PART 3 - RF Board Construction
The whole original concept of started out based on the BITX, a 20 meter all transistor transceiver, [11] shrinking it down using SOT packages, so there are some carryover circuits from that design. The R/T switching was based on the WA7JHZ 75 Meter Transceiver project design. [12] (Which, by the way, is intended to provide “…several hours of quality entertainment”) The clincher to go ahead and use integrated circuits, such as this design used, was that a quick supplier check found a version of the venerable SA602s were still available, even in SO8 packages, just in case I need to go to smaller versions. The interesting thing to me in that design was that it used the balanced inputs as separate single ended paths for both receive and transmit sides. It really saves on the switching, since it is all mechanical now, although the gains of each stage are cut effectively in half. The LM386 is also still available and in SO8 packages too. The switching was done in that design by mechanically connecting several slide switches together to form a multi-pole switch. Finally, a search turned up a fairly common 6PDT version of the E-Switch PBH series switches [P/N PBH6UEENAG1DBLK] were available and in stock and no modifications to slide switches needed.
The original proto soon showed that the receiver’s gain was too low. I needed more receive amplification so I based the RF front end amplifier on the W6IBC design [13] with an 8-MHZ IF [14] and also added an audio preamp. The TR switching was simplified somewhat, based in part on that design, but still using the same mechanical switch, versus the electronic TR control. The gain distribution was corrected and it now had full speaker volume. I have not had any audio howl as in the first design either. Sensitivity seemed OK; I heard all the usual suspects on 20 and even picked out some DX stations with my station’s vertical antenna attached to the proto. It looks like about 90-100 dB of receiver system gain. There is a bit more receiver noise (hiss) than I would like but the result seems OK. I used two 7mm IF can transformers for each of the two band pass filters. This proto has a reasonable response for 14MHz and the design has been tested by simple capacitor value changes up to 30MHz can be fine tuned with the transformer adjustments and take up very little board space. I will need to re-visit the design especially for the 50MHz version, perhaps with smaller shielded inductors and trimmer caps. The squelch circuit was probably the only real original addition to this design, the comparator circuit uses the signal strength indication through its own audio amplifier and rectifier and chain to control the power to the audio amp. It uses the “leftover” comparator intended to delay the transmitter audio circuits, and a voltage divider formed by the squelch control pot.
The second proto started by building the receiver and transmitter circuits “ugly style” on both sides of a double sided PCB scrap. Although the proto started to come to life in this form it was soon abandoned. Not only would never be able to fit in the case but would not be an accurate test bed for any of the layout’s interactions separated by the ground planes. The actual PCB layout was now almost completed so I started the third (and final) proto using the "pretty ugly" technique which used the PCB drill file as the layout template. The receiver and transmitter circuits were squeezed pretty tight to fit combined on that form fitting board. As the circuits were built and tweaked the PCB was finalized and completely routed. I guess I was a bit cautious by building three prototypes, I was skeptical the whole thing would even work packed so tightly on one board. Looking back, it probably would have been OK to simply make the PCB. The major circuits changed little, and the layout worked, but the added cost of proto boards made me want to prove out the concept first before committing.
Building this board is relatively straight forward. There are a large number of components involved so I would recommend a divide-and-conquer method of building and tuning in stages or functional blocks. It makes life easier to correct any small mistakes as you go along instead of trying to track one down when completely built. I usually start with power circuits. I find it a little easier to pre-tune the filters before the rest of the radio components are in place, but that depends on what test equipment is available, these are also be tuned for maximum signal after the whole radio is assembled anyway. I usually start with the receiver circuits, from the antenna to the speaker, tuning as I go along for signal strength and finally the BFO for intelligibility. Then adding the squelch control comparator last, since it will be always on for testing without it in place. Followed by the transmit side circuits. These are only tested for maximum signal output by first using an intermittent injected CW or speaking into the mcrophone. The final adjustments are done with the LPA in place. I then finish the transmitter circuits by adjusting the mic gain and touching up the BFO for the best signal while monitoring the output on another HF rig. It may be a good idea to drill a small access hole in the case over it to do final adjustments with the case on. I would guess similar holes could be made for the BPF and BFO tuning adjustments, if desired, although I have never done so. The microphone and LED can actully be mounted on the back (or solder side) of the board. This saves on routing as they can be lined up directly to holes on the front panel. I would recommend carfeully measuring and drilling the holes before final assembly! I also found since the RF board will be directly under the speaker it can be tacked, with a small dab of hotmelt glue or RTV, to the back of the board just behind the speaker also makes an acceptable arrangement, without the need to make a dedicated microphone hole, the speaker cone is transparent to sound.
