THE PHOENIX RECEIVER
                                              Evolution of a Valve Communications Receiver Project
I had at last decided after many years to build the PW 1961 Project "An Amateur Communications Receiver" I had built this receiver when I was aged 16yrs and amazingly got it going! (after a fashion that is)

And as the years went by and I gained more knowledge I thought it would be a nice nostalgia trip to build it again as the original had long gone, cannibalized into other projects, and it was also a chance to try my hand at building a dial cord dial system, something I’d never tried, AND of course a chance to build yet another Valve Project!! I have always owed this old article a great deal as not only did it teach me an awful lot about valves and their circuitry it also set me on the path of electronics which was to stay with me and provide me with a good income over the rest of my working lifetime.I had been working on the receiver metalwork for sometime and finally got to the stage where the dial cord drive was almost complete, there wasn’t a dial drive on the original article as such, just a reduction drive.
So I spent quite a few hours experimenting, Dials and drives on Homebrew projects have always been a stumbling block, I hate spending lots of money on now hard to find dial drives, and yet a poor looking drive really puts a homebrew project into the "also ran" class no matter how good its performance, each time I had made a drive as you can see from my other projects its been a matter of original thinking and economy, the heart of home brewing in my opinion J and of course it has to look at least reasonable! This time I thought I had cracked it, with a few simple bits of angle aluminium and some laminated print outs from the faithful PC I hope to have produced a good working dialdrive, stringing the dial drive nearly drove me mad!!! stringing a dial drive is not easy at the best of times but stringing one from scratch that has never been done before is a nightmare. I intend to write up the whole proceedure at a later date for fellow dial drive homebrewers..
And now with most of the hardware mounted and PSU wired up and tested, It’s Time for the interesting bit, the VFO the heart of any receiver, I had intended it to be as the article design a full coverage 1.5 to 30mhz general coverage, the original design used a coil pack which is no longer made so undaunted I set out to build one, I had a set of coils removed from a very old and defunct marine transceiver and had fitted all the oscillator coils and circuitry and started measuring the frequency spread. and stability of the VFO,

About this time I started having serious doubts about the original design, it was a basic single conversion superhet with 3 IF stages to obtain the selectivity And I knew from previous experience the front end image rejection was going to be pretty poor although when I built it all those years ago it was (as far as I was concerned then) rocket science J I was pleased to hear anything on it, let alone worry about image rejection!! But now older and wiser and considering how much work was going to go into it, I reluctantly decided to alter the design into a double conversion superhet, This now meant a change in the VFO frequency from 455khz to about 1.6mhz above the incoming signal, but try as I might I just simply couldn’t get full coverage with the 315 pf 3 gang variable tuning capacity that I was using in my dial drive, so having a fresh start I decided to just fit in just as many amateur bands that I could, again this was a problem as it meant either losing Top Band or 80mtrs, also 40mtrs was far to cramped, just a quarter turn of the main tuning cap for total coverage!! Yet again time for a complete rethink, to be honest to myself, as a general coverage receiver I probably wasn’t going to use it much, better I suppose to make it into something I’m really into, even if it makes it a bit unusual, and just keep the bare bones of the original design.
SO I decided to make it a Three Band Amateur radio only, Top Band 80 and 40mtrs, with balanced tuning so the whole scale was used for the band selected, this way I could use the ladder filter and other refinements I was thinking of adding to good effect on the CW ends, but with full coverage though the SSB end as well, not forgetting the new extension to the 40mtr Band,A few tests and checks and I soon had the VFO covering perfectly the selected frequencies needed, almost as if this is what had been intended all along. Time for a tidy up and a few stability checks and its then on to the 1st Mixer wiring,BIG problems the mixer on Topband at 1.8 to 2mhz did not take kindly to an IF of 1.6mhz and also I realised a double conversion on Topband with a 1.6mhz 1st IF was not really worth having,

So I decided that when Topband was selected it would just be single conversion at 455khz by setting the VFO to a 455khz shift and switching off the 2nd conversion oscillator and using the 2nd mixer as another IF stage at 455khz, so all set, now for the RF amplifier to my amazement it worked perfectly first time, no instability and nice smooth RF gain control, By using another receiver set to 1.6mhz and thus acting as the IF and rest of the receiver I could test out the whole of the front end, and it worked well on 80 and 40, Topband I decided to leave until later.Well, I say it worked well but something was not quite right, it didn’t sound right, the CW note sounded very slightly distorted, not really enough to worry but irritating, also the SSB sounded a bit gravely I thought at first it was a bit of ripple on the HT power supply but that checked out ok, after much testing and checking I finally come to the conclusion the mains transformer field was reacting with the dust iron cores in the VFO circuitry and Frequency modulating the oscillator, this was bad news indeed, as it meant a complete redesign of the layout, and or a change of mains transformer type, Time again for one of these major rethinks which simply have to happen on what was fast becoming an original design, Another alternative was an external Power Supply, this was quite common in early Amateur valve equipment many manufacturers did this Heathkit and KW and the like it does have other advantages, but somehow I always thought it a bit messy, however it meant another matching case and that could be interesting, I could even mount aspeaker in it, so external PSU it was, and out came all the PSU components onto another chassis and straight away it was lovely T9 notes all round…
Time to move on to the 2nd frequency conversion, this proved to be fairly easy, the only difficult bit was making an IF transformer with a frequency on around 1.6mhz as that was going to be the 1st IF, economy in mind I again used what I had, a 2.08125 crystal in my junk box would serve as conversion crystal, I also had some 461.5 khz crystals which I hoped to use as a ladder filter in the 2nd IF so the simple calculation 2.08125 - .4615 = 1.61975 giving the 1st IF and adding this to the VFO gives the VFO operating range, ie for 7.0 to 7.2 mhz gives 8.61975 to 8.81975, I used an ECF82 as the triode pentode mixer and it fired up perfectly without any problems, 

