How to wind phasing transformers?

[neazoi]

Hello,
how should I wind the phasing transformers in this schematic?
http://www.neazoi.com/technology/diodes_only/mxr80.htm

They are two toroid cores.

I am confused about the turns direction both on top and bottom

 

[chuckey]

The dots on the schematic show the start of each winding. So a piece of tape on the start of your first winding (sec start?) wind turns repeat for other winding (pri start). Find other end of winding with a meter. The toroids are extremely small ~3mm outside diam. In the picture of the layout, you can see them as the two little spots just above the tagboard.
Frank

 

[sky_123]

The blobs in the diagram indicate which wire goes where. I don't know what a saturable mixer is, but, are you sure you're supposed to be using toroids? The diagram seems to indicate an adjustment, like in the typical metal can transformers. You can't do that with toroids. No idea if that is important to you or not.

 

[betwixt]

I'm puzzled at the bottom photograph, is it suggesting the cores for SR1 and SR2 are from old computer core memories?

I was already working in the computer industry when they were "hi-tech" but they were very tiny and only capable of passing three wires (X,Y and sense) though their holes. I'm guessing the outside diameter of the cores was only about 1mm.

Brian.

 

[neazoi]

I have already made this and I received radio Romania, about 1000Km away, using an audio amplifier of course.
My setup was different so that I could receive the MW band and I also used larger toroids, not the tiny ones he describes.

I do not know if smaller cores will be better because of the saturation but it works in my case.

The problem is that no matter how I wind the low turns sides, the mixer works either way.
In the photo you may be able to see what I have done.

I am a bit confused about the polarity of the coils and how to connect the two toroids together.
Which coils should I wind left hand sided and which right hand sided?Or the only thing that matters is the start and the end?

 

[neazoi]

Hello,
I need a little bit of help here please?

 

[chuckey]

Left and right handed does not matter, its direction of the turns that is important. So hold a core up at eye level, with its hole going upwards, now think is this winding going clockwise as I put on more turns or anti clockwise? It does not matter is the turns are being put on wards or away from you. The phasing of the coils is arranged so that the local oscillator and the input signal do not appear at the output terminal. It is helpful if the aerial connection (50 ohms) will conduct audio frequencies down to earth, which means that it should be bridged with a radio frequency choke.
Frank

 

[tpetar]

Useful link:

**broken link removed**


Also I suggest to see rest on Danyk web **broken link removed**

 

[neazoi]

Useful link:

**broken link removed**


Also I suggest to see rest on Danyk web **broken link removed**

These diagrams illustrate that winding direction is the one that plays the role in phasing!
Which of the two are correct?

Ok, let me think about it practically in this example. The cores he winds these coils are so tiny that direction of the coils (coil goes to the right as turns progressed or to the left) cannot be maintained. All the turns appear at about the same point in these tiny cores.
So I think what defines the phasing here is the polarity of the coil (turns screwed as you look the toroid, or unscrewed).

What do you think?

I think the high part(signal turns) is easy to make, just wind turns at one core, then continue to the other core in the same polarity.
I still do not know how to wind the lower parts (oscillator side)

 

[sky_123]

Per toroid, you can wind them any way you want. There is no interaction between the two toroids. All that matters is that per toroid, you observe the blobs on the schematic so that you can wire it into your circuit correctly.

 

[neazoi]

Per toroid, you can wind them any way you want. There is no interaction between the two toroids. All that matters is that per toroid, you observe the blobs on the schematic so that you can wire it into your circuit correctly.

Ok so basically I wind up any way I want each toroid, and then connect them accordingly to the schematic. If I find problems on the lower part (oscillator) I can swap the connections of one of the two toroids.

Would this work?

 

[sky_123]

If I find problems on the lower part (oscillator) I can swap the connections of one of the two toroids

If you wind two windings on a toroid, and mark the beginnings of each winding with some marker or paint spot, then
you can relate those to the blobs on the schematic, so no need to swap connections, because you will be 100% sure, since
you wound the windings yourself.
Now the only risk is that you may not be able to identify windings, and that can be resolved by using two different colors
of enamelled wire, or use a continuity tester to buzz them out, or have different lengths of wire ends so you don't forget,
or any other manner you wish.

