[The last renewal of this page: November 7, 2010]

I decided to build a single super receiver from 3.5MHz to 4.0MHz, and a crystal controled convertor for 7/14/21/28/50MHz to compose a receiving system.
So for 3.5MHz it is a single super (heterodyne) receiver, and for 7MHz to 50MHz it is a doube super (heterodyne) receiver.

I put the importance on an easiness to build and a certainty, and selected the way to take it easy.

The below is a brief specs.

Receiving Frequency	3.500MHz�`4.000MHz
Receiving Mode	AM/CW(SSB)
IF Frequency	455KHz
Structure of amplifire	RF1+IF3
Selectivity Measure	Mechanical filter for SSB (with a Q-dump rsistor), and a Crystal filter for CW (500Hz)
Detection	Product Detection
BFO	Varialble with a BFO coil
Noise Measure	ANL with diodes
Frequency Calibration	100KHz Marker

This series of equipments started with an idea of "External appearance before everything else", I completed the front panel with only what I have in my mind as a block diagram, before the circuit design is completed.
It looks it works actually, but it doen not. hi hi




This diagram is only for the time of starting work, and it will be amended lots of changes as the actual process goes by. (The final diagram will be shown after the completion.)

I used a 6AK5 for the RF amplifire because it has extremely low inner noise. I used to use a 6BE6 for a mixer in the past, but as far as inner noise leverl concerns, it is one of the worst tube, so this time, I used 6U8 instead. For the local oscilator, I selected a 6C4 which is very easy to use and produces very little heat in may experience, avoiding composite tubes. A constant voltage discharge tube is used in +B circuit, of course.

For filters, I obtained TRIO's old IFT, the highest peak of all the IFT, "T-11" for this receiver. This is because I experienced unbeleivable wide selectivity of a receiver with just three normal IFTs. Of course, this will not be able to give me a satisfaction. So this time, I am going to use a mechanical filter for SSB(455KHz) which I removed from a junk of YAESU FLDX400, and a Crystal filter for CW(455KHz) which I obtained from Yahoo Auction before. To avoid to lose a good audio quality of AM, I will add a Q-dump resistor in parallel of the SSB filter and see the result. If the result is not acceptable, I will repalce it with something different or an IFT.

Considering a large insert loss of a mechanical filter, I designed three stages for IF amplifire. I will adjust at screen grid resistor to let it work lightly.

For the detection, it shoud be a product detection without any hesitation. For this, I avoided 6BE6 again, and used a triode 12AU7, which circuit is very simple and looks nice. I have never had a good experience with a mixer circuit that mixture signal to the grid through a capacitor. The case of 3.5/7MHz protable transceiver was no exception, and I made it my mind never to use again this method. I selected the method to use a coil for BFO, because it is very convenience to be able to adjust the beat frequency especially for CW, and a stability of oscilated frequency is not a question considering the stability of the VFO of receiver.

There are many ways to drive S-meter, but considering the stability of the zero-point, activity of the movement of the meter, I selected the way to use a pair of triodes in a differental detection circuit. This is what I have proved with the 3.5/7MHz portable trasceiver which I built recently.

I inserted a simple ANL circuit. This is a good one which I have confirmed with YAESU FRDX-400, and not only for the noise but also for the audio frequency, which is changed somewhat quiet, it is effective.

To calibrate the frequency, I added a marker of 100KHz. It is nice if I could be divided to 10KHz, but for the time being I think I will try as it is.

The stand-by circuit is not the way to cut off completely during the transmission, but the way to keep monitoring (This is good especially for the CW to get a tone, and to follow the QRH of the VFO). While the transmission, the RF/IF gain is reduced to the set level with a monitor lever nob.

---

So, let's see what's going on!




What I started with was a dial which is the most important "face" of the receiver. I was lucky to be able to find a junk of old MITSUMI's MD-5, a vernier dial with a large square scale plate, so I decided to mount this putting my deep nostalgia in it.
For the AM transmitter and receiver which I build in my student era 40 years ago, I used a LEAD's cabinet with this MD5. I recall everything with nostalgia to melt into teas, hi hi.




The MD-5 that I obtained had a scale plate which is printed divisions for the TRIO's coil kit for all wave receiver. This is nothing to do for me, so I started with modifying the plate. I fristly captured the scale with a scanner, and then using Adobe's Photoshop I drew divisions for 500KHz/full-scale with leaving the most outer division as a protractor.
In the beggining, I tried to print the image to a special paper with a seal in its back, but I could not take away small air bubbles when I tried to stick it to the aluminum plate. So, after all, I printed it to the paper for photograph, and then have it laminated, and cut it exactly the same shape as the aluminum panel, drill holes for screws, and mount it to the dial instead of the aluminum panel. This was well done and perfect.




