Hammarlund HQ145 Rebuild
  

I acquired this receiver from the estate of a good friend. The major part of his collection was Hallicrafters and National. It is significant that this particular radio, a Hammarlund, occupied the main AM operating position. It was his ONLY Hammarlund.

I acquired it because I knew its properties as a good fidelity AM receiver because of the 455 KC IF. However, owing to the dual conversion, the image performance on 20 meters and up would be superior.

It is dual conversion with first IF at 3.035 MC and second IF at 455 KC. It is similar in performance to the earlier relatives of the HQ 129, the HQ-120X, HQ-140X, HQ-150. The HQ-120 was actually a scaled down Super Pro, similar to the BC-775, BC-1004, and BC-779. Look inside and you will see the heritage. The difference between this radio and the HQ-180 is the 180 has a third IF at 60 KC, and is very sharply selective. The 180 is a great receiver for night time operation when the 75 meter band is crowded, with its sharp IF skirts and the ability to select the upper or lower sideband of an AM signal. The HQ-145 is essentially a HQ-180 without the 60 KC IF. In 2017, I began work on an HQ-170 for a friend, and my stable of HQ-180s. Some of that repair information may be helpful to you on an HQ-145. Here is the link to the HQ-170 & HQ-180 page: HammarlundHQ170-180.html.

The HQ-145 variable selectivity crystal filter will outperform the simple filter in the Hallicrafters SX-99 that I rebuilt in another article. "Wide open" the radio sounds great listening to the AM Class E rigs. The first crystal position is nice for AM with some interference. The notch is effective on heterodynes. The first three settings of the crystal filter are good for SSB. The 4 and 5 settings are great for single signal CW reception. The noise limiter actually works fairly well on power line pulse noise with AM reception.

It lacks a product detector, which means that the AGC is not operational on SSB or CW. You have to run the AF gain wide open, and use the RF gain for the volume. Turn it up for a weak signal, and prepare to have your earphones blown off your head by a strong signal that jumps in on the frequency. I may add this feature to the radio at some point in the future. There is a circuit that uses two diodes (Schottky would be better than the original 1965 handbook design). A buffer amp would have to be added to keep the BFO out of the AGC circuitry provided by the 6AL5. Switching of the audio output would have to be added also.

The Crystal Calibrator was an option. I home brewed one from the schematic in the manual. It did not work well, possibly due to my crystal. I eventually went with a circuit from a 1959 handbook using a 6AU6. The 6BZ6 in the Hammarlund calibrator had quirky performance owing to its tendency to take off on VHF parasitic frequencies. More on this later. I also used a 200 KC crystal to prevent confusion on the higher frequencies, owing to dial resolution. The actual tracking of the bandspread dial was amazingly good, and I found no need for the 100 KC on the lower frequencies.

The alignment of the 455 KC IF is always a challenge. If you have a sweep generator, it is easy. If you do not, check this article out: We dont need no stinkin' sweep generator by KD6RF.

The design problem that can complicate alignment of the front end is the use of the 6BZ6. For some reason, these tubes like to take off on VHF frequencies. The function of R1 and R3 is to suppress these oscillations. It does not work. The long pieces of buss wire going to the tuning condensers form a VHF tank circuit. Getting down to the tube socket to change these resistors to something different is impossible without complete disassembly of the band switch. Ugly work. I prefer the 6EW6. It has higher transconductance and is better behaved. It is pin for pin compatible with the 6BZ6. NOTE: Without an antenna or resistive load attached to the antenna terminals, the front end may still take off, but it is OK with something connected.

The older versions of these receivers, including the HQ-180 and HQ-170 are less prone to this problem that the A versions. The later versions are worse for VHF oscillation because the antenna trimmer capacitor is mounted to the front panel, eliminating the dial stringing inside the radio. This capacitor is attached by a short piece of coaxial cable that is also resonant in the VHF region. A small resistance in series can tame it, but keep it as low in resistance as possible. Forget ever getting a HQ-170 to ever work on 6 meters properly. The noise figure is terrible and the stability is unusable. Get a converter, if you have to use it for 6 meters. They will work OK on 10 meters and down.

