2007-03-24
MK484 MW Receiver
I've grown tired of my regenerative AM broadcast receiver's lack of AGC and its RFI problems, so I decided to build something based on the MK484. I've had them laying around for years in the junk box, and my recent work in using them for an AM receiver IF prompted this project.
Other requirements for this project were a small size, something smaller than the current plastic tub I carry in my backpack to work every day, and something with a calibrated tuning dial to take the guess work out of tuning. I also set myself a challenge to build it using only a single 1.5 volt cell for the supply, and driving conventional 32 Ohm dynamic earphones.
The RF section is very simple, basically the single cell "reference design" MK484 receiver. The ferrite rod is the small rectangular prism unit available from Jaycar or Electus (catalogue number LF1016), I measured its mu-rod at about 65 and calculated the turns required for 400-450 uH or so, required to cover the extended AM broadcast band completely using the 165 + 62 pF tuning gangs also available from Australian electronic stores.
The RF circuit was constructed on a small scrap of veroboard and connections made between it, the rest of the components using PCB pins. At this point the smallest grey ABS box from Jaycar was selected as the enclosure and the polyvaricon tuning gang mounted in the middle of the lid. A knob was selected, large enough for good feel, but small enough to allow a well marked dial to be added later. The adapter hardware from Doug Hendricks KI6DS QRP Kits was again used to mate the polyvaricon with the 1/4" grub screw retained knob. The RF board was "mounted" with stiff wire on the back of the polyvaricon after its trimmers were set for minimum capacitance (a piece of cardboard insulating the track side), and the tuning range customised by removing several turns on the loopstick until it covered from about 538 kHz to 1791 kHz - my overshooting the inductance no doubt a result of the distributed capacitance of the windings. The windings were fixed in place with wax using a candle and the heat gun. The loopstick was then secured to the underside of the enclosure lid with two blobs of 2-part epoxy.
With the RF side of the receiver complete, now came the more challenging bit, an AF amplifier of reasonable performance powered by only a single cell. There have been many "optimised" single-cell headphone amplifiers published over the years, most of them heavily using boot strapping and being quite complex with lots of transistors and large value electrolytic capacitors. I had no where near enough room to fit such a design, so at first I went with a classic two transistor line up that fits on a 4x10 hole piece of veroboard.
In this circuit DC passes through the speakers, and despite some careful design to improve its performance compared to the similar circuit in the Regenerative FM Receiver, I found its performance fairly poor. It offers sufficient gain and undistorted output for a quiet room, but it is completely unsuitable for the kinds of environment I wanted to use the radio in, on my daily commute to work. Its response rolls off at about 15-18 kHz, and the LF corner is quite good due to the largish coupling caps. From 3 volts up, such a circuit is quite usable. If you put a 1K:8 Ohm transformer in the final collector circuit and modified the feedback a little it might actually offer quite acceptable performance.
After using the radio with this Mark I AF amplifier, I decided to modify it for better performance. I picked a transformer coupled circuit to get the best output possible with the low supply voltage. Unfortunately this meant replacing the AA cell with an AAA cell to make sufficient room for the audio transformer.
The emitter resistor bypass on the final transistor is probably redundant, it varies the gain only slightly, you can omit it to save space if you wish. The emitter resistance is important to minimise distortion however. The LF corner is somewhat worse than the original circuit, and the HF response rolls off a little earlier, but the result sounds quite natural. Most importantly there is much more power gain available, delivering ear-bleeding levels at the onset of objectionable distortion.
The finishing touches were some labelling using a Dynamo label maker, and the creation of a nicely laminated dial for the tuning knob. I used the XYLs Xyron 510 machine to make a piece of cardboard into a sticky-backed label which was pasted into place. This was marked with the assistance of the signal generator (and off-air signals). This was then peeled off and a final version made using the original as a template. The final version was run though the Xyron machine again using a laminate cartridge to produce a very pleasing glossy result.
Notes
The wall of the plastic enclosure is too thick for the 3.5 mm stereo socket mounting thread. I ended up epoxying it into the wall. With some care I guess you could thread it into a slightly undersized hole, but gluing it right into the hole is much less mechanically challenging, especially if you didn't notice the problem until after you drilled the hole like I did!
As usual, a switched pot would avoid the need for a separate power switch. It would also make the radio more resistant to accidental turning on while it is knocking around in your bag. If only suitable switched pots were a still available from Jaycar or DSE...
The selectivity could be better, 2BL (702 kHz) is an enormous signal at my QTH and can be heard between channels near the centre of the tuning range. It is not all that objectionable, except when trying to chase DX at night - which was never a design requirement. I probably should have used Litz wire for the coil, or tried using one of the commercial pre-wound coils (like Jaycar catalogue number LF1020). Such coils are designed to match the tuning gang I am using, but past experience has suggested their precision leaves a lot to be desired, often being incapable of tuning the entire band even when moved physically on the ferrite bar and using the trimmers on the tuning gang. They are a cheap source of an approximately correct coil that can be tweaked though.
A Q-booster using a simple FET circuit could tighten the selectivity, but it would chew more power and isn't really required for local station work. I am just being picky really, the performance as-is exceeds most bargain-store superhets.
The Xyron 510 machine is a damn handy piece of kit. It can place adhesive on the back of any thin, fairly flat, object, or laminate both sides, or one side and put adhesive on the other. Cartridges are available with a choice of re-positionable or permanent adhesive, it can even make fridge magnets. Its main target market is the scrap booking community, but homebrew electronic hobbyists would find it very useful for panel work.
6 comments.
Attachments
| title | type | size |
|---|---|---|
| RF Board Circuit Source | application/postscript | 13.346 kbytes |
| AF Board (Mark I) Source | application/postscript | 13.053 kbytes |
| AF Board (Mark II) Source | application/postscript | 14.849 kbytes |