The 7MHz SSB / CW radio receiver has a Fine Tune by hs8jyx

[Garyl]

Hello again!
I am still studying various receivers projects and now I've stumbled upon the hs8jyx creation described here:
https://hs8jyx.blogspot.com/2013/08/ssbcw-7-mhz-fine-tune.html
From what I see it's a simple receiver for 7MHz (40m) band, which should be able to receive something at night.

This is the schematic:


The PCB and connections diagram:


PCB PDF:
View attachment 7mhz-pcb.pdf

Description (translated by Google Translate):

The 7MHz SSB / CW radio receiver has a Fine Tune
First of all, thanks to your fellow members for your interest in this project. This project is the same circuit as the 14 MHz version (https://hs8jyx.blogspot.com/2013/06/14-mhz-vfo.html), but we made a new circuit board to simplify the sealing box. For a frequency of 7 MHz the frequency is quite constant. There is no frequency slider so much that I can not accept it.

The 7 MHz frequency will be able to communicate well in the night. Members who have not played this frequency yet. Do not be surprised if the day can not get anyone. Or get a strange noise.
Download the PCB at https://drive.google.com/file/d/0BxgJ9QX1BWUyWnBVbE1BSlJfR1E/edit?usp=sharing.


1.5 uH coil using 19 copper coil on T37-2 (red) coil (round counter of spindle https://15meter-radio.blogspot.com/2013/02/toroid.html )


The L1 coil can be wrapped around 21-22 on T50-7 or 22-23 on T50-6 (this is because the coil can be squeezed or loosened. Number of rounds can change a little)

(Example of reception I do not have a 7 MHz antenna. Use a long cable instead of an antenna.) Figure down the device and the connection.



Wide Bandwidth - Narrow What determines

From the figure is our VFO circuit. Adjusting the frequency of the VFO circuit can be done in two ways: adjust the L (coil) and adjust the C (capacitor). Here we use the C modulation using the capacity. At 1N4001 diodes, the higher the voltage range, the wider the bandwidth. (Ie 0-8 volts is wider than 0-5 volts, for example).

C2 and C3 also determine the bandwidth. up to a certain point, it does not matter L1 C1 will affect Bandwidth much. That if I Bandwidth widened to the L1 high value C1 less (frequency remains the same, but Bandwidth increases ) that this is coming from the C in the diode will affect the frequency, if connected to a C. a few examples receiver 7.1 MHz (to practice it), we assume a diode 1N4001 in tune with the 2-25 pF (by tuning voltage maximum and minimum).


Frequency Resonance Formula


Circuit 1 uses a low coil (using 1uH). We need to use a C value of 502 pF (500pF + 2 pF of the smallest diodes) to obtain a 7.1 MHz frequency.
At the lowest voltage, we get a frequency of 7.100 MHz ( 1uH // 500 + 2pF)
when the maximum pressure is not frequency MHz is 6.950 (500 +, 1uH // 25pF)

a bandwidth = 150 KHz,

the circuit uses two coils over (use 2 uH)
we need to use the C 502 pF (. 249pF + 2 pF diode, as a minimum) to 7.1 MHz frequency

when the minimum voltage to frequency 7.100 MHz (1uH // 249 + 2pF
p. When the maximum voltage, we have a frequency of 6.800 MHz (1uH // 249 + 25pF) Bandwidth = 300 KHz

Photos of original circuit made by author:


And now, I want to replicate this circuit but I have several questions:


1. Is this diode correctly connected? It looks like a direct connection to ground, but well, it's at the gate of the MOSFET...


2. The tuning circuit - are they using 1N4001 diode as varicap?

3. The tuning circuit - from what I see, I can start that VFO without NE602 for testing on the separate board? I will measure frequency with oscilloscope...

4. The tuning circuit - is the 2N3819 JFET crucial here or can I use other MOSFET or even BJT?

5. How crucial are those toroids, listed as "T37-2", "T50-7", "T50-6"? Can I get it working with the "no name" toroids from the electronic waste?



6. Why can't I just use a normal 1.5uH axial trough-hole inductor 1.5uH in the double pass filter? Is this related to current issues or what? Does it really have to be 19 turns T37-2 (red core)?

7. There is LM386 connected to the TDA2030 by potentiometer. They are both audio amplifiers, then, why are there two amplifier stages used? Is this justified in this circuit, or is this just for... louder sound?

8. What kind of antenna do I need for that receiver? Does it have a good sensitivity?

 

[betwixt]

1. It is correct and it should be a small signal diode. That stage is an oscillator (VFO) and it's purpose is to stabilize the level of the oscillation.

2. Nothing wrong with that at all. The way a variable capacitance diode works is it uses the width of the depletion region of the PN junction much in the same way as the distance between plates of a capacitor determine its value. 1N4001 is useless as high frequency rectifier because it can't start and stop conduction quickly enough but in this application it is used only in non-conducting mode (reversed voltage) and in some respects the larger junction in a power rated diode is beneficial because it has a higher surface area and therefore higher capacitance.

3. Yes, unlike the MC3361 where the oscillator is inside the IC, this design uses an external oscillator. You can measure it's frequency where marked by the red circle.

4. It has to be a JFET I'm afraid. It should be possible to design an equivalent circuit using other types of transistor but it would be quite different. The 2N3819 has been obsolete for many years but you can use almost any small N-channel JFET as an equivalent.

5. Not critical at all, they are only used because they allow lots of inductance in a tiny area. The only problem is the 'Al' property of different toroid size and material means you get a different inductance per turn. You might have to experiment with the number of turns to get the frequency right.

6. The 'Q' factor of axial inductors tend to be less which means a small loss of signal but for this application I doubt it would be big enough to notice. Current rating is completely irrelevant as it will never be more than a few nano-Amps.

7. The design is strange in that respect. The LM386 makes an extremely poor pre-amplifier so why the designer used it is a mystery. It probably does need an amplifier stage but a conventional low-voltage op-amp would work far better. I would replace it with a TL071 or similar and add the extra few components to bias it properly.

8. Sensitivity should be about the same as the MC3361, good enough to pick up local stations during the day and maybe other continents at night, even with a simple wire antenna.

Brian.

 

[vfone]

The differential connection (and all the benefits of it) between NE602 and LM386 is widely used in schematics like this.
Now, to put a 10% distortions amplifier (at 0.5W) in front of a 0.1% distortions amplifier (at 8W), doesn't make sense.

 

[Garyl]

7. The design is strange in that respect. The LM386 makes an extremely poor pre-amplifier so why the designer used it is a mystery. It probably does need an amplifier stage but a conventional low-voltage op-amp would work far better. I would replace it with a TL071 or similar and add the extra few components to bias it properly.

Would an op amp like MCP601, MCP606 or maybe MCP618 work?

Anyway, thanks for the input! I have collected most of the components and I will etch the PCB soon!
My NE602 is in SMD case but I also have a SMD2DIP board. The A2030H (TDA2030) is taken from my old parts bin, I guess I desoldered it from somewhere when I was a kid, but I hope it's still working.

Unfortunatelly I don't have 1N4001 yet. Can I use 1N4148?
And are those variable capacitors at antenna input crucial? How do I tune them?

I've been also wondering if it would be possible to create the other band (for example 80m or 20m) version of this circuit. I have looked into NE602 datasheet:

it seems that this chip can support higher frequencies.
Would it be reasonable to receive 20m/80m with circuit with changed VFO capacitor value and the input filter?



I also need some way to test the receiver after soldering it, especially because I haven't any good antenna. Would a crystal oscillator work here as a test signal source?
The nearest crystal value to 7MHz I have managed to find so far is:

 

[E-design]

In an article by **broken link removed** he claimed that the gate clamping diode degraded the phase-noise of the oscillator.

 

[betwixt]

Would an op amp like MCP601, MCP606 or maybe MCP618 work?

