Building the Simple 80 Metre DSB Minitransceiver by VK3AJG?

Before even thinking about building that, do you understand it is illegal to operate it in your country without a license?

To answer your questions:
1. They are the 10mm square metal cans you find in older AM/FM domestic radios, usually with a hole in the top and a screwdriver slot inside it. The frequency doesn't matter because you will be discarding any capacitors inside it (usually wired between the pins underneath the can) and the coil itself then winding your own coil on it. Basically all you are doing is salvaging the mechanical parts. Warning: NEVER use a metal screwdriver on the gray core, they are very brittle and will snap very easily!

2. It only tunes part of the 80m band and it transmits a DSB signal. There is nothing wrong with that but almost every signal on the 80m band is SSB not DSB which is more complicated to generate. The transmitted power is enough to cover tens of Km but the exact range depends on the antennas at both ends, the sensitivity of the receiver and whether there are any obstructions to the signal.

3. It can be used 'walkie talkie' mode but because of the limited antenna length if you are carrying it, the range will be quite short. When 80m is used in mobile communications, a loading coil is used to make the antenna appear longer than it really is, it makes the transmitter more efficient so it increases the range a little but the coil itself would be big enough to make it awkward to carry by hand.

4. The alternative would be a quartz crystal. As the tuning is achieved by 'pulling' the resonant frequency, quartz would limit the tuning range to maybe less than 1KHz. Ceramic is less stable but can be pulled more than quartz.

5. These are best made from small rings of ferrite also known as 'toroids' (maybe 10mm diameter with a 6mm hole). It might be possible to make them from normal interference limiting beads but the hole is probably too small to fit all the wires through.

6. It isn't critical as long as the wires are insulated so the turns cant short to each other. It is important that the windings are identical so the trick to winding them is to tie lengths of wire together and wind them simultaneously rather than winding one coil after another. It makes sure the magnetic coupling as a near identical from one winding to the others.

7. If it is a double sided board (which would work best) just solder the ground wires of components to the top and bottom sides.

8. 8 Ohms will work but a higher impedance speaker would be better. It will only produce low volume levels anyway but higher impedance speakers or headphones would be a bit louder. The circuit expects an 'electret' microphone with built in pre-amplifier, the kind that looks like a small metal can with two pins that you find in old cassette recorders and some webcams.

9. The tuning RANGE is only 80-90KHz (it depends on the ceramic resonator) so for example it may be from 3.500 to 3.590 KHz. The center frequency depends on the resonant frequency of the ceramic resonator. So basically, it only covers a small portion of the band.

Brian.

Thanks for clarification!

I am well aware of that, I respect the laws and I am not going to make any pirate transmissions, I just want to build it for educational purposes, and in future I will surely get license!
I even consider building only the "receiving" parts of this circuit.

Anyway, now almost everything is clear, there are only few more things which are bugging me:
- what are the values of the diodes near the speaker? I don't even know what is their role, but it looks like amplifier, maybe I could replace it with LM386.
- where should I measure the frequency with my oscilloscope when testing/tuning it to 3600kHz or so?
- is the any preferred type of capacitors used in this circuit? I've read that "monolytic" are better than ceramic

You can replace the whole top right part of the schematic with an LM386, it is just the audio amplifier to drive the loudspeaker.
If you do build it as shown, the diodes should be small signal silicon types such as the 1N914 or 1N4148. Don't use Schottky or any special diodes as they are there to ensure the two output transistor are biased properly and they should drop the same voltage as the Base-Emitter voltages of the two transistors. In fact for best results the diodes should be thermally bonded to the transistors so they are at the same temperature.

Unfortunately, unlike the previous design you looked at, the frequency in this receiver is fixed by the ceramic resonator. There is no 'pre-set' adjustment to center it on the frequency you want. To select other frequencies you have to change the resonator but you will find 'uncommon' frequencies very hard to obtain. The 3.58MHz one is a very common frequency because it is mass produced for NTSC television sets and also frequently used in telephone tone dialers.

Don't worry about the capacitors, each type has it's good and bad points but only ceramic is made is such a wide range of values and with wire ends. From a functional point of view you wouldn't see any difference at all from one type to another.

Brian.

A 80m Transceiver
http://www.ozqrp.com/docs/MST2_80M_manual_V1.pdf

It seems a bit too complicated as for the start, also I wouldn't be able to manufacture such a quality twosided PCB at home...



Are you talkingn about that circuit on the schematic:
Attachment 140662
It applies only to "transmitter" or could it improve my receiver reception?