One other reason to prefer the non-A versions is that the later versions have a small transformer which keeps the filaments lit on the 6C4 HFO and 6BE6 conversion tubes, thus assuring a speedy death. I have serviced these radios for friends that had tubes in those positions that did not budge the meter on my tube checker at all. The cathode glazes over. I tried without success to rejuvenate them in a process similar to picture tube rejuvenation; no success with that. Best solution is to rewire the primary of the small keep alive filament transformer to the AC switch, or operate the radio from an outlet strip that can be turned off completely when not in use.

One other treatment I always apply to these Hammarlunds is to completely eliminate the balanced input wiring. I go with a short piece of RG-174 from the coax connector directly to the bandswitch for the antenna coils. I ground the unused balanced side of the coil directly at the T1 to T4 antenna coil. This prevents some coupling from the input to output of the RF stage, causing oscillation on the operational frequency.

The reason this is problematic is that inexperienced technicians may mistake the oscillations for the HF oscillator being way out of alignment. This is what happened to mine. Someone had forced the adjustment slug into L6 so hard that it could not be budged, in an attempt to align the radio. As a result, the tracking on AM with the dial calibration was terrible, and sensitivity was way below par. I eventually used a coil from an old broadcast receiver to substitute for the existing frozen coil. Luckily, it was not a shortwave coil. The substitute coil was a bit high in inductance. I measured the old one and took enough turns off the new coil to allow calibration of the dial and it turned this radio into a stellar AM DX radio.

The other problem I encountered was an oscillation of the RF stage (using either the 6BZ6 or 6EW6) was oscillation on the Broadcast band! I had never encountered this before. I found that a 10K resistor across the winding of L4 used for the broadcast band tamed it without excessive loss of sensitivity. This problem likely was the reason the oscillator coil for the BCB was trashed.

By the way, the factory schematic is incorrect for the pin out of the 6BZ6. Pins 2 and 7 are swapped.

Solid state the power supply to remove heat. I used a 300 ohm 10 watt resistor between the cathodes of the diodes and L9 and C62A. The capacitors were all leaky, so I soldered the hot terminals together on the old can and grounded them. I attached a new terminal strip by soldering its ground terminal to the terminals all joined as described. The remaining terminals were used to connect the new capacitors to replace C62, except for C26C, which is the cathode bypass resistor on the 6AQ5.

Now the radio was running factory stock, with the crystal filter accessory.

Parts to get you this far may be available at sources listed at: http://www.k7jrl.com/pub/manuals/hammarlu/hq170notes/HQ170Notes.pdf.

I did the audio mods found at: http://www.amwindow.org/tech/htm/hqaudiomods.htm. These are easy to do, and work excellently. You will be really pleased.

Getting it back into the cabinet without mangling the clock setting shaft is a trick. Here's what I do. I slide the chassis into the cabinet about 70% of the way. I slide a piece of shrink sleeve big enough to guide the shaft, but small enough to pass through the rear cabinet hole, over the rear of the shaft. Now as you insert the chassis into the cabinet, the shrink sleeve guides the clock shaft right through the hole. SLICK!

Be sure to attach a decent speaker to fully appreciate the beautiful audio response of this fine radio. A Hammarlund S-100 or S-200 will have a nice presence rise and full range response from an elliptical speaker inside. I have one, and it sounds natural. I have also found that thrift store stereo speakers from boom boxes or smaller stereo systems are a good choice. Often they are very efficient, making good use of the limited power of the 6AQ5. They often have surprisingly good frequency response. Look for a separate tweeter and ports for extended bass frequencies. One of the best I have was snagged out of the dump discretely.