Unfortunately, no for two reasons, first is they can't survive supply voltages as high as 8V, second is they have no internal bias circuits. The LM386 is unusual in that it's inputs can be taken down to ground level but it suffers very high noise levels and quite severe distortion as a consequence. You could use one of those if you connected its supply to the 5V regulator output and added some bias resistors. For example add a 100K resistor from the + input to ground, a 100K resistor from the + input to 5V and a 47K resistor from the output pin to the - input. You might have to play with the feedback resistor (47K one) to get the gain as you need it. If you research differential amplifiers it will explain how they work.

Unfortunatelly I don't have 1N4001 yet. Can I use 1N4148?

Again no, it might work but you would be exceeding their current rating. You could use any of the 1N400x series of diodes though.

And are those variable capacitors at antenna input crucial? How do I tune them?

The types and values are not critical but don't stray too far off the value. In engineering talk "tune for maximum smoke", in plain language, find a weak signal somewhere near the middle of the band and adjust them for best strength.

Would it be reasonable to receive 20m/80m with circuit with changed VFO capacitor value and the input filter?

Yes, even 10m but I would expect some degrading of sensitivity as the frequency increases. You would almost certainly have to change the tuning inductor and reduce the VFO feedback capacitors to tune higher than 40m.

6.9375 is only just below the 40m band so yes you could use it as a test signal.

E-design's comment about the clamping diode is correct but I would still leave it in place. It degrades the phase noise slightly but probably not enough to notice in such a simple receiver. It does however, stabilize the VFO output voltage which is more important in that design.

Brian.

 

[Garyl]

Okay, so I have a progress.

I am using 1N4002 instead of 1N4001.

1. I have etched the PCB.
2. I have soldered the audio stage with 2030 - the speaker reacted to touching the input.
3. I have soldered the VFO - I didn't check it very carefully, but it gave me no output.
4. I have soldered the Mixer chip and the LM386.
NOTE: the input antenna filter is still not soldered! Other components are in place.

Now:
- On the scope I can see a jerky 8MHz waveform signal on pin 6 of mixer
- Changing pots isn't changing anything in the waveform (Or maybe the change is so small that I can't see it!)
- Touching LM386 pins makes speaker react
- Touching pin 6 of mixer makes speaker react
- LM386 is getting warm fast.

I have used the exact components values from the PCB image (the blue one).
I only put 2pF capacitor instead of 1.5pF

What is wrong? That LM386 is heating too much, but I can't figure out why... the mixer seems to be working.
I can show HD photos of circuit or scope signal if it's needed.

Also, do I need to change anything here more than crystal value to generate 7MHz test signal:

?

 

[betwixt]

The waveform on mixer pin 6 should be stable but you must use a low capacitance scope probe (x10 probe) and ground it close to the VFO to see the waveform properly. The frequency isn't too important at the moment, you can adjust it later but the VFO is obviously working which is a good sign. The potentiometers should have little or no effect on the waveform.

I'm not sure why the LM386 is getting hot but as stated earlier, it isn't a good choice of pre-amplifier. I would guess it is unstable, probably because it is driving a capacitive load, adding a resistor of about 47 Ohms between pin 5 and the 10uF output capacitor should fix that. Good design practice is to add a 100nF ceramic capacitor across the LM386 supply pins (across the 100uF capacitor beside it).

The crystal oscillator should work fine over a reasonable frequency range by just swapping crystals.

Brian.

 

[Garyl]

The waveform on mixer pin 6 should be stable but you must use a low capacitance scope probe (x10 probe) and ground it close to the VFO to see the waveform properly.

I am using the x10 from the start, but my ground connection wasn't indeed very close. I will try again soon.

I'm not sure why the LM386 is getting hot but as stated earlier, it isn't a good choice of pre-amplifier. I would guess it is unstable, probably because it is driving a capacitive load, adding a resistor of about 47 Ohms between pin 5 and the 10uF output capacitor should fix that.

Good design practice is to add a 100nF ceramic capacitor across the LM386 supply pins (across the 100uF capacitor beside it).