I put two mini-bulbs for illumination. I am satisfied with a result which is very beautiful, nice looking, with a good atmosphere.




Now, I woked with a S-meter. This is also an important part of the receiver in its appearance, so I did not like to do a careless work. I wanted to use a fifty years old round shape black meter of 300 micro ampere which I obtained from the Yahoo Auction. I loved a scale of S-meter used for YAESU FRDX400 (A square meter you see on the left top is that one), so I captured it in an image file. I also captured a scale of the balck meter and I composed them using Adobe Photshop using an ultra-C technique. It is finished just like a brand new antique style S-meter. I installed two mini-bulbs, and I will never be got tired just looking at it throughout the day!....Oh, is it too much to say so?




I spend a lot of time for the two parts which are going to be the "faces" of the receiver, and therefore, it came out fairly well. I exchanged a nob of the dial to a lager and heavier one. This nob provides much better feeling than before. My heart beats first simply by imaging the movement of the S-meter back and forth.




After that, I fristly tried to complete the transmitter and the VFO, so I got back with the receiver after about one month.
Sometimes we forget about the receiver has a RF circuit, and if you do careless work then it will happen undesired oscilation and feed back. So, I paid a carefull attention the route of signal flow, and the structure of the shield panels.
Firstly, drew a rough disposition with a felt pen.




Place the parts flowing to the drawings, and give a fine adjustment. After the determination of disposition of the parts, put the mark where the center of the parts is placed.




Drill holes with the blade of 3.5mm to where the center mark is painted, and then you will feel refreshed because there will be no way to get back anymore! (No need to worry about disposition or circuit.)
After you finish socket holes and IFT holes, no amendment would be acceptable ! Just go ahead! hi hi..




Mayor parts above the chassis such as tubes, IFTs, BFO coil, mechanical filter, antenna coil, RF coil, OSC coil, output transformer, etc. are mounted. A L/C box hasnot been mounted yet.




I found a suitable cabinet (TAKACHI brand) for L/C box in Akihabara. This compact box just fit to the very limited narrow space of this receiver. I was lucky to be exempted from a tiresome sheet metal processing work.




Now, the back of the chassis. I paid a carefull attention to avoid feed back problems. ( As I spent a hard time with the feed back problem in the receiver circuit of 3.57MHz portable transceiver before.) I examined thoroughly the disposition of the terminals and socket at the back of the cabinet, flow of the RF signal, flow of the AF signal, and etc. I set shield plates from RF amplifire to the mixer, just before the pass it over to the IF amplifire. The shield plate between different stage was set right over the pins of the socket of the tube. Each shielded section is not covered and closed, so it can not be called as a perfect one, but it looks much better than the previous case.(3.5/7MHz portable transceiver) The shield plates between secctions in RF amplifire, I needed to mount VCs being modified to being coupled in series, and I spent a hard time to get corrct alignment. So, taking whole one day, I finished to mount mayor parts in the back of the chassis. Finally, the time for wiring.




Finished wiring for Power supply, Heater line, and Ground line. At this point, I checked the heaters on , voltage check, illuminations of the dial on,and its voltage check.
A 12V DC is required for the relays and dial illuminations, so in the beggining I was planning to get 12.6V/AV connecting two heater terminals for 6.3Vof the transformer in series, but for some reason, the voltage drops down to 7V with the load. I tried a double voltage rectifcation from the 6.3V, but for some reason it stays at 9V only. (Incredible strange transformer, indeed!) After all, I used a series output of two 6.3V terminals for a double voltage rectification, and dropping the voltage with a 150ohm resistor to get DC12.6V with load.




In the left half of the chassis, wiring s for AF amplifire, product ddetection, ANL, IF amplifire after the filters, and BFO have been completed.
The BFO transformer seemed to be extremely old (perhaps 60years old), and I was worrying about it, but using frequency counter and a syncro scope, I confirmed its oscilation. (I am not sure if it produce a clean tone or not with such a shabby wave form...)




In the center area surrounded with shield plates, 100KHz marker, S-meter circuit, and VFO circuit(Except LC box) have been wired.
100KHz marker was also confirmed its funcion with a frequency counter and a syncro scope.
I adjusted S-meter circuit giving a signal of 455KHz from SG to the top of the IF amplifire, confirming its funciton. Cores of each IFT have been adjusted at the same time, as well.




