Skip to content

High Quality Broadcast Band Radio

February 20, 2020

This article replaces a previous one. The back story is that I am a “senior citizen” piano student working for the ABRSM Grade 8 exam (final). It’s a major challenge at an age where not only cerebral memory but muscle memory is fragile and I’d practised myself pretty much to a stand still. I needed something to chew on and having acquired the major parts for a radio years ago, decided to build one and this is the result.

This was a good opportunity to learn a little about RF tube electronics and this is the first radio that I have constructed. My previous RF experience is limited to restoring a Bush AC91 table radio, a Halicrafters SX-71 receiver and a Heathkit SB-2 transceiver plus various items of radio specific test equipment. The broadcast band is badly polluted by RFI, presumably coming from the proliferation of switched mode power supplies, computers and other broadband noise generators. The noise actually made attempting to use the radio annoying. I did a little research on the net and discovered the Tecsun tuneable broadcast loop antenna. It works extremely well. The trick is to scan the band slowly with the radio, following with the antenna tuning, rocking it gently to help pull in the transmissions. I am now mostly using it to listen to internet radio (largely BBC) using a FCC Code 15 legal AM transmitter that I built years ago, akin to tube era BlueTooth.

I think the way I’ll structure this article is to start with such features as it has, then lots of hopefully interesting pictures with some notes about construction and get into technical depth at the end. First some technical terms. Heterodyne is the term used for mixing two frequencies to get a sum or difference frequency. Superhet simply refers to this process at supersonic frequencies. A superhet radio is tuned into the desired receive frequency and presents this to a mixer or frequency changer. A “local” oscillator provides a second frequency that tracks the tuning of the receive frequency such that a constant difference frequency is available at the output of the mixer, this being referred to as an intermediate frequency, IF. The advantage is that the signal can then be amplified by a amplifier stage that is designed to operate at just one frequency and is thus equally effective across the entire band that the radio is designed to receive. AGC refers to Automatic Gain Control that helps to level the response of the radio to strong and weak signals. It also helps to compensate for the coming and going of the signal due to ionospheric activity. Detection refers to separating the audio signal from the AM intermediate frequency.

It is a five (valve or tube) plus one superhet, the plus one being the inclusion of a tuned RF stage having user variable gain on the front end. The coverage is 550 to 1600 kHz, known to the British as the medium wave. I also decided to include a magic eye for fun and with a tube rectifier, this accounts for the “Super Seven” moniker on the front. The superhet circuit is generic except, I split the AGC and detection circuits (the 6AT6 tube has two diodes) and provided an additional germanium diode, doubler fashion, in the AGC circuit to increase the available control voltage. The AGC may also be switched off and thus holding the radio at its maximum sensitivity. The other area where it departs from the norm is the output stage which is not only triode connected but uses cathode  coupled negative feedback. There is so much gain that I was trying to lose some in the AF stages and it made sense to do this in a way that increases audio quality. The tube lineup is, 6BA6, 6BE6, 6BA6, 6AT6, 6AQ5, 5AR4, 6EC5.

So to the pictures somewhat in order of construction:

The chassis is made of 1/8th thick copper clad GRP, soldered together. I found my Hacko iron was powerful enough to run a fillet along the inside joints.


Here’s the bent plywood case and wood sub front panel:

Wooden Sub Panel

The case was made of 3mm plywood on a former. I cut kerfs using a pocket sized table saw from Harbor Freight to aid the bends. It is shown standing on a sheet of 1/4 lexan which became the base.

