[AMRadio] FW: Homebrew receiver

[Brett Gazdzinski]

This is what I plan on sending to Electric Radio, along with
pictures.
What do you guys think?

Brett
N2DTS



 I wanted a complete home brew station, and since I have
 a homebrew pair of 813,s, modulated by one of two modulator
 decks, push pull parallel 100TH,s, or a pair of 4sc250b,s,
 and a classic push pull rig with link coupling, using 812,s 
 modulated by
 a pair of 811,s, only a receiver was needed.
 At first, I thought I would build something simple that worked
 just well enough to be able to copy AM under good conditions, just
 so I could say I had a home brew station.
 But I wanted something a little better than the regen receiver
 type of radio, maybe a simple superhetrodyne.
 I did loads of research, looked in Bill Orr, and all my old
 ARRL handbooks, looking for simple receivers.
 All the circuits had some sort of problem, complex tapped coils,
 hard to get parts, poor designs, etc.
 I also looked at the diagrams for things like my Gonset G76, the Scott
 model SLRM I have, the Hallicrafters sx17, and the R390.
 
 I decided to base the receiver on the Scott SLRM, since it works
 very well, has good fidelity, uses 8 pin tubes and a 455Khz IF.
 I ran into problems though, as the Scott was built to reduce emissions
 out the antenna, with loads of shielding and an rf amp with 
 tuned circuits.
 
 I accumulated parts, and started construction with the basic layout
 of two tuned circuits on the antenna input, an RF amp, a separate
 local oscillator and mixer, two stages of IF amplification, hifi
 detector, s meter circuit, agc circuit, and power supply.
 
 Since it was to be experimental, I used octal sockets for everything,
 the antenna coils, the local oscillator coils, and the IF 
 transformers.
 The receiver started out with plug in coils to change bands.
 
 I laid out all the parts, leaving room between things to allow
 room for experimentation, and mounted the basic parts.
 
 I tried various circuits for the local oscillator, using coils
 wound on ceramic forms, B+W coil stock, and slug tuned ceramic
 coil forms.
 This step would have been very difficult without the aid of a 
 very nice  spectrum analyzer I have through work. It allowed me 
 to look at the  frequency output, harmonics, hash, drift, 
 frequency range, amplitude, all at the same time.
 
 At first, I went with plug in coils in the local oscillator, 
 used the rf amp,  using the spectrum analyzer to peak things 
 and check gain.  The mixer was easy, then to a filter.
 I planed on using a mechanical filter, but they are 
 expensive, and a little tricky to put in the circuit.
 I found a company on the web, kiwi, who makes various filters, and
 went with one that has an op amp input, three filters of slightly
 different center frequencies (sets bandwidth) and an op amp output,
 and runs off 10 to 30 volts dc.
 There is no loss through the filter, and its quite similar in results
 to a mechanical filter. I used a 5.5kc model.
 It mounts on Velcro, and has pig tail shielded wires to hook up
 to the IF system.
 This filter is easy to add to any receiver using 455 KHz as an IF,
 and really works fantastic.
 
 I copied the IF system out of the Scott, and used a hifi detector
 on one of the AM web pages.
 It took some experimentation to get the agc takeoff and IF gain
 control systems working well, then I added the S meter circuit I stole
 out of the Bill Orr handbook using a 6SN7.
 
 Taking the receiver for a test drive revealed problems.
 Startup drift was excessive, muting the receiver seemed impossible,
 the RF amp caused all sorts of problems, and the if amps were 
 unstable.
 
 As a test, I hooked the antenna up to the mixer input, and bypassed
 the rf amp, and had very good results, so I removed the rf amp
 completely, and went with two tuned circuits then into the mixer.
 Some experimentation with the antenna link on the input coil boosted
 gain quite a bit.
 I ordered a selection of NPO caps, and did weeks of experimentation
 on the local oscillator stability, changing components, 
 design, putting the coil in a metal plug in can to shield it, and 
 got the stability much better, but still have startup drift for 
 the first 5 minutes.
 
 Careful shielding and reducing the gain of the IF eliminated the odd
 oscillations I got at times, and the receiver was working quite well.
 
 I did not like the tuning dials I had, marking the frequency was hard
 with the drift, and I have a real problem marking the frequency
 so it looks nice on the dial.
 I needed something better, and found the almost all digital 
 electronics digital frequency readouts, basically a frequency counter 
 with a selectable frequency offset.
 You program the thing to offset the IF frequency, in my case, 455Khz
 lower, and all you need to do is get the pickup close to the
 local oscillator tube, and the display reads the exact 
 receive frequency down to 1000 Hz.
 I used their backlit display, which looks nice, and a real accurate
 frequency readout is very nice to have.
 