I started to work inside of the LC box. I recognized an OSC coil of TRIO's S-series has a single coil with a tap, and is for Hartley OSC circuit.(It's too late!) There is no reason to use Hartley for the VFO which requires only 500KHz of spread. So I decided to make the original coil for Clapp OSC circuit. So the tube is also changed from a triode 6C4 to pentode 6BA6.
I made a coil using 0.5mm of enameled wire winding to the ceramic bobin that I had as many turn as possible. It does give oscilation, but its seems too much capacitive and with a 30PF VC, the frequency range stays around 150 to 300KHz only. I need to get 0.3mm enameled wire and rewind the coil again.




With a few times trial to wind 0.33mm enameled wire in different turns, I got 520KHz of range with a 30PF VC. For the time being, I think I go with this set up.

It is what I had been worrying about, but actually it is hard to fit 500KHz of spread acurately to the full scale. As I have designed, adjusting additional VC put in series and another VC put in parallel back and forth, must be the required method.
It is not impossible, but as the clear space in the LC box is extremely limited, I need to get small ceramic trimmers to achive that. Unfortunately, I do not have them now, and are difficult parts to find, it is a homework for the time being.

Moreover, I was planned to adjust the linearity of the VC by vending the wing of the VC, but if I do it now, the variable range will also varies, and it's not desirable now. With the situation nothing has been done, 0KHz and 300KHz on the scale are right on the real frequency, while 20KHz difference at 100KHz and 200KHz, 25KHz difference at 400KHz, and 50KHz difference at 500KHz on the scale.
I have 100KHz marker, so there will not be a much problem in real operation if I get used to it. In the worst case, I can redraw the division of the scale according to the actual frequency, so I leave it as it is, and I will think what to do later on.




All the wirings have been completed. (In the picture, wirings around the filters are removed for its adjustment.)
I recognized that the old TRIO's coil set which I obtained from YAHOO Auction was that for above 10MHz, and then I had to rewind the coil by my self.
I had to do a cut and try for a few times before I reached to get enough inductance for 3.5MHz, but I spent a lot of time for it.
I started adjustment, and I can demodulate the SSB very comfortably. I was surprised to find exellent stability of the VFO, which let me monitor the ragchews on 3.5MHz for more than 20 minutes without any toutch to the dial while I am doing the adjusting work, just same as commercial made SSB transceiver. I am sure that this receiver will work fine for practical use after the enough adjustment.




I found a big loss in the old mechanical filter, which I removed from the junk of YAESU all tube SSB transmitter FLDX400. It was what I had been worrying about. I then took apart it and ..... Oh! terrible! as what I have heard before, the sponge is died and got hard and fragil, and acutally it was locking the filter element's resonance movement. There is no way to get good resonance on 455KHz. I decide to clean it with a dehydrated ethanol later, and gave it a try to use another mechanical filter that I had. It was better, but it still has a big loss of the gain.
By the way, I was very much impressed with a SSB filter which absolutely get everything so quiet and let the receiver to be a high class. With only IFTs, the band is really noisy with a lot of signals, but with the SSB filter ON, it gets so quiet as if we got into the different band (maybe UHF) and very comfortable to monitor the band just as a normal SSB transceiver.
TRIO's IFT, T-11 has still lot of popularity to get the high price in the Auctions, but before a SSB filter, it has no meaning at all, hi hi.
Of course, this series of equipments are for AM, and the SSB filter would take away a good feature of AM, i.e., a nice audio. I have to think about the measure for it.




I cleaned the filter element with a dehydrated ethanol. The ethanol smelled so nice and I had to endure my eager to dirink it all, hi hi.
The sticked pices of sponge are removed, and all seems to be brand new. I put it back to the case wraping it with a piece of cotton.

I set it back to the circuit, and I was surprised to find the out put level was tremendously increased. VFO is very stable, the product detector is producing a clean audio, marker and ANL are properly working. Mechanical filter is great, and it's comfortable listening to a SSB QSO. However, it still need more S/N. It seems there is a problem somewhere between a mixer and RF amplifire. Somewhere there is a part that does not amplify the enough gain. It can't be so, as this is a receiver of RF1+IF"3"! It could be possible that a homebrew coil does not perform well.




After all, I rewinded the coil again. At the start, I modified the second coil of the TRIO's coil (at the lower left of the picture) to 21 turns with 0.5mm urethane wire. But as the sensitivity remained very low, rewinded it 40 turns with 0.33mm urethane wire.
Adjusting screw at above the chassis close to the front panel is for the antenna coil set below the chassis to avoid link with the RF coil, and this one is also rewinded in the same way.