Cabinet Former

Trial kerf cutting:Cutting Kerfs


Here’s some details of the tuning drive and magic eye:

Magic Eye & Slow Drive

Slow drive and back light LEDS:Dial & Slow Drive

The slow drive spindle is an adapted Stanley quick drill chuck adaptor. I used this because it is rubber covered which helps to reduce slipping of the rather crude rubber O ring drive belt:Slow Drive Spindle

Initial calibration (more on this in the tech section):Initial Calibration and Magic Eye

The final dial. The numbers appear behind a fixed cursor in a viewing port, pre WW2 style and are backlit. The dial itself is simply a disc of paper attached to an acrylic (I think) disc.Dial

Dial aperture and magic eye

Here’s the design of the aluminium front panel, exported from the Front Panel Express design program:

Panel PDF from Designer

And so it came together:

Aspect View

I affixed the schematic to the top which being hand drawn adds interest in my eyes.Top with SchematicRear View

Here’s the underside again, this time as best I could, through the lexan base.Lexan Underside

The design is inspired by the Angelfire 4 tube superhet but frankly such radios are pretty generic. An aspect of the design that differs from the “American Five” is the provision of a tuned RF stage on the front end. A stated reason for this is to improve AGC action by applying the AGC line to the RF stage also. I had intended to include an RF stage so the Angelfire design was most helpful as a reference. Another good reference is the circuit 22-3 given in the RCA Receiving Tube Manual. My interpretation includes variable gain for the RF stage by the simple expedient of placing a 1k pot in series with a 100 Ohm resistor in the cathode. It also has a “top cut” tone control, which really does help in noisy conditions. A significant modification is that I took advantage of the two diodes provided in the 6AT6 by separating the detector and AGC rectifier. This gave an improvement in audio quality. Further, I added a germanium diode to the AGC rectifier circuit, doubler fashion, to increase the AGC voltage. (I used a 1N277 simply because I had one in my sand box.) I found that even with the RF stage, the AGC action did not please me. It’s pretty much a parts bin special. Years ago, I had purchased 460kHz IF transformers from RadioJohn on eBay. In addition, I had a suitable 3 section tuning condenser, power transformer, choke and output transformer, also speakers. The antenna, RF and oscillator coils came from Antique Electronic Supply. The recommended load for a 6AQ5 in beam tetrode mode is 5k and my output transformer is 5k into 4 Ohms which fits nicely with the two 8 Ohm speakers in parallel. (I now have the 6AQ5 connected as a triode with cathode coupled negative feedback, there’s more than enough gain.) Initially, I went with a choke input power supply, otherwise the voltage would be too high. Using a 5AR4, the B+ was around 245V. Later, I added a 1µF input capacitor and changed the rectifier to a 5Y3 resulting in around 250V.

Setting it up was a challenge to me. Performing an alignment on a working radio is one thing, setting one up from scratch quite another. The deal is to get the oscillator to track the tuned frequency such that the IF frequency (460kHz) remains constant across the whole band. This is very important to maintain good sensitivity across the band. It’s no good being tuned to a frequency if the IF frequency is off. The general approach is to tune things at the low end (550kHz) using the coil slugs and at the top end (1600kHz) using the padder capacitors that are provided on the tuning condenser. I found that I could measure the local oscillator frequency with my frequency meter by just placing a probe near the oscillator coil. This is good because it minimises any loading of the oscillator which would affect the frequency. Similarly, I put a couple of turns of wire around the connection from the secondary of the RF coil to the B section of the tuning capacitor and using a scope could see precisely when the RF circuits were peaked at a given frequency. Knowing the tuned frequency, I could tune the oscillator to cause the correct IF frequency, then tune the IF transformers using a meter on the AGC line, looking for the voltage to peak (negative). Fiddling with the oscillator padder capacitor helped to get the tracking right but it took a lot of iteration. I was checking the oscillator and tuned frequencies at 100kHz intervals. I kept at it for a bit and the result was quite good. I really can’t suggest a technique to get everything in the ball park when starting out though, other than patience, observation and thought.

I should point out that since I separated the detector and AGC, the second IF transformer is only partially in the AGC act, so final IF alignment is best performed by measuring the audio signal level from a modulated test signal at the output stage. I found it better to do the RF and oscillator alignment with just a carrier.

Just for completeness:

From → All Posts

  1. As always, your descriptions and projects are top notch!

  2. Kevin Kennedy permalink

    Very cool, love the look, and I’d guess performance would be on par with really good 1930s vintage superhet broadcast receivers which would be about the pinnacle for such things.

    • Hi Kevin!
      Yes, that sounds dead on as far as I can tell. Backwards into the future I go……

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: