H's Comet

by John M.Bogath

I suppose that most of you are now experiencing the wrath of the winter season as you mad this. Therefore, you probably won't mind me writing about events that occurred last summer. Let's regress a bit in time. It was to be a very warm day back in July, and I was prepared for it. I had tanked up on water, covered myself with stinky sunscreen, and dressed myself in breathable, light-colored clothing. I chose very comfortable sneakers for my feet, because I expected to do a lot of walking. I was going to attend the big Sussex County (N) hamfest.

The Sussex County hamfest is put on by the Sussex County Amateur Radio Club (no surprise here, HI); and I must say that these guys really know how to throw a party. Acres of parking, acres of tailgating/and a huge indoor facility draws people from all over the northeast, mid-west, and mid-Atlantic areas. At a time when hamfest seem to be dwindling and disappearing, this one is getting even larger.

Well there I was; walking passed table-after-table-after-table of all kinds of electronics goodies, including many boatanchors. There's an "SX-l00", "NC-183", and "Invader 2000"; soon followed by an "NC-300", Eico "720" with "730", Globe King 500C", and a Johnson "Valiant". Now, I see a "Ranger", "HQ-140", "R-388", "HQ-180", "Comet Pro", and an "SP-600". SCREE-EE-CH! "COMET PRO" ?!? BACK UP!! Yes, buried underneath a couple of other boatanchors was a fairly decent looking Hammarlund "Comet Pro", with the crystal filter option. I was focused. All other pieces of equipment disappeared from my sight as the vendor and I carefully dug the ancient-looking "Comet Pro" from out of the pile. After a brief examination of my discovery and a few questions, I made the vendor an offer which was accepted. I tingled with excitement and anticipation as I carried away my newest treasure.

Now, if you are long-time CXer, like Wilson (W4BOH), you know why I was so excited to acquire one of these receivers. Wilson is the only other CXer I know who has a Hammarlund "Comet Pro". He inherited it from his father. The Hammarlund "Comet" shortwave receiver became a reality towards the end of 1931. It was simply the very first commercially produced superheterodyne shortwave communications receiver ever made. The key word here is "superheterodyne". It had intermediate frequency (LF.) stages and could exhibit the same high degree of selectivity on 20 meters that it had on 160 meters. No previous commercially-made HF receiver could claim that. Imagine what our VHF/UHF bands would be like today if all receivers used were like those in the Heathkit "Lunchboxes". Selectivity is an extremely important characteristic in all receivers, and the Hammarlund "Comet" is the "granddaddy" of all modern HF and above communications receivers, in this respect.

Throughout 1931 and much of 1932, Hammarlund continued to make improvements to their "Comet". Later in 1932, Hammarlund felt that the current "Comet" was ready for the professional market and renamed it the "Comet Pro". In 1933, Hammarlund offered two very useful options for the "Comet Pro": a crystal filter and automatic gain control. After that, the "Comet Pro" appears to have been essentially left alone so that the Hammarlund engineers could concentrate on the next generation of this "Pro" line: the famous "Super Pro" series. However, between 1931 and 1935, Hammarlund appears to have sold many "Comets", making it a very successful venture.

All "Comets" look like something from out of the 1920s.

They all have a plain black panel with plain and smallish black knobs and little "peek-hole" dials. There are no labels on any controls, so you need to remember what each control does; or forever fumble. (HI) The three tuning windows (i.e., "peek-holes") only reveal logging scales; no frequency markings.

One nice feature, though, is the somewhat ornate central window which displays a band spread scale.

Behind its frosted plastic window is a drum that has a celluloid logging scale attached to it. This drum and scale rotate when the center (band spread) knob is turned. Its logging scale looks like CI photographic film negative that is backlit by an internal lamp. That causes the scale to be projected onto the back of the frosted viewing lens. This feature eliminates parallax problems, and enables one to easily return to the exact same frequency of a desired station.

Front Panel Of Comet Pro.

The bottom-right toggle- switch was added by someone (not factory) to provide "stand-by" capability. The factory did offer this option later on and located the toggle-switch for it in the same place. It is interesting that the AVC option also placed a toggle-switch there, as well. I wonder what happened if a customer ordered both of these options together ... footswitch! (HI)

Looking at the chassis bottom

If one looks inside of one of these receivers, the first thing you will notice is lots of aluminum shielding.
The I.F. coils are shielded, the crystal filter is shielded, all non-choke coils are shielded, the tubes are shielded (except for the rectifier and audio output), many capacitors are shielded, the audio low-pass filter is shielded, and there is even a partition shield under the chassis for the I.F. stages. You could not afford to build this receiver today because of all of the metal work; and I haven't even mentioned the really heavy-duty steel chassis, front panel, and cabinet.

Anotherlook at the chassis bottom shows it with several shield cans removed.