The result is very succcesful, and a lot of strong signals of ragchew came in. Even for the receiving circuit, the resonance circuit is not enough with the simple resonance but does need a Q. This experience certainly taught me a lesson.
Putting a 0dB micro signal into the antenna terminal from the SG, and it was readable the modulated tone with enough S/N. I compared it with YAESU FT-817, but no special difference was observed. Now the sensitivity problem was completely resolved.




I recognized the linearity of S-meter was not good enough. I was checking the circuit and found the AGC voltage range is too wide and the input of the comparator of the S-meter was saturated. So, I inserted 1M ohm gain controller, and adjust all the level controllers so that S-meter, which is being set to indicate S-9 with the signal of SG, indicates compromised level with the signal 60dB stronger from the SG. After all, for only the S-meter, I needed three level controllers such as Zero adjust, Amplification gain, and Input level.

Now the filters. The SSB filter works great after the cleaning, and there is nothing to complain for the SSB reception, but this is the AM/CW receiver, so this can not be all right. On the other hand, 500Hz CW filter (Crystal filter) cuts the pass band very sharp but the output level is extremely low. After the examinations, the most important thing is to receive AM with a good audio quality, and for CW, a SSB filter is good enough as a receiver on 1960's.(I have never seen a CW filter in the homebrew receivers in that era.). So I decided to use an IFT for AM and the mechanical filter for CW.

The picture shows where I was trying to see the result switching a filter to an IFT temporarily.
Oh, fantastic! This is the AM sound! With the IFT, here come millions of signals and sounds! Fresh voice of announcer of Pyongyang Radio right int the middle of the amateur band! After switching it to the SSB filter, all disappeared in silence, except clear signal of the targeted SSB and CW. Oh, it's a nice combinations.




I found a brandnew MURATA's 6KHz ceramic filter that I bought before in Akihabara was kept in the parts box, and tried it. It is excellent!
It is not too narrow nor too wide. It demodulated AM beautifully. It's very different from crazy festival world with IFT. It costed me just JPY500(US$5.00) or so, very cheap. I was surprused with it high performance.
After all, I gave it my mind the IFT and use this ceramic filter for AM, and a mechanical filter for CW.

The stand-by circuit has been a pending, but now decided to give minus 100V to the AGC line to cut off the tubes. So I mounted a small junk transformer to get AC 100V. (The tranformer was used in the junk measuring instrument which I used to build 3.5/7MHz portable transceiver. What a great junk!)




I found it the vernier dial is good enough to tune the frequency even for the SSB, so I threw away the risk to extend the wire drilling a hole to the LC box to extend the wire to the spread VC on the front panel.
Instead, I added IF GAIN controller in addition to the RF GAIN controller. It is useful to adjust a large difference of gain between two filters.
MURATA's ceramic filter is such a small as in a picture, but works just great.




As a matter of fact, the calibration and redrawing of the dial scale have not been done yet, but using the receiver actually, I was not much frustrated. (The acutual frequency is understandable roughly, and the marker tells exact point every 100KHz.) So, I will work with this later when I feel I need it.

If I maybe allowed to wish so much, adjustment never be finished, so I decided to declare the completion here.




Only except the crystal convertor, Transmitter + VFO + Receiver are completed. For 3.5MHz, the crystal convertor is not necesarry, so now I can work 3.5MHz without any problem.
Now, the debut time has come!

I called CQ on 3.757MHz for seven times, but nobody called me back, hi hi. But you see how much I enjoyed that? My heart beats fast and excited so much!
For the microphone, I use my treasure, a velocity microphone (ribbon microphone). It is very heavy. I use a tube head amplifire using one 12AX7A before I connect it to the transmitter. Monitoring audio with other receiver, the audio is super!

I will find a couter party of QSO taking a time, and I asked one of my local station JL1IHE to get on 3.765MHz to make the 1st QSO. Of course the signal report was 59++ with his ICOM IC-756PRO, and he gave me a report that the audio quality is just wonderful. The VFO seems to be very stable, and while I was trying to repeat stand-by switching fast manytimes, what he has received was a tone of a clean carrier being intermittent just same as CW signal.

Well, a crystal convertor is only left. All the series will be completed, soon.


This is the final circuit diagram. If you need detailed image, please download a PDF file






In case you hit this page directly from a Search Engine, you can reach to all of the pages of MAFNET from Top Page