So it makes no sense to use quartz? I've considered using one of the TV ones:

3.575611MHz			PAL		PAL M color subcarrier
3.579545MHz			NTSC		NTSC M color subcarrier; see colorburst. More specifically, 315/88 = 3.57954 MHz. Because these are very common and inexpensive they are used in many other applications, for example DTMF generators
3.582056MHz				PAL		PAL-N color subcarrier
3.595295MHz				NTSC		NTSC M color subcarrier, plus horizontal scan rate (15,750). Used for a rainbow color test, produces color through the entire 360 degrees of phase shift. Unusual.[2]

Similar but not quite the same. Note the length 'l' is still 41m so for a walkie-talkie you would need a helium balloon on the top end or it would be difficult to hold upright
For a transmitter to work most efficiently, its output impedance should be the same as the antenna impedance. As it would be virtually impossible to carry around a 41m vertical antenna, the trick used is to wind it into a coil. It keeps the transmitter happy so it can produce most power but only the remaining length radiates significant signal, the coiled part still radiates a little but much of it is lost. It's a trade-off between efficiency and portability.

Using Quartz instead of a ceramic resonator will restrict the tuning range to only about 2KHz, too narrow to be of much use.
As long as you use the same in the transmitter and receiver it should be OK but such precise frequencies usually mean they are made of quartz so they will suffer the same limited 'pull' range. A single 3.579MHz ceramic will probably cover all those frequencies with the existing tuning circuit.

Brian.

Okay, I will stick with ceramic ones for now.
I have collected the following parts so far:

1. Toroids:
At first I wanted to salvage them but they were too big and almost every one was different:

1. The big toroids are for EMC filtering, you fit them over power cables or wind just a few turns of thick wire around them to stop interference from switch mode power supplies going down the power lines where it can reach other equipment. They are designed for low frequencies and normally to stop signals passing through, the opposite of what you want!

The smaller T32 toroids are ideal, the idea is they trap the magnetic field inside the core and therefore make it more efficient. A rod core radiates a field outside the coil so it loses some of the signal. The coil winding is OK, I probably wouldn't have twisted the wires so tightly but it wouldn't make much difference in your application. The idea is that the coils are as identical as possible. The important factor is the magnetic coupling between them should induce the same voltage across the coil ends and for that the number of turns must be the same, how far apart the turns are spaced from each other and the core, and the capacitance between the wires should be the same. If you over twist them the capacitance between the wires increases which may not be helpful but the idea of twisting them at all is that they follow as near identical route through the core. Be very careful to wire the start and ends of each coil correctly, if you reverse one of the windings (swap wire ends) it will cancel the signal instead of reinforcing it!

2. Those are the right parts. There is some natural variance in their resonant frequency, ceramic is not as accurate as quartz so you might find one that gives you slightly better frequency coverage.

3. The wire thickness isn't very important, if you can find thinner wire it will make it easier to fit all the turns but the whole idea of the movable core is you can adjust it's value so it may be OK as it is.

4. The earphone may be OK but a small loudspeaker will be better. The microphone probably isn't suitable but it might work. The design expects an 'Electret' microphone which looks like a small (~10mm) cylinder with two pins at one end and a small hole at the other. One of the pins will be visibly joined to the cylinder, that is the ground side. Electret microphones use an electrostatic diaphragm and have an internal FET pre-amplifier so they have a relatively high output voltage, if you use the type in the intercom it is probably a dynamic type and it may not produce enough volume.

5. I wouldn't do that, you will find it difficult to fabricate and connect to the internal ground layer. I can't see any reason why it shouldn't work on single sided board at 80m. On a double sided board you get the benefit of a continuous ground but you have to solder the ground connections on both sides and also keep clearance around all the other component legs so they don't short to ground. You can see that a 'sandwich' of two boards makes that difficult to do.

6. It might work but the problem it would introduce is the volume control. As you adjust it, the load on Q1/Q2 would change and you might get strange results. The idea behind this circuit is that Q1 and Q2 are exactly balanced but carrying signals out of phase (one upside down with respect to the other) and the next stage is there to maintain a constant load on it. I would keep the following stage and connect your LM386 to the existing volume control connection.

Brian.

They supply it being as kit.
http://www.ozqrp.com/shopDX.html

I have etched the two-sided board and I am assembling this circuit right now.

The audio amplifier stage is already working - I can hear the static noise and it reacts to touching the solder joints.