 

[neazoi]

If you wind two windings on a toroid, and mark the beginnings of each winding with some marker or paint spot, then
you can relate those to the blobs on the schematic, so no need to swap connections, because you will be 100% sure, since
you wound the windings yourself.
Now the only risk is that you may not be able to identify windings, and that can be resolved by using two different colors
of enamelled wire, or use a continuity tester to buzz them out, or have different lengths of wire ends so you don't forget,
or any other manner you wish.

Yes I did this. Since the windings ratio is large, you can easily identify them, since the resistances differ. The 50 turns winding gave a resistanve of 1.8-2 ohms, whereas the 10 turns gave something like 0.7 ohms (all errors from the multimeter assummed).

The signals connection on the high-windings side is not difficult and since in so small toroids coil direction (not winding direction - polarity) does not seem to have a smooth pattern (turns may easilly overlap), One could easily swap the coil ends, to get the desired connection, without affecting the performance I think.

Then I wound the low-turns and tested the circuit. I received just one station on MW, radio Romania, about 1000Km away. When I swapped the connections of the low-turns, Sensitivity went up and I could receive 5 stations on the MW band, including Radio China International. Am I at the correct phasing of the coils? I have no idea if correct coil phasing causes the signal go up.

 

[betwixt]

Are you driving enough oscillator voltage into the coils? In order to work as mixers they must be driven hard so the cores saturate. The original memory cores were specifically designed to saturate at low current, that's how they stored a bit of data. I wonder if you are seeing leakage through the pi filter and the transistor is working as a detector instead. I presume you are using a germanium transistor with very low fT , the 2N35 has a unity gain frequency of only 400KHz.

While the working method is interesting, you do realize I hope that it is very inefficient compared to other mixer systems.

Brian.

 

[neazoi]

Are you driving enough oscillator voltage into the coils? In order to work as mixers they must be driven hard so the cores saturate. The original memory cores were specifically designed to saturate at low current, that's how they stored a bit of data. I wonder if you are seeing leakage through the pi filter and the transistor is working as a detector instead. I presume you are using a germanium transistor with very low fT , the 2N35 has a unity gain frequency of only 400KHz.

While the working method is interesting, you do realize I hope that it is very inefficient compared to other mixer systems.

Brian.

Yes of course, this is very inefficient 30-40db loss more than a diode mixer. I just thought it would be interesting to try such an old technology.
I am driving it with a signal generator at half the frequency, as this is a harmonics mixer, like the anti-parallel diodes one.
The voltage measured out of the generator is more than 2vpp, I think the cores can saturate at this voltage. I think, if they could not, I would not hear anything at the audio range, because they will not drop into the non-linear region and no mixing would be done, is that right?

I have not used a LPF yet neither a transistor as an amplifier. I have used these so called "amplified speakers" from commerce, so I hope no rectifying should occur in these. I have tried swapping the low-Z coils ends and I can listen to more signals with the same setup, so I believe the cores are doing the detection and not the "amplified speakers". What do you think?

 

[betwixt]

It sounds like you are doing it right with the signal generator but I would be cautious about the amplified speakers if you have not used the LPF. I have 'computer' monitor speakers here that can even pick up my cordless phone on 1.9GHz - and that's after I opened then and fitted filters on the power wires, input wires and an LPF on the input of the amplifier. They also make rather good monitors for my Ham radio transmitters although I wish they didn't :shock:

It would be worth trying the LPF just to be sure, 2V of RF from the oscillator is quite a large amount to have connected anywhere near an amplifier.

Brian.

 

[neazoi]

It sounds like you are doing it right with the signal generator but I would be cautious about the amplified speakers if you have not used the LPF. I have 'computer' monitor speakers here that can even pick up my cordless phone on 1.9GHz - and that's after I opened then and fitted filters on the power wires, input wires and an LPF on the input of the amplifier. They also make rather good monitors for my Ham radio transmitters although I wish they didn't :shock:

It would be worth trying the LPF just to be sure, 2V of RF from the oscillator is quite a large amount to have connected anywhere near an amplifier.

Brian.

Thank you very much, I will definitely try the LPF asap.
Just for interest, here is a photo of the mounting of these transformers.
They have been installed on the same size of ceramic material as before 1.5cm x 2cm. Tiny, aren't they?

 

[betwixt]

From their size I would guess those really are ferrte memory cores so you have done well to wind so many turns on them. In their original use they only had three wires passing through the hole.

Brian.