Crystal Radios Of The 2008 Contest Entrants
This year's 1AD set is a Bazian Reflex circuit, modified for RF regeneration, as first suggested to me by Macrohenry. This set is hot but very temperamental to use. During three years' collaboration there were many occasions where one of us could not reproduce what worked well for the other. There are too many subtle, easily overlooked factors at play in this circuit.
Like last year's Homodyne, this is a built-in antenna set. The 7-1/2 inch long ferrite rod L1 is all the antenna there is. The set is rotated on a lazy susan. Key to the Bazian's performance is the MPSA18 low noise, high gain (around 1000) silicon transistor used.
RF regeneration is had by ever-so-slightly coupling the reflex RF transformer T1 back to L1 in the primary tuning circuit. Rotating T1 relative to L1 controls regeneration. Starting from "zero-coupling" (T1 axis perpendicular to L1 axis), rotating T1 brings the set to a precarious state of extraordinary sensitivity and selectivity. But, if the set is over-coupled by the slightest amount, it locks up with a loud pop and goes dead. Then, coupling must be decreased well below optimal coupling to bring it back to life. Sometimes, at optimal coupling, a rise in the station's signal level, fading in, will trip the set off. There are ways to tame this violent action, but so far I haven't found one that doesn't tame performance as well.
Needless to say, vernier adjustment of coupling is a must. Over the entire broadcast band only +/- 5 degrees rotation is used. It's easy to overdo selectivity in this set. That results in muffled audio detrimental to understanding weak DX signals. Occasionally it is best to back off the selectivity
in order to sharpen up the sound.
Because of the loud discharge pops, I opted for a less sensitive pair of U.S. Instrument UA729-7 elements from a surplus handset. A simple line transformer matches the phones to the set's low impedance output. The Bazian's performance doesn't match the Homodyne's, but on a per-transistor basis the Bazian is more efficient.
A convenient feature of this set is it loves dead 9-volt batteries. I've run them down to less than 2.5 volts, at which point they refuse to deliver any more current. Rejects from smoke detectors provide me a lifetime supply.
Check out this URL for more info pictures of Tom Polk's radio: Tom Polk's 2008 1AD Contest Radio
For more info, visit Dave's website at:
Dave Schmarder's 2008 Tube Radio Contest Entry
The receiver is a two tube, two active device set. The first stage, V1, is a converter using a 6BA7, and is based on the KE4ID one tube superhet. I simplified the front end to a single tuned stage mainly due to lack of parts. Some component values were tweaked a bit. AVC was also added. Lack of a double tuned front end has caused image problems with strong local transmitters, but the addition of a bank of wave-traps when using the long wire antenna has minimized the problem.
The second stage, V2, is a reflexed IF and AF amplifier using a 6HJ8. IF regen was added to this stage later, after it was up and running reliably. A few points of interest, regarding this stage, follow:
- The regeneration is largely capacitively coupled. I had noted that it would work with the tickler coil leads reversed. I later disconnected the ground end of the tickler, and moved the pot to the high end, and the regen still worked. However, it works best with the tickler connected as shown in the schematic, and there is a difference in operation (especially when the AVC is on) depending on which way the tickler leads are connected.
- Increasing the regen to the point where oscillation begins will cause the set to operate in synchrodyne mode. In this mode, selectivity of the IF becomes extremely sharp and it will separate out adjacent channels unless they are extremely powerful. I believe that the presence of AVC voltage helps to regulate the amount of oscillation, and keep the sensitivity up. I notice no discernible decrease in sensitivity in synchrodyne mode. One thing of note is that due to the mainly capacitive coupling of the regeneration, the IF frequency pulls about 20 kHz high when it goes into oscillation. Hence the IF shifts up to about 282 kHz. Oddly, this doesn't seem to hurt anything. Future plans are to move the tickler coil to the cathode side of V2 to see if the proportion of capacitive coupling can be reduced without harming performance.
- The audio filter provides a DC path back to the V2 grid providing approximately 50% AVC, while 100% AVC is fed back to V1. This arrangement has a side effect of increasing low frequency audio response at low volume settings, and better high frequency response at high volume settings.
- The V2 screen grid is connected to the high side of the second IF transformer primary, and the low side of the audio output transformer primary. In effect this gives high gain pentode operation for the IF signals, and lower gain lower impedance triode operation for the audio signals. Compared to having the screen connected to B+, this has less tendency to oscillate at audio frequencies, although the audio gain is somewhat reduced. A milliammeter connected in the screen grid circuit acts as a simple signal strength meter. It's relatively useless when the receiver is in synchrodyne mode.
- It occurred to me after the set was up and running, that there are three different modes of detection taking place in this circuit (sometimes all at the same time). There is of course diode detection from the diode section of V2. With strong local signals, there is also square law detection happening on the IF signal on its first pass through V2. Finally, there is synchrodyne detection when the regen is turned up to sustain oscillation. If I had considered this before I had built the set, I likely would have thought that it wouldn't work and may not have built it at all. However, these three detection modes seem to coexist quite happily.