While the HQ145 is not a Super Pro or an SX28, it sure is nice. It is also very affordable by comparison. It is sensitive and capable, for a radio of its age. It is surprisingly stable mechanically and thermally. Later radios with crystal controlled first oscillators will be better, but this is as good as it gets for this period of technology. Below 15 meters, it is not an issue. For AM work at 15 or 10 meters, it is not a problem either. It can get dicey on SSB up there, but CW is useable. The 20 meter band on the HQ-145 has a special band spread position that is well done.

I do not favor radios with gratuitous complexity, such as early synthesizers, which are a nightmare to service and have gawdawful phase noise. This HQ-145 is just about right and has that lovely Hammarlund old buzzardly look to it that just screams here is a real radio.

Here is another take on what is the best AM receiver: Johnny Novice's Guide to Receivers.

This radio is configured in my main AM station as an alternate receiver to use with the Valiant under good conditions.

THE TURBO HAMMARLUND HQ145! UPDATED 3/2014

This is a beautiful looking classic radio, from the same parentage as the legendary HQ170 and HQ180. There are Hammarlunds similar to the HQ145 that are ham band only. I prefer a radio that covers all the WARC bands, since I own a Viking 2 that will do the WARC bands. The HQ170 has a drift problem on 40 meters that sometimes makes the upper part of the band inaccessible, unless you are very careful to fudge the alignment just right. This is really annoying to an AM operator, since the window is 7290. I don't know if the scaled down cousins of the HQ170 have this problem. Also, I like BC band dxing and SWLing, and the ham band only radios do not have WWV on them to check the dial calibrator. The HQ145 was introduced as a replacement for the HQ129. In actuality, it is a big improvement owing to the dual conversion, more modern components, much better sensitivity on 20 meters and up, and the notch filter. It was mildly costlier than the Nationals and Hallicrafters in its performance and price range, but in my opinion, worth the difference. The reason Hammarlund lost out in the race for performance was their failure to incorporate a crystal controlled first conversion oscillator for better 20 through 10 meter stability, very important to the upcoming SSB revolution. National, Hallicrafters, Collins, Heathkit, and even Lafayette offered cheaper radios that offered that feature soon after the introduction of the 145, 170, and 180.

The original part of this article was to describe getting the HQ-145 running in essentially stock form. This part takes it to the next level. As I mentioned in the first part, the HQ-145 is essentially a HQ-180 without the 60 KC IF and the product detector and hang AGC. For AM work, that would have done the job. Recently, I used this receiver in an AWA CW QSO Party that required pre 1960 gear. I found that the lack of AGC in CW mode was intolerable. I had forgotten how bad the good old days were. I was spoiled. But this radio is not beyond help. We can fix it. Better than before. We have the technology.

As has been said: "One thing I've learned in all my years is nothing works out quite the way you want it to." But we have choices.

I alluded to this previously. I had installed a product detector in a Hammarlund Super Pro 210 many years ago. I dug into my files and pulled out a proven design. Even the tube used was perfect for the job: a 6SN7. The socket vacated by the 5U4 when I solid stated the power supply was empty. I did not want new holes in the front panel or chassis. Under the chassis was fair game. So the "TURBO" mod adding the product detector is at the end of this article. But first the basics.


SOLID STATE POWER SUPPLY AND 125 VOLT OPERATION

I thought I would first include some photos to explain the previous work. First, mount the new silicon diode rectifiers on a terminal strip mounted to the transformer screw as shown. Use a 150 Ohm 5 Watt resistor in series with the diodes feeding the power supply choke and input condenser to protect against surge on power up. The high voltage secondary wires will reach easily to this location. Leave the red/yellow center tap wire attached to the ground on the 5U4 tube socket.

Use the yellow 5 Volt filament wires from the 5U4 in the correct phase (You MUST REDUCE AC voltage to the primary of the transformer) to operate a 115 Volt radio on 125 volt mains. Connect between the ADDED 2 AMP FUSE and the terminal strip feeding the radio as shown. You MUST do this or at only 120 Volts the 6.3VAC measures over 6.6 Volts. This will shorten tube life by straining the filaments. This reduced B+ voltage and created another problem which I had to fix.