Now I wonder, the IF Can on the schematic (L7) is connected to ground by JP5 in one case. Does it mean that I don't need to solder L7 (IF can) when I only want to receive?

Also, should VFO (3.58 oscillator) work without L6 soldered?

JP5 couples the antenna signal to the mixer (through C25 / J2) so the top two pins as shown on the schematic must be linked for it to work.

It should do, but obviously you wont receive anything.

Brian.

I have added two jumpers, the speaker is working, but I can't measure oscillator signal (I tried in multiple places).

The voltage at the Q3 (BC549C, VFO) base seems to be 1.3V.

I still haven't soldered mic, and the transmitting parts...

Any suggestions what might be wrong or what else can I measure?



I have added two jumpers, the speaker is working, but I can't measure oscillator signal (I tried in multiple places).

The voltage at the Q3 (BC549C, VFO) base seems to be 1.3V.

I still haven't soldered mic, and the transmitting parts...

Any suggestions what might be wrong or what else can I measure?

The only place you can measure the VFO output is across R2 using a x10 probe.

If you have 1.3V on the base, which seems reasonable, you should have around 0.6V on the emitter and something a bit higher, maybe 4V on it's collector.
If the collector voltage is lower than about 3V, check the voltages on Q1 and Q2 and let me know what they are.

Brian.

I have 1.3V on the base, and 0.6V on emitter and collector.

Q1 and Q2 common emitter has voltage 0.6V.
Q1 base has voltage 1.3V.
Q1 collector has voltage 0.6V.
Q2 base has voltage 1.3V.
Q2 collector has also voltage 0.6V (well, Q1 and Q2 collectors are connected by the inductor, so it's not suprising...)

Looks like the wiring around L1 is wrong then.

Q1 and Q2 collector should have approximately the same DC voltage on them but they are not linked through the inductor. In fact in normal operation the signal on them should be out of phase, as Q1 voltage rises, Q2 falls and vice versa.

The collector voltage on VFO transistor comes from the current being passed by Q1 and Q2 and clearly there isn't any. The current path is through potentiometer R11, then the two windings of L1 then Q1 and Q2. The collector voltage on Q1 and Q2 should be quite high compared to their emitters. Check the value of R11 is correct and that it has 8.2V on it's slider and then the path through L1, it has to be broken somewhere in that area.

Brian.

On the R11 middle pin there is 8.4V (because of the Zener diode).
R11 is a trimpot market as 202, which means 20 * 10^2 = 20*100 = 2000Ohm.
The remaining R11 pins also have 8.4V.
And then... I don't know where to look.
Is my L1 correctly connected?
I have marked the windings connections on the schematic:

The 8.2v on the other pins of the potentiometer means no current is flowing through it - makes sense in view of the other voltages you reported.
The problem is the wiring of L1, the top and bottom red lines in the layout are indicating the coil connections but they should be vertical, not horizontal.
The middle line is OK, that's the output coupling coil.

I think if you can rewire L1 it will fix all the problems so far.

Brian.

Well, I didn't see your post but I changed two coil wires on one side so it's connected like that:

I'm glad it was so simple to fix - and my diagnosis was correct!

For receiving, leave the two controls in mid position for now. Their purpose is to exactly balance the signal levels in Q1 and Q2. It's called a "balanced mixer" circuit, Q1 and Q2 work like a "see saw", as the current in one increases, it drops in the other. The idea is that if they are exactly balanced and the coils in L1 are matched, the average signal in L1 core is zero because the magnetic fields are equal but opposite and they cancel out. The third winding (the middle one in the schematic) therefore picks up no signal and no RF is amplified or transmitted. The microphone signal only connects to one side of the balanced mixer, to Q1 via L6 so the audio unbalances them and they no longer cancel. It's the unbalanced signal that forms the DSB transmission. So when it is all working, you make sure no audio enters the mic input then adjust the two potentiometers for minimum (ideally zero) output.

When you hear a station near 3.58MHz, adjust the 'can' transformer for maximum signal. It is the equivalent of the bandpass filter on your Lidia receiver but because the tuning range is so narrow on this one, a peaking filter is more appropriate to use.

The oscillator should run no matter where the control is set, if it is stopping it is probably at the extreme of minimum value. You could try adding a small capacitor across it to prevent the value reaching so low (~5pF) or you could try changing the transistor for one with higher gain. Some experimentation may be required but don't forget the resonators are not intended to be 'pulled' off frequency anyway so the circuit takes advantage of their weaknesses and performance may vary from one resonator to another.

Brian.