The antenna (for the 1st weekend of the 1AD contest) was a 60 foot long, 12 foot high inverted 'L' long wire antenna pointing west.
A recent addition to this radio was a 30" diameter tuned loop antenna, consisting of a ten turn tuned loop and a one turn pickup loop. This was completed after the first weekend of the 1AD contest, and was used exclusively for the second weekend. Since the receiver's tuning capacitor has 3 gangs, I used the spare gang as the loop tuning capacitor. The loop inductance was designed to match the RF coil inductance so that they would track properly. Tracking was improved later by the addition of a padder capacitor, but performance was quite good. Overall signal strength was noticeably less than with the long wire antenna, but there was considerably less noise, and the receiver has more than enough gain to compensate. The loop directionality was a major benefit in separating multiple stations on the same frequency.
Headphones are Sony Sport Walkman MDR-W08--certainly not sensitive, but adequate for the audio signal that the receiver puts out.
My thanks to Jack Bryant, Dave Schmarder and Bill Meacham for their helpful advice as I was debugging this set.
Here is the 1AD Superhet before regeneneration was added and before the front panel was crunched.
Before I knew it, the 1AD Contest was almost upon me. I was busy disconnecting my antennas in my radio room. The antennas come in through a metal plate mounted in the window. My shelf in front of the window is stacked with radios. As I finished with the antenna connection, I bumped my1AD superhet. It slid off the shelf, fell three feet, and landed on my 1 tube regen! The front panel of the superhet sliced the RF choke in the regen as the front panel itself crumpled over....bummer. At least it spurred me on to get rid of some clutter! My collection is now several radios and tuners slimmer, and the shelf is not quite so full.
Several days passed, and I was still not ready for the contests. I could see that the regen required surgery before I could use it. I decided to try the superhet. I hand wrestled with the front panel and straightened it enough so that I could turn the knobs. I had previously been experimenting with the superhet, adding regeneration with a link from the plate circuit to the RF coil in the grid. Well, I connected up the superhet, and it worked,! It took a bit of effort to get it working across the whole band, but it was sensitive and as selective as before. This was going to be the 2008 1AD radio. Details on my original 1AD set are at this link: Jack's 1AD Superhet Radio.
I started early Friday morning, working DX stations during the gray line. I decided to keep at it. I started with WSB at 1201. I kept at it until 1612 UTC catching 46 stations, a mixture of local and DX. The regeneration aided reception of weak locals. I took a break during the day and started at back at 2348 UTC.
By midnight I had copied 87 stations. I kept thinking that just maybe I could reach 100 stations if I stayed with it! I was sleepy and tired, but the lure of 100 stations within the first 24 hours of the contest was appealing. Finally I copied the 99th station. It was WHBQ in Memphis. I grew up listening to Elvis, Charlie Rich, The Beatles, Sam and Dave and many other great artists that were on that station in the 50s and 60s. Fourteen minutes later I copied WGY, the 100th station.
I went listened a little bit more that weekend, but decided I would trim my log down and just go with my count in the first 24 hours, 100 stations. It sounds like a good target for me for the next contest.
When I told Jack Bryant that I had an idea for a OneAD set using a JFET LO and a ring mixer, he said he wanted to encourage me to build a regenerative set instead and would even sell me the parts cheap. Time was short, only two weeks until the OneAD contest, so when the BC-221 and some other parts arrived, I had to begin a crash effort to get it done. I slavishly followed his layout and construction techniques, because I know regen sets can be pretty finicky about layout.
Even though I got my ham ticket in 1954, when a tube regen set was still the receiver project in "How To Become A Radio Amateur", I had never actually built one. I had operated one and was not impressed. The excellent tuning capacitor and vernier dial drive in the 221 is the big factor in this set's performance. Virtually everything else went into the trash. The 12 volt tube was a plus, since I didn't need to build a high voltage power supply.
It all went together with no problems. My antenna for the contest was about a 40 foot wire inside the house and an 8 foot ground stake, the same as I used in the Crystal Set Contest, but the best I could do in the allotted time and in my covenant restricted neighborhood.
It took a little while to get the hang of regen set operation. The picture shows the BC-221 based receiver and a box next to it that says "Frequency Counter Controller". This is a DDS signal generator. I injected a little signal from the DDS into the antenna lead using a toroid coupling transformer. I could zero beat stations and determine the frequency exactly. That was a big factor in identifying stations. Please note that the DDS was switched off when tuning so as not to be guilty of "exalted carrier" reception, however I found that it contributed nothing to the sensitivity of the regen set.
The sound power phones were found on E-Bay, manufactured by Dynalec. My ears, manufactured in 1941 by my mother, are not very sensitive anymore and seriously limited my listening. But changes in propagation and station antenna patterns actually allowed me to hear two local stations only 10KHz apart. Jack was impressed.
In conclusion, I had a great time, but the antenna and my poor ears resulted in not many points. I'm looking forward to a new antenna and next year's contest.