Time now for the 2nd IF strip at 461.5 khz, this frequency is very close to the standard valve 455 khz IF transformers and I had no trouble re-tuning them to cover this frequency, experience had taught me that the wiring up on a high gain IF strip has to be very carefully done to avoid instability and sure enough after wiring up all the IF transformers the whole lot went completely unstable, time to add a few more earth points and a 100k resister across the primary windings of the Transformers, this had the effect of flattening the tuning peaks and thus reducing the gain a bit, I should be able to recover this later with the filter in place in the mean time I put two transformers back to back linking them with a 50pf mica capacitor and so now had Eight tuned circuits at 461.5 khz and the resulting selectivity was quite good, I tested this by just fitting a diode across the last IF transformer secondary and connecting it to a small amplifier I couldn’t resist at this stage trying out the ladder filter design I wanted to include in the finished radio, I had not seen any circuits for a low frequency one like I hoped to use so set off by just linking the crystals into a ladder and trying different coupling and resonating capacitors, I found by trial and error that just three crystals was quite enough to provide a very sharp smooth filter for CW use, and switching it out provided an acceptable bandwidth for SSB, Time to move on to the Audio stages BFO and detector, I had intended to use an ECL80 as the triode pentode Audio amplifier, in the original article the designer had used the triode as the BFO and the reason for this soon became clear as getting an ECL80 with its common cathode to work reliably was very difficult as it tended to go into very low frequency oscillation, so I used a separate twin triode valve ECC81 to provide the audio pre-amp and triode BFO, Using a 455 khz IF transformer as an oscillator transformer proved to be a very easy and produced a nice stable BFO and the separate triode audio pre-amp cured the audio problems, time now for the detector, I had previously in another project used a simple product detector, I was so impressed I decided to use it again in this project and at first all seemed well, time to now to align all the circuits and air test, Aligning the now quite complex radio was fairly easy, using a signal generator set to 461.5 khz applied to the 2nd mixer I carefully tuned all the cores in the IF transformers the BFO being set to produce about a 1khz note, after tuning for maximum I then switched in the ladder filter and by varying the BFO above and below the central IF frequency I could by careful adjustment of the filter coupling coils get a nice linear curve to the filter passband sides, at this time I noticed a very odd problem, the BFO tended to jump in frequency when close to zero beat with the incoming signal, after a few hours of checking and rechecking it appeared that the product detector was trying to synchronise the BFO with the incoming signal, I found the cure for this was to simply reduce the levels from the IF and BFO into the detector, then I set the signal generator to 7mhz and adjusted the VFO until I could hear a signal and tuned the cores of the 1st IF transformer for maximum as this must be the correct frequency for it, then I reset the VFO for the correct place on the dial I had made for 7mhz, time now to re-peak the front end and mixer coils for maximum, and also check the tracking over the range, MORE big problems, getting the tracking right over even a small range proved to be quite difficult, I was using small value capacitors in series with the main 3 gang tuning capacitor and in the end I had to use the spare contacts I fortunately had on the band switch to provide different capacitors for each band, Time AT LAST to settle down and try the receiver out on 40 and 80 mtrs
I’ve often felt you can’t spend enough time doing this as it is at this point the character of the receiver starts to appear, sounds silly but I’ve found over the years that all communication receivers have a certain feel about them that you either like or dislike, this particular receiver after a few problems of dry joints and instability was performing quite well BUT I seemed to have both the audio and RF gain controls well up all the time, obviously a bit lacking in gain also attempts at wiring the S meter circuits had produced the same problem, so as I had a spare triode in the ECL80 I thought I might use this to provide a bit of extra IF gain and it proved to be the answer, I had to extend the RF gain control from the RF amplifier into the 1st 461.5 amplifier and reset all the S meter 

sensitivity control, the improvement was wonderful after a few days of listening on 80 and 40 mtrs I had a long think about the spare band, I decided after a lot of consideration to abandon the Top Band idea and go for 20mtrs so the receiver would now be 80 40 and 20 mtrs only, my main bands of interest, fitting the 20mtr coils proved fairly easy but I had a lot of problems with IF breakthrough for some reason? however slightly different coupling to the front end 20mtr antenna coil solved the problem, so now the Phoenix Receiver was almost complete electronically, and performing extremely well, the best valve receiver project I had made so far, as all valve receivers she requires about 30 minutes to settle down and then becomes very stable, also the dial drive works well with no backlash and the SSB reception is very good without any extra filters, the CW ladder filter took a bit of getting used to but after a while I started to like the way it sounds even if it is a bit tricky to get the best out of it As for what remains of the original 1961 receiver, well it still retains the three EF91 valve front end and two of the EF91 IF amplifiers and the ECL80 audio output, otherwise its become another G3YUH special J I like to think an improvement, but that’s where the fun is for me, using my old skills to produce a working example of vintage radio technology, whether it will be kept as that or stripped down and sold for bits as many homebrew projects end up who knows, but that won’t take away the fun its given me in building it and even more using it……..

 

There is still much to do, finish the final case, possibly a paint spray job, make the PSU case etc but the main electronic part of the project is over and I can use the receiver as I finish it.