I have added 100nF ceramic capacitors at the copper side of the board, just at the LM386 supply pin and at the TDA2030 supply pin.
It was still warming, so first I disconnected LM386 pin 5 and 10uF cap track.
Now the LM386 was cold and receiver only gave sound when I connected audio pot or the TDA2030 tracks.
Then I added the resistor (47 as you said), and now the LM386 is not heating, and TDA2030 is heating only a little.
But I feel that the sound static changed slightly....


Also, now I have the 50pF input antenna trimmers left to solder, but I realized that my trimmers from drawer have no value description, they are just in different colours. One is yellow and one is brown.... one of them is 40pF and second is 60pF... is there some kind of single-colour code for them? Or maybe I can use just 27pF caps instead of them...

 

[betwixt]

Then I added the resistor (47 as you said), and now the LM386 is not heating, and TDA2030 is heating only a little.
But I feel that the sound static changed slightly....

A small amount of heating, especially in the TDA2030 is normal. I would guess the change in 'static' sound is good, if the LM386 was oscillating it would produce some strange sounds of it's own. You are probably hearing normal background noise now and it was wrong before. The 47 Ohm resistor will make almost no change in the volume or tone quality of the sound, it just adds some isolation between the output of the LM386 and the capacitor across the volume control because like most amplifiers they become unstable when driving a capacitive load.

lso, now I have the 50pF input antenna trimmers left to solder, but I realized that my trimmers from drawer have no value description, they are just in different colours. One is yellow and one is brown.... one of them is 40pF and second is 60pF... is there some kind of single-colour code for them? Or maybe I can use just 27pF caps instead of them...

Years ago there was a color coding system but when Asian production ramped up, it was ignored and these days the color means nothing. However, bear in mind the value of the trimmer capacitor depends on the number of 'vanes' on the shaft and the area interleaved between them. They have a maximum value when all the vanes are intermeshed and a minimum value which is from proximity when they are widest apart. That means for the most part that you can use a higher value than specified if you adjust it to a lower value. Equally, if the value isn't going high enough, you can add a fixed capacitor across it. For example, if you needed 50pF and you had a 20pF trimmer you could add 40pF in parallel with it so the range covered say 42pF to 60pF. I assumed 2pF as the minimum trimmer value.

You can use fixed capacitors instead of the trimmers but the reason they are used is to allow adjustment for the tolerance of the inductors and 220pF/270pF capacitors so you may not get optimal results.

Brian.

 

[Garyl]

I will most likely use them with 10pF capacitors and just try to tune.

Right now I have tested my old 3.6864MHz tested with the scope:


and I have also prepared a new one - 6.9375MHz:

I don't know why the signals have different amplitude, I have used the very same power supply for both (stabilized 12V).


The generated frequencies are OK so I started testing the receiver.
Unfortunatelly, I haven't managed to pick up the test signal.
Also I'm not sure about the VFO state, because today I haven't managed to scope it successfully - just no signal at all shows up but TRIGG'D, but maybe my old scope can't catch it somehow or I am using it wrong. But hey, I hadn't any trouble with scoping the previous project VFO signal, nor with scoping the test signals....


PS: The interesting thing is that my receiver is somehow picking up the interference from mobile phones. I mean that commonly known "speaker interference" sound. That sound: https://www.youtube.com/watch?v=R0O1_FYzbjw

What do you suggest to do next? I've been thinking about adding a variable capacitor to VFO. Maybe replace the "Main Tune capacitor" with variable one...

 

[betwixt]

The different amplitude may be nothing more than the feedback capacitors being too high in value for the new crystal frequency but it makes little difference as the amount you need to produce the test signal is many times smaller than even the smaller waveform.

There is one thing worth checking though: You are using a x10 probe which is good, but have you adjusted it's compensation control? Your oscilloscope almost certainly has a calibrator test point which carries a square wave signal, connect thge ground and the x10 probe to the test point and adjust so you see the waveform. It should have a flat top and flat bottom, if you see a slope or curve it is under-compensated, if you see sharp points on the edges (overshoot) it is over-compensated. On the probe there will be a small adjustment hole, be very gentle as the control is tiny and fragile, if you break it the probe is useless. Turn the control, preferably using a non-conductive screwdriver (a plastic one or a trimming tool) until the waveform is as near a perfect square wave as you can manage. If the compensation is wrong, the amplitude of the waveform will change according to frequency so you get an inaccurate reading. Note that if you have a switched x1/x10 probe, the compensation only works in the x10 position so don't try adjusting in x1 position.

