2007-01-07
VXO 80 metre receiver
I recently emailed Hy-Q asking what range of ceramic resonator devices they stocked. They were very helpful as usual, but only carried 2 items as stock that were of immediate interest: 3.58 and 3.68 MHz. I purchased 40 of each for experiments (they are extremely cheap), along with some other assorted crystals.
I started by building a VXO circuit, to see just how far these things could be pulled. It was found to be quite practical to pull them about 100 kHz and still have a stable oscillator, meaning the two available devices could cover most of the interesting parts of 80 metres.
It was only a short leap to making a full direct conversion receiver from parts in the junkbox. The VXO buffer delivers about 6 dBm, so it was practical to drive a diode DBM directly. A simple two-stage AF pre-amp was constructed and a simple three transistor power amp suitable for driving headphones or a small speaker completed the AF side of things.
With my dipole for 40 metres connected into the DBM RF port I could hear the ARNSW 3.699 MHz CW practice beacon quite clearly, and easily distinguish when the antenna was connected and disconnected. However, I felt I needed more front-end gain, and pre-selection. I discovered that the DBM would harmonically resolve signals on 40 metres quite well. In fact I was tuning around 80 metres on my Yaesu FRG-7700 looking for a QSO I heard on this receiver in vain, only to make the connection some time later and tune up to 40 where it was heard.
The front-end was built with a double-tuned band-pass filter and a buffer amplifier supplying enough gain to more than make up for the DBM losses, giving this receiver a much better noise figure than it actually needs on 80 metres. The harmonic reception problem is gone, and the radio in general performs very well indeed.
You can think of this radio as a "Mark II" version of my old altoids tin 80 metre RX. It has a much superior front-end, similar or perhaps better VFO stability, but far less tuning range. It is extremely simple, as receivers go, and can be put together in a day or less. None of the devices are especially exotic, all except the ceramic resonator are available from DSE, Jaycar or Rockby. It could be miniaturised somewhat by using ferrite beads or T25-43 ferrites for the DBM (or perhaps a commercial mixer package). It is quite practical to build it on a PCB to pack the components closer, it could definitely be built into a altoids tin, even ugly style, if so desired.
I haven't put it in a box yet. I haven't really found a nice metal box to suit it. I am considering building a matching DSB TX and using it along with that as a full 80 metres transceiver. ARNSW is having a challenge this year, to build a crystal locked 80 metre 20 Watt TX, so a 20 Watt DSB rig might be admissible.
Construction Notes
The "rubbering" inductor in the VXO circuit is something you will need to experiment with to get the best range/stability trade-off for your particular resonator device. The properties of the ferrite bead used will drastically affect the stability of the oscillator thermally. It also must not move physically, I potted mine in wax once I was happy with the range available. Without the inductor you can still pull the resonator *much* farther than a colourburst crystal.
I am currently using the oscillator side of the broadcast tuning gangs DSE/Jaycar sell for the frequency control. I may add a 1N4004 "varactor" fine control, but so far the tuning is fine enough to resolve SSB with little drama. The long wires I have hanging off the VXO board to the capacitor let me "fine-tune" by body capacitance - HI HI! Once inside a nice metal box it should be quite stable, and a large knob will make it very usable I suspect. A vernier doesn't seem to really be needed.
Update: Boxed Up
I finally found a fairly nice Aluminum and Steel enclosure for the receiver. It was sitting on a shelf out of place in Jaycar's Sydney CBD office, it whispered gently "buy me"... Anyway, it was a bit of a squeeze to get it all in there, but it fits. There is a close-up of the ugly board for your study if the circuit diagram is unclear. The signal path is folded into a inverted "U" shape, RF entering at the bottom left, passing through to the mixer at the top where it is converted to baseband, then down through the AF strip and out to the headphones/speaker. The LO just hangs off the side and injects its output into the mixer.
The tuning cap 1/4" shaft adapter and mounting screws comes from Doug Hendricks KI6DS. The qrpkits.com site lists them as "Polyvaricon Hardware Pack". In the past I've had *great* trouble locating hardware compatible with the plastic tuning caps available here in VK-land shops (i.e. Jaycar, DSE, and Rockby), so this find was an big deal for me!
I want a switched pot for the volume control, so it can act as a power switch too, but no one seems to carry them any more in Sydney (at least not miniature ones)? I've ordered some in from eBay, when they arrive I might try putting a 9 or 12 V battery pack in the box too, switched through the external DC plug. The receiver will run of as little as 6 volts, but the RF gain suffers and the bias on the AF final is a bit off. The radio pulls too much current - about 50 mA at 13.8 Volts (because of the large LO drive, needed for good strong-signal behavior and sensitivity) for a "216" 9 Volt battery supply.
Operational Experience and Thoughts
The excuse for a Diplexers I put in really sucks. It terminates the mixer IF-port at RF with basically a short-circuit, reflecting the LO out RF-port somewhat more than I'd like, and gives the LO buffer a terrible load to work into. Next time I have the radio apart I think I will modify it, at the very least terminate the RF into a resistor, if not add an inductor to keep it out of the AF section. It still works pretty well considering the fairly horrific design though.
The AF strip isn't too bad considering the design was back-of-the-envelope stuff. I'll have to sweep it and measure its real response, because it sounds pretty good for something so simple. The collector-base feedback cap in the first stage might be a bit too large, I increased it empirically from my calculated value, and in hind-site this was probably needed because of diplexers related issues.
An active antenna for this radio would be handy. I discovered the "pa0rdt-Mini-Whip©" today, I might have a go at building something similar - although it pulls as much current as the radio does! I know this is to give it strong-signal capability, which I probably don't need if I make something narrow-banded.
I've been designing a better AF strip for another project, in which I am considering a diplexers with a cross-over at about 100 kHz. This is fairly high, but makes the inductors more practical to wind by hand rather than using commercial chokes. This means the termination needs to be pretty flat to at least 1 MHz. A common-base amplifier would be a good match I think, the emitter resistance can be set up to terminate the AF port of the diplexers quite well to beyond its pass-band.
A common-base amplifier in the front-end would give the receiver a better defined input impedance as well. Currently, because of the symmetrical nature of the BPF, the relatively high input impedance of the LNA is reflected through the filter and to the antenna connector. This is probably overkill for a low-HF radio like this, however it would improve the reverse-isolation of the amp stage (and considering the current diplexers, this wouldn't hurt). In a VHF radio it would be almost mandatory, but a FET would be preferred for a better noise figure.
9 comments.
Attachments
| title | type | size |
|---|---|---|
| VXO Circuit Source | application/postscript | 14.072 kbytes |
| DBM Circuit Source | application/postscript | 12.241 kbytes |
| AF Amplifier Circuit Source | application/postscript | 15.636 kbytes |
| BPF and Amplifier "Front-End" Source | application/postscript | 13.950 kbytes |