If one continues looking, you will eventually see that underneath the chassis there are no terminal strips or binding posts to secure circuit wires to (What the __ _ _ ?). Instead, the components are soldered to switch terminals/ tube socket lugs, or whatever, on one end; then the other end sucks up into the air to meet with wires room other "free" components or to a general interconnecting wire/bus that may run for a considerable distance as it ties a series of components together. Such assembly work is often referred to as being "sky-wired" and is rarely seen in more modern radios.

Another interesting feature of the "Comet" is its mechanical fine adjustment control for the BFO frequency.

It looks like an old automotive on-the-column gear- shift-lever sticking out of the top of what looks like a large I.F. can.
To get to it, you must lift the cabinet's trap-door access cover and reach back to the right-rear corner of the receiver. Oh; those crafty Hammarlund engineers!? (HI)

Don't look for a volume control. There isn't one. Receiver sensitivity and volume are both controlled by one LF. gain control. I guess if you're not going to label anything on your front-panel, you might be wise to minimize the number of knobs and switches on it. (HI)

As with the non-existent volume control, don't look for an AVC/AGC switch either; unless your "Comet Pro" has the AVC option installed. Mine doesn't, and I am surprised to say that I don't miss it. Manual control of the I.F. gain circuits seems to do the job just fine with this receiver.

This receiver does have a "Tone" control, however. It is attached to the A.C. power switch and when advanced, gradually cuts off the receiver's high frequency audio response. I found this feature to be useful as a low-pass filter that proved to be effective in reducing QRM on the CW bands. On the Fone bands, it can be used to reduce background hiss and make copy more pleasurable.

All of the labeled capacitors (or should I say "condensers" ?) in my receiver were made by Aerovox, Hammarlund was located in New York City at that time, and so was Aerovox, Hammarlund may have preferred to have their vendors close by. The electrolytic capacitors are interesting. They are "wet" electrolytes and differ from both modern electrolytic and oil-filled capacitors. They are filled with a water-like liquid that has a chemical dissolved in it. I assume that the chemical increases the dielectric constant of the liquid which boosts the capacity (i.e., more microfarads per given size) of these capacitors, If the capacitor dries out, a white powder residue is left behind (i.e. chemical )

This shows an interesting item of historic intrigue: the metal plate attached to the rear apron of my receiver's chassis.

This plate says, "Licensed Only for Radio Amateur Experimental and Broadcast Reception by Radio Corporation of America". At that time, RCA held the patent for the superheterodyne receiver and anyone wanting to make a receiver of such design would have to pay RCA licensing fees for the privilege. As written, it would appear that my receiver was NOT licensed for "professional'' use and could not be legally sold to the government or to a corporation. Did that "privilege" come later or did receivers in the hands of "professional" users have a different licensing plate affixed to them? ... one with a greater fee attached to it, perhaps ?

This shows where the plate was attached.

I have a 1930 AM broadcast band radio that is of superheterodyne design. I wonder what held up the making of a shortwave version of this receiver. The only thing I can come up with is that the engineers of that era were having problems designing an acceptably low-noise mixer that exhibited reasonable conversion loss over a wide range of frequencies (e.g., 0.5 to 20 MHz). Mixers can be "touchy" and lose effectiveness easily if circuit conditions change. Keeping those conditions constant over a considerable frequency range may have been the "hurdle" that the engineers were having problems clearing. Often, an RF amplifier ahead of the mixer can be used to "hide" some of the mixer's ills. That's why I was surprised to learn that the "Comet" does not have an RF amplifier. Did Hammarlund's engineers "hit a home run" with their mixer design; or is the performance of the "Comet" seriously compromised by a mediocre mixer and no RF amplifier to help it out ?

To find out how my "Comet Pro" performs, I first needed to get it working. Most CXers know what that means. You don't just blindly power up an 83 year old radio, that's essentially original condition, and start looking for smoke! Or do you!!! (HI) A 20 amp fuse in my "Comet's" fuse holder and electrolyte oozing from an electrolytic cap told me that someone had already tried (OUCH I). I went to work; and since I'd be "preaching to the choir", I'll spare you CXers the details as to what and how I did it. Let it suffice to say that because of the age and historical significance of this radio, I kept it looking "original" throughout. The only obvious non-original part is the new 3-wire A.C. power cord. Safety trumps "pretty"; especially in a piece of equipment that won't just sit on a shelf. This "baby" is going to "prowl" plenty.

So, just how good is the "Comet Pro"? On 80 meters, a 1.0 microvolt signal (50 ohms input impedance) produced "armchair" CW copy. A 0.3 microvolt signal was barely copyable. This was with a set of "high mileage" tubes in the receiver and no crystal filter in use. I feel that the crystal filter in my "Comet Pro" could be performing better, so I am not using it for the purposes of this article. I hope to investigate this filter circuit more fully in the near future. It would certainly appear that the Hammarlund engineers did indeed "hit a home run" with their mixer design; at least as far as 80 meters is concerned.