Mobile phone pick up is the AM envelope of the phone transmitter pulsing on and off, you hear the bursts of data but not because it is receiving them in the conventional radio sense, it's because the wiring and tracks on the PCB are acting as antennas and the tiny voltage they pick up are upsetting the amplifier bias. It's one of the reasons why shielding is sometimes needed but then not many people use their phones while listening on 40m!

What do you suggest to do next? I've been thinking about adding a variable capacitor to VFO. Maybe replace the "Main Tune capacitor" with variable one...

The only reason for using the potentiometers is they allow wires to run to the controls if they are mounted off the board. A variable capacitor will certainly work and may be a solution to the tuning problem but remember that all wires and PCB traces have inductance so if you connect the the variable capacitor 'off PCB' the wire lengths will have an effect on the frequency. Also, if you do use a variable capacitor, make sure the 'body' and the shaft are connected to the circuit ground or it will seriously go off frequency if you touch it. If it isn't obvious which side is ground, do a continuity test between the shaft and connecting pins, only the ground side should show conductivity.

Brian.

 

[Garyl]

I have written a long post, but the forum has broken it again (it frozen after I hit "upload image" and the page became grey and I was unable to copy it out...).
So I will keep short this time.
.. but that's not the first time, it's annoying, also, why do I keep getting "reply notifications" when I reply in my own topics?

So, the oscilloscope is calibrated well, but indeed I have never done that before and you've gave a good hint with it.

The VFO might have infact never worked.

I have meaesured the voltage at the sources of JFETs at it was around 0V.

After a little consulting I have decided to replace 2N3819 with BC547. The 5V-powered BC547 got extra 100k restistor between base and collector, and the second one got 150k one. I also removed 1N4148.

The signal from VFO is now very clear, take a look:

The main tune pot is changing it clearly, but it's also changing this signal amplitude...

What do you think about this change? Tomorrow I will perform tests with test signal generator and finally solder remaining trimpots to antenna filter.

 

[betwixt]

The change should be OK although the FET version should work equally well.

The tuning pot should alter the frequency but don't worry if the amplitude changes a little as well, it is normal. The amplitude depends on lots of things, in particular the feedback capacitors and the coupling capacitors. There is an optimum value for each frequency but obviously you can't change them as you tune the VFO so expect some variation in level.

Brian.

 

[Garyl]

I still haven't soldered the 50pF trimmers to antenna filter, but I have tested the circuit with 6.9375MHz frequency generator.

With the frequency generator connected directly to antenna input, I can clearly hear the signal, but I can hear it in more than one frequency potentiometer position...

With just the very long antenna connection, I can also receive the test signal, but only at one knob position, but the noise is way bigger.

Well, it seems it's time to try receiving some stations at night, right? At the moment I have only old repurposed cable hanging from the roof which I use as antenna, I hope it will be enough...

 

[Garyl]

It seems that receiver is working. I have received some strange signals and one radio station, which seems very strange and I'm unsure if it's a commercial one or what. Also I can't recognize the language.

Can anyone hear the recordings and tell me what is that, or at least recognize the language?


2017_09_06_17_52_08_40m_radio_startsWithMusic.mp3 - strange music and then radio... sounds like some commercial radio station?
View attachment 2017_09_06_17_52_08_40m_radio_startsWithMusic.zip

2017_09_06_22_05_04_40m_unknownsig.mp3 - what is this signal? Sounds too fast for the CW?
View attachment 2017_09_06_22_05_04_40m_unknownsig.zip

2017_09_07_13_12_34long40mradio.mp3 - 15 minutes recording, there is a strange tone at 00:24 to 00:55. Another tune is at 6:10-6:20. Another 13:44 to 13:58
View attachment 2017_09_07_13_12_34long40mradio.zip

Is this possible that I am receiving a non-40m radio with my receiver?