This manifested another problem. The VR tube now would extinguish in some circumstances and be too bright under others. I did some measurements, and found that in standby the VR tube was running near or above 40 mA. The 7 pin VR tubes are only rated for 30 mA. The 6C4 oscillator, which runs during standby, draws 4.3 mA. You can measure the VR tube current at the 4K resistor easily without disconnecting anything. Originally I had a 300 Ohm resistor in the solid state diode feed. After all of the math was done, the 150 Ohm resistor in series with the power supply diodes reduced the standby voltage feeding R41. But it was necessary to add a 270 Ohm 5 Watt resistor in series with R41 (4K) to get the right combination that would keep the tube lit during SSB operation, while within its maximum ratings in standby mode. I suspect this problem was inherent in the original design and went undetected like the errors on the schematic I have noted. I did not want to complicate the station with yet another outboard bucking transformer sitting on the floor, if I could easily update the existing radio so that I could plug it into any outlet.


CRYSTAL CALIBRATOR AND ANTENNA TRIM DIFFERENCES

The antenna trim capacitor on the A version of the HQ145, HQ170, and HQ180 is mounted to the front panel directly and attached via a coax cable. This results in a VHF parasitic circuit. This photo shows why an earlier non-A version is preferable.

The home brew crystal calibrator which could be a clone of the Hammarlund unit is mounted in the same place as the factory provided unit. I used a 200 KC crystal to make spotting the band spread onto the correct harmonic on the highest band less ambiguous. This is why I used a different circuit. I also used a gimmick capacitor of wire coupled to the solid lead of the main tuning capacitor to get enough injection, instead of the wire to the antenna screws on the back of the set. The screw terminals have been disconnected and the balanced wiring all removed anyway. I found replacing all that with RG174 and directly grounding the appropriate terminal of the antenna coil improves stability, and most people do not use balanced wire to the receiver anyway. It is usually converted in the antenna match unit to coax anyway. The adjustment of the crystal calibrator with the BFO on in the stock radio is difficult, because the BFO frequency "pulls" and if the ratio of all signal strengths is not just right, the beat note is ambiguous. With the product detector modification, it is a lot easier. You can put the mode switch to AM and test clip the crystal calibrator B+ input directly to the VR tube B+ temporarily to do the zero adjustment of the calibrator. In normal use, to set the band spread, you would normally use the BFO; so this is not a major defect once you have the calibrator calibrated.


S METER CALIBRATION AND IF STAGE TROUBLESHOOTING

The S meter zero drifted badly during the recent QSO party. The manual suggests replacing the 6BA6 IF tubes, or swapping them. That did not help. I found that a 2.2K resistor R19 had changed value to over 5K. Replacing this corrected the bridge balance for V4 and V5. By the way, the schematic is incorrect about the pin 1 and 6 of V4 6BA6. They are swapped. The pins on V5 are correct. If you have this problem, check all the 2.2K resistors involved with the IF tubes V4 and V5, and replace all that are out of tolerance.

I then found that after the repairs and the 6EW6 mod for the RF amp and a proper alignment of the 3035 KC IF cans, the S meter often pinned on strong signals, the overall gain had improved so much. I found it necessary to add a pot to reduce the full scale reading (as the 170 and 180 do for calibration) in series with the wiper of the existing zero adjust pot. Quelle bonne surprise! I use optimum resonant antennas for each band. You may not find this necessary. I reckon that WWCR, local AM stations, and stations on 3875 should read on scale near the maximum deflection. The 50 uV standard is irrelevant to me, but you can do that if you have a signal generator that is calibrated for it. However, the "standard" never specifies what antenna delivers the 50 uV, so I contend this is the right way for me.

 

NOTCH TUNING LIMITATIONS

One alignment note: The front panel NOTCH TUNING does not fully move the notch outside the excellent passband during AM reception. If you do not use the notch much, adjust the internal calibration slug so it is offset to one side or the other. The center of the pass band may be at plus or minus 2 KC indicated, but it is still useable in most cases. At least this way, the response is symmetrical as observed on a sweep generator, if you leave the front panel tuning to the side most removed from center frequency. Maybe I am just being picky, this is not essential. The HQ 170 AND 180 do not have this problem since their widest bandpass with the 60 KC IF does not exceed 6 KC total; the notch is placed outside their response when properly aligned.

MUTING THE HQ-145 FOR T/R SWITCHING

If you plan to use the HQ-145 as part of an amateur radio station, you will have to provide for muting to prevent squealing feedback on AM. The existing method provided by Hammarlund via the AC outlet connection on the rear panel can work but it is clunky and requires switching high voltage in the external T/R switch. There's another way provided by D-Labs with a relay kit which can be installed inside the HQ-145, which removes high voltage from external T/R systems. Order it here: https://d-labelectronics.com/d-lab-push-to-talk-module

The installation is simple; just follow this schematic, provided by Terry, N6TLU.  

OTHER POSSIBLE IMPROVEMENTS

If you were going to install a mechanical filter, you would probably put it in place of the notch filter. I may do this someday, as I have some surplus 2.1 KC units. Beware that the input transformer has a SERIES tuned secondary, so this could be trickier than first anticipated. But this radio is used for AM and CW, not SSB, so that is not as likely as other choices. I may convert the notch components to a Q multiplier. This is more versatile than the phasing notch on the existing crystal filter. Plus it provides a slick vernier tuning for a narrow CW peak that could work better than the existing crystal filter on bands above 30 Meters. This is not a problem with the crystal filter; it is a problem with the tuning getting touchy on the upper bands. I will provide an update if I do this, as it may be of interest to others who wish to soup up this otherwise excellent radio. The improvement would only be useful for CW on the highest bands when you needed a narrow band pass to eliminate QRM. The complications involved are the series tuned IF can, and the necessity of adding a triode tube somewhere.

PRODUCT DETECTOR FOR HQ145 PLUS SLOW AGC
THE TURBO HQ145 MODIFICATION FOR CW & SSB

Well, that is all you need to get the receiver running basically factory stock with a few improvements. Now comes the "TURBO" mod to get great CW and SSB performance. ESSB signals below 3.7 MC sound pretty nice on this setup. After you have cleaned up the octal socket that used to be the home for a 5U4, you now have a space for the new 6SN7. This octal tube is very close in characteristics to the 12AU7 used in the HQ170 and 180 for product detector. The circuit I use is also similar, with some adjustments for 455 KC IF. The original 12 pF coupling capacitor for injecting BFO to the 6AL5 is used in its original physical location. I just disconnected the original coax from the BFO and added a new coax RG58 to the detector side of the 6SN7. I disconnected the other end of the stock coax from the BFO end. Leave the original wire in place and put a piece of shrink sleeve over the end to prevent it from shorting to anything. Add a new piece of RG58 from the BFO (use a 100 pF mica blocking condenser) to the new product detector BFO injection port. I may reduce the 12 pF even more to see if it makes the strong signal performance better on SSB. Be aware that many SSB signals are distorted when they leave the station of origin, so do not blame this radio. Some modern finals running off 12 volts do not have good distortion figures, especially when pushed too hard on the ALC.

All the new resistors and capacitors for the product detector can be mounted on terminal strips attached by screws mounting the power supply filter choke.

The modification can be tested in the original circuit without the plates of the 6SN7 wired up for proper voltages. This is how I selected the 12 pF and 100pF coupling condensers. The signal port (pin 4 of 6SN7) should have about 2 Volts peak to peak max using a 10X probe on your scope to prevent loading of the circuit. The BFO port (6SN7 pin 1) should have about 8 Volts peak to peak. All these are at 455 KC, so most TV service scopes should suffice for measurements.

