and 1st Mixer tubes.  A second source is filament voltage change with varying line voltage, which is solved by using a new regulated DC supply for the VFO and first Mixer tubes.

This modification alone promoted the HQ145 from near worthlessness on the upper bands with several KHz to a few 10’s of KHz drift, to a very usable few tens to a few hundred Hz of drift under warm-up and line-load conditions.

This new filament supply (for the VFO and 1st Mixer tube filaments) is regulated DC at 6.3 Volts when the receiver is turned ON, and regulated down to 5 volts (presumably to extend tube lifetime) when OFF.  This unit also supplies the +6.3 Volts (called +6V) and the -10 Volts for the other modifications.

The transformer is mounted on the inside rear chassis wall.  Semiconductors are “hidden” underneath the other components built onto a new terminal strip mid chassis.

 

BANDSPREAD
On some versions of the HQ145, Hammarlund added an unlabeled sort-of calibration and mechanical bandpread control.  On my unit, it simply wasn’t all that consistent or useful.

Here we show that it’s a simple matter to replace the mechanical control with a compression type mica trimmer capacitor.  It may be that a small conventional unit could also be used, but I found the multi-turn functionality of the compression mica trimmer to be more useful from 40 M up through 10 M.  One pair of plates is used, and the unit is simply wired in parallel with the Main VFO tuning capacitor.

A longer brass screw was used in place of the trimmer’s standard adjustment screw.   Once tightened in place, a 1/4 inch threaded standoff was screwed and tacked onto the compression mica trimmer’s new threaded screw to allow for the use of the original knob.

The compression mica trimmer is snuggled into the stock chassis cutout in the stock location.

 

SENSITIVITY ON THE HIGH BAND
Hammarlund designed double conversion architecture into the HQ145 on the highest band for better image rejection.  Unfortunately this was done by using a penta-grid converter tube as the first frequency mixer which was used as a low gain RF amplifier on the lower bands.  Therefore sensitivity suffers on the higher bands where it is needed most.

The low gain problem was solved by building in an FET amplifier at the first IF of about 12 dB gain.   Receiver noise figure was good enough as-is such that further RF amplification was not required. 

The unit was constructed on a small circuit board using surface mount components.  It was then encapsulated in epoxy and painted drab green to resemble a “hybrid” commonly used in this era.

 

BOTTOM VIEW
Casual inspection shows nothing glaringly out of place or period, with the possible exception of modern-ish looking electrolytic capacitors.

 

Dave Benzel – KD6RF – 2017-Jan-6

 

10,255 total views, 4 views today