PS: The trimmers at antenna filter are still not soldered...

 

[betwixt]

There are broadcast stations in and around the 40m band so it isn't unusual to hear broadcast signals mixed with amateur ones. The band edges are also different in some regions of the World.

The first recording is a classic demonstration of why direct conversion receivers are not used commercially for AM signals. The background tone is the carrier being converted down to audio frequencies and the garbled music/voice is where you are 'almost' tuned to the carrier frequency but the pitch isn't quite right. It is theoretically possible to recover AM accurately but keeping the frequency exactly matched is difficult and you will never get the phase to match (same frequency but the waveform shifted in time slightly) so you get a warble on the notes.

The second recording is a radio teleprinter (RTTY), it sends binary numbers as two tones, one for a '1' and another for a '0'. The data is usually encoded though so except for amateur transmissions where plain text is used, you can only recover what looks like random characters.

The third recording would take me too long to download at the moment but I'll try again later when the school kids have gone to bed tonight and the internet speeds up again.

Brian.

 

[Garyl]

There is still one thing I don't understand about this schematic...
I have read about NE602/SA602 a bit more and it seems that has built-in oscillator which can work with few passive extenral components.
Why wasn't it used here?
Why it was substituted with external VFO circuit with two JFETs and many other components?
Or maybe I am understanding something worng..


PS: @betwixt, did you manage to download third sample? Or at least recognize the language from the recordings?

 

[betwixt]

I'm not really sure why the designer followed that thinking. If I had to guess, it would be that the NE602 is optimized for higher frequencies, typically around 45MHz where it would work as the oscillator and mixer to produce a fixed IF output. That doesn't stop it being used at lower frequencies but the noise and intermodulation (= overload) characteristics might not be so good. I would think the 'front end' of the NE602 and MC3361 are virtually identical but the 3361 has the added discriminator and audio amplifier.

I managed to download the file, I don't recognize the language but it sounds 'Eastern European', maybe Hungarian or Romanian. I used to work in Romania but I didn't pick up enough language skills to be able to understand more than a few words. I suspect your tuning is a little off, the majority of AM stations are around 7.2 to 7.4MHZ so you might be a little higher in frequency than you thought.

Brian.

 

[Garyl]

I have tested this receiver several days already and I have following conclusions:
- it seems I've been able to receive only the single particular station so far. There are NO other stations.
- tuning with this receiver is relatively easy, much easier than my 80m MC3361 one
- voice quality is also way better (on 80m MC3361 I struggle to receive readable voice, even while there are much more stations hearable there, but all are too disorted and hard to tune)

Now I also know what kind of station I've been receiving.
On 2017 09 30 I have heard an English transmission. It's the Voice of Turkey. Their reception is easy with the receiver from this topic.
Here are the recordings:

2017_09_30_14_52_07_40m_letterbox.mp3
View attachment 2017_09_30_14_52_07_40m_letterbox.zip
This is a longer recording of their "letterbox". They are reading emails and talking about QSL cards. They also mention some called "Question of the month".
The interesting moments are:
0:55 They are reading reception report on 9700kHz
7:40-8:10 (or so) - what happens there? Clearly a second voice is heard. And I am pretty sure that was from my receiver, there was no TV nor other radio running near my receiver when I recorded it.
9:40-10:00 They are saying their contact address and email address.

2017_09_30_15_02_41_40m_questionOfTheMonth.mp3
View attachment 2017_09_30_15_02_41_40m_questionOfTheMonth.zip
This is their "Question of the month" recording. It's related to the some kind of earthquake in turkey. One can send they answers, even by email. They also repeat their email and address again.

I also googled for related stuff a bit and found that blog entry:
https://shortwavearchive.com/archive/voice-of-turkey-german-language-service-september-10-2017
Guy reports that he received "The Voice of Turkey's German language service, recorded on September 10, 2017 starting at 2300 UTC on 9830 kHz."

Those numbers (9830 and 9700) makes me wonder, are they transmitting on various frequencies, or is my receiver really that off?