 The performance of the receiver was astounding!
 With the transmitting antenna used, sensitivity was very 
 good, fidelity was great, I use a marantz amp on all the receivers
 in the shack, to a big three way speaker, and the homebrew sounds
 the best, because of the low distortion fi fi detector I guess.
 The biggest surprise is the noise level.
 Since the tube count is low, and the mixer design is a quiet one,
 the receiver is incredibly quiet.
 Its MUCH quieter than anything else I have, or have ever had.
 Forget the modern rigs, the IC chips just can not run quiet, and
 there are so many of them in modern rigs that the noise and distortion
 in any modern rig I ever used is way high.
 Comparison to my very well working r390a was dramatic, I could CLEARLY
 hear signals that were well under the hash level of the r390a,
 the signals were unreadable on the R390a, but very good 
 comfortable copy on the homebrew receiver.
 
 After the results I got out of the homebrew, the plan changed from
 something I could use sometimes, just to have a complete home 
 brew station, to the receiver of choice.
 This caused problems.
 I had the receiver mounted in a rack cabinet, and had to run around
 back to change the plug in coils, a real pain in the butt over time.
 So out came the receiver, and a new front panel and band switching
 was added, along with 160 meters.
 Tuning was changed to a system using TWO back to back vernier
 drives, the tuning range was changed to cover only part of the
 ham bands, giving very nice slow tuning range.
 A bfo was needed for zero beating AM signals, so I found and built
 a 455Khz crystal oscillator circuit, with a variable output 
 level by way of a pot in the screen voltage.
 The level control is on the front panel.
 The bfo also allows me to copy cw and ssb quite well, without
 a product detector, so I can listen to the ssb guys complain about
 AM.
 The receiver moved into a cabinet on the operating desk, and 
 was integrated in the shack with muting and so on, and is 
 the main receiver now, the others are almost never used...
 The only problem the receiver has, and it does not bother me, is
 the startup drift. From a cold start, it drifts about 1kc
 over about 5 minutes, then is rock stable.
 This might be due to the choice of octal tubes, the actual tube
 used effects the drift quite a bit.
 Experimentation with npo caps can reduce the drift, but it starts
 drifting the other way over longer periods of time, and I think its
 better to have 5 minutes of drift and stop, rather than drift less
 but over longer periods of time.
 
 
 
 I was quite surprised about how easy it was to build, and how much
 raw fun it was to design the thing, and do all the testing
 and development.
 You sure do learn a lot when you build something step by step, without
 any overall design to start with.
 Every system must be analyzed, built, tested, changed, other 
 things tried, etc.
 
 The end result looks a little rough inside, as it was 
 changed quite a bit, deleting the RF amp, adding band switching,
 etc, but it still looks ok.
 Its been totally reliable and stable for about a year now, with
 quite a lot of use. 
 
 My next project is a superhet receiver using 7 and 9 pin tubes, using
 things I learned from the first one:
 
 Start off with band switching,
 Forget the RF amp, its not needed on the low bands at all,
 Do NOT leave a lot of space between things, but put the tube sockets
 and IF cans close together as possible, along with the local 
 oscillator parts and band switch.
 
 It will also include two filters, 4.5Kc, and 5.5Kc.
 The 5.5 was a great overall choice, but a 4.5 will help
 when things get crowded on the bands.
 
 Building a good receiver for AM reception is not as hard as most
 people think, and I encourage people to give it a try.
 
 I have no formal electronics background, all I know I got out
 of books and by playing around, so if I can do it, almost
 anyone can.
 
 Parts are not a limitation, although it may take some time to assemble
 all you need at a reasonable cost.
 Things like IF cans can be got out of old tube radios, old table
 top AM radios are a good source of parts, as well as mouser 
 electronics, Antique electronic supply, ham fests, even radio shack.
 
 
 You may find you can build something better than anything you can
 buy for almost any price, as YOU pick what is important,
 I only wanted part of 80 and 40 meter coverage, low noise, and hi
 fidelity, along with reasonable frequency resolution.
 
 Old tube receivers like the Scott SLRM, SX17, SX28 can be quiet and
 hi fidelity, but lack frequency resolution, good filters, and cover 
 more bands than I need.
 Newer tube receivers like the R390 series, the Collins 75a series,
 the National nc300/303, Drake and others have some good points,
 but lack fidelity, bandwidth choices, look ugly as stink, or have some
 other drawback.
 All new ham equipment seems to be very high in noise and distortion,
 and you may THINK some of that stuff sounds good, until you
 compare it to a good AM signal through an old tube hi fidelity
 receiver like the Scott SLRM or the SX17.
 Even with output from the detector into a good hi fidelity 
 amp and speaker, there is no comparison between the new and 
 old stuff on AM.
 I integrate all the receivers into the Marantz amp, and I can 
 jump between various receivers quickly, all tuned to the same
 signal, and the difference is dramatic.
 I have tested many, Kenwood ts440, icom 735, Kennwood r1000, 
 IC 756pro, and others, and they are all poor receivers for AM
 if you want fidelity.
  
 Brett
 N2DTS