How can you recognize what the announcer on an old fashioned AM radio is saying when it produces only the vowels of speech? All the important consonants of speech that reach 14kHz are missing. Consonants are needed to understand speech, which is why most speech on a telephone is, "What? What did you say?".

 

Eh? What are you saying?
Ain't matter to me anyway because I am almost deaf!

 

I am old and have normal-for-my-age (75) high frequency hearing loss. I like music and conversations so I got high quality hearing aids to hear normally again.
I became completely blind with cataracts. So I got cataracts surgery to make my vision better than ever before.
I was almost killed with a heart attack so I got stents surgery to fix me.
My tonsils, hernias and other things also got fixed.
I am young again.

 

One author named Walter James May published a 30 page book on classic crystal radio:
He studied 12 of what he considered the best crystal sets.
At the time of the writing in 1954 the radio became more compact than the earlier designs.
Some of the elements that surround the crystal and it's function are illustrated in this book.
The pictorial type diagrams were more common, used to improve replication.
https://worldradiohistory.com/BOOKSHELF-ARH/Technology/Boys-Book-Of-Crystal-Sets.pdf

 

One author named Walter James May published a 30 page book on classic crystal radio:
He studied 12 of what he considered the best crystal sets.
At the time of the writing in 1954 the radio became more compact than the earlier designs.
Some of the elements that surround the crystal and it's function are illustrated in this book.
The pictorial type diagrams were more common, used to improve replication.
https://worldradiohistory.com/BOOKSHELF-ARH/Technology/Boys-Book-Of-Crystal-Sets.pdf

66 years ago there were only a few AM radio stations and they were only in large cities. Then the poor selectivity of a crystal radio with only one LC stage was fine.
Today with many AM radio stations in and around town a crystal radio would probably play a few stations at the same time because REAL radios have at least 5 LC stages for good selectivity.

 

Just an update, the circuit is working but I received no output from the speaker. The problem is I wound my own transformer and the transformer does not respond at the desired frequency the antenna tuner is operating on. I believe it is the ferrite material I have used? SP1 it is, please see ferrite brochure.

Under test with a frequency generator, the transformer seems to operate around the 3 MHz range. I'll be doing my research to get this transformer working i.e. sourcing the correct core material.

Note I had to use a 144 turn to 12 turn ratio (12:1) to suit the 4 Ohm speaker.

 

Hello Audio Guru, you must have acquired good genes, your mind is quick.
I think the design for AM and Shortwave having overlapping signals can be expected.
Compared to a radio today the small parts and the complexity can be a fairly advanced topic.

 

Just an update, the circuit is working but I received no output from the speaker. The problem is I wound my own transformer and the transformer does not respond at the desired frequency the antenna tuner is operating on. I believe it is the ferrite material I have used? SP1 it is, please see ferrite brochure.

Under test with a frequency generator, the transformer seems to operate around the 3 MHz range. I'll be doing my research to get this transformer working i.e. sourcing the correct core material.

Note I had to use a 144 turn to 12 turn ratio (12:1) to suit the 4 Ohm speaker.

It never would have crossed my mind to build my own AF output transformer. Here is a picture showing the relative size of a transistor radio output transformer (the red component to the left of the loudspeaker).





If you cannot get hold of a transformer why don't you build the circuit on post #37, shown here again. You do not have to use the exact transistors noted except that Q2 and Q3 should be a complementary matched pair.

 

I wound my own transformer and the transformer does not respond at the desired frequency the antenna tuner is operating on.

It dos not operate at the RF frequency, it operates at the audio frequenccies, or about a max of 5000Hz for AM radio.

Note