I then wired the new product detector temporarily in circuit and tried it. The AGC and S meter worked. No deflection of the S meter occurred when the BFO ran, unlike before when the S meter read half scale when the BFO was on. SSB signals were strong and undistorted. CW signals were held at a reasonable variation from barely readable to above S9. Adjustment of the frequency of the crystal calibrator is much easier now, since the BFO frequency does not pull due to overload.

So the final problem to be solved: how do I switch from the original 6AL5 AM envelope detector to the new 6SN7 product detector with the existing function switch? The BFO B+ is switched for the AM to CW mode change. I used that voltage through a 20K resistor to a 5 Volt reed relay (SPDT). It only needed a coil current of 10 mA to actuate the contacts. I verified this by test of the relay with a small variable power supply. Try to keep the current draw low with a sensitive relay. See the schematic for details. I use a 0.01 uF coupling capacitor from the product detector to limit the low frequency response on SSB. I upped the coupling condenser from the 6AL5 AM detector for better low frequency response, and ran BOTH signals thru it, as selected by the reed relay.

Once this step is completed, you have basically a radio that performs like a HQ170 or 180 with wider bandwidth on the OFF position of the SELECTIVITY switch. Of course, the sideband selection function has to be taken care of with careful BFO adjustment as you would with the non-A versions of the 170 and 180. And you do not have the fine vernier tuning of the 170 or 180, which is handy for SSB. But this is a big improvement over the stock HQ145, particularly for AGC use on SSB and CW. The AGC time constant is switched by the original AGC on-off switch. Now it is FAST-SLOW AGC. With a system this nice, I NEVER find the need to turn AGC off. I did not ever use that function on the 170 or 180 either. I used 0.27 uF added to the AGC circuit to get a slow AGC response. It mounts to the AGC on/off switch in series with the ground lead to add capacitance.


I find that SELECTIVITY settings of 1, 2, and 3 are good for SSB. For CW, I use 4 after I pick the station I want, or 5 (minimum bandwidth) if there is QRM. The lack of a FINE VERNIER like the 170 or 180 makes use of SELECTIVITY setting 5 on 20 meters and up a bit touchy, but it can be done with a gentle touch. Below 30 Meters, band spread tuning is OK for CW using SELECTIVITY 5.

I love Hammarlunds for the heavy front panel and mechanical stability, a big step up from most National and Hallicrafters radios in the same price and performance range. With the new goodies, this is a really hot receiver on 10 meters. Even CB signals are capable of driving the S meter to S9 or better, and ham signals on 10 and 15 frequently go well above 20 over. CW signals when working traffic nets or pileups are not so fatiguing due to wildly varying audio levels.


IMPORTANT HEADPHONE SAFETY NOTE

It is still a good idea to use an audio attenuator for headphones on this radio, since the direct audio speaker output is available on the front panel. This was designed with 2000 Ohm headsets in mind. I like Radio Shack light weight low-Z monaural full frequency response headsets for AM use, so I can monitor the outgoing transmissions. But I use a separate headphone breakout box for multiple headsets that I constructed for field day with individual volume controls (one for me, one for my logger). Why not just put a new headphone jack with switching contacts and a resistor to fix this? You could. The way I use my station, I share the excellent Hammarlund speaker with the radios on my operating position. I do not want to have to pull the radio out to connect it on the back, so I plug the speaker into the headphone jack of the radio I am using at the time. An attenuator in the headphone jack would prevent that desirable feature.

EVEN MORE TURBO HQ145 TO COME!

I may install a Bishop noise limiter in this radio later, which is in the last IF, not the audio. That way, the noise limiter will work on AM and SSB/CW. This is similar to the one in the Heathkit SB310, and it works great.

  

73,
Janis
AB2RA
Text size:  +   –      Back to Vintage Equipment
Back to Receivers
Back to the Home Page