In order to check the "Comet Pro's" sensitivity on other bands, I needed more coil sets. A coil set for this receiver consists of two coils; an oscillator coil and a wavelength (abbreviated "W.L.") coil, which tunes the mixer's input to the desired receive frequency. Since I may not live long enough to locate such coils just "floating around" somewhere (HI), I decided to dive into the old "junk box" and see what I could find. The "box" yielded four good coil forms and some good sizes of enameled wire; enough to make two more coil sets. I then proceeded to wind my own coil sets that would include 40 and 20 meters.

Sensitivity of the "Comet Pro" was the same on 40 meters as for 80. Twenty meters was `only slightly less sensitive. One microvolt still provided an easy to copy CW signal on 14 MHz or is that "MC"?), while a 0.4 microvolt signal was the minimum needed to still be copvable. I did not notice any images on 80 or 40 meters, but did notice several on 20. This was expected in a receiver with an I.F. frequency under 0.5 MHz and only one tuned circuit preceding its mixer. A good amplified preselector could improve this receiver's performance significantly on the frequencies above about 40 meters.

Don't, however, get the impression that this receiver is not a stand-alone performer on 20 meters. With just a 40 meter dipole for an antenna (think "mismatch" here), I easily tuned in dozens of DX stations on both 20 meter CW and SSB. Yes, the "Comet Pro" is quite a SSB performer. At one point, I tuned In a western-U.S./Canada roundtable on 20 meter SSB, and after 15 minutes of listening, I still did not need to correct the tuning for drift. All participants were "Q5" copy, as well. I did not expect that. I also did not expect the ease with which these SSB signals can be tuned in, especially since the dial mechanism and tuning capacitors are all dial-cord driven! I had to replace the dial-cord in mine and am still amazed that the tuning is essentially solid and backlash-free with such a design.

Speaking of tuning, the "Comet Pro" tunes as if its selectivity skirts are quite steep; especially for a receiver with only two I.F. stages and no filters (remember; for the purposes of this paper, I am not using the crystal filter option installed in my radio). Hammarlund claimed a "Q" of 115 with their 465 KHz ("KC") I.F. coils (there are three of them). These coils were wound with "Litz" wire and tuned with air variable capacitors, instead of the usual mica-compression units. The extra effort in the manufacture of these assemblies really shows in this radio's performance.

So, could a "Comet Pro" be used as a primary receiver in an amateur radio station today? My answer would be a definite "yes", with a few cautions. First, the operator would need to understand the radio and know how to operate it to full capability in the mode being used. It would not be a good receiver choice for an avid DXer or serious contester. However, having said that, the "Comet Pro" is up to the challenge to provide a skilled operator with many DX contacts and many points in less intensive on-the-air events. I know it "runs circles around" my first ham receiver, a Lafayette "HE-10"; and I made many, many QSOs with that one. Clearly, the Hammarlund people squeezed every bit of performance possible out of this basic design, because of that effort, the "Comet Pro" still remains a serious piece of communications equipment, even 83 years after its manufacture.

I deliberately titled this article "H's Comet". How many of you CXers out there only had thoughts of outer space and "Halley's Comet" when you read that? Come on; admit it. OK? From now on, if you see the title "H's Comet", or even "Hailey's Comet", will you also think of "Hammarlund's Comet" as well? If so, then my purpose In writing this will have been fulfilled.


Figures 1 through 8


Figure 1



The three "peek holes". The one on the left shows oscillator frequency adjustment (i.e., coarse frequency); the one on the right shows mixer input frequency tuning (i.e., wavelength). The ornate center dial is the bandspread tuning readout.



Figure 2



"Comet Pro" tuning mechanism. Note the celluloid bandspread scale in center-foreground. Shield-covers have been removed to reveal the plug-in oscillator ("OSC") and wavelength ("W.L.") coils for 56 to 120 meter coverage.



Figure 3



Front-panel view of "Comet Pro". The bottom-right toggle- switch was added by someone (not factory) to provide "stand-by" capability. The factory did offer this option later on and located the toggle-switch for it in the same place. It is interesting that the AVC option also placed a toggle-switch there, as well. I wonder what happened if a customer ordered both of these options together ... footswitch! (HI).



Figure 4



Chassis bottom with all shields in place. The large, green, wire- wound resistor was added to lower the 120 VAC input voltage.



Figure 5



Bottom view with shield cans removed from BFO circuits and audio low-pass filter.



Figure 6



BFO tuning lever. Can you guess what gear this receiver is cruising in ?'! (HI)



Figure 7



Direct photoscan copy of rear-apron plate describing RCA licensing terms.



Figure 8



Rear view. All shields are in place. Note the shiny rectangle near the center of the rear-apron. That is where the RCA licensing tag was located.



Copyright John M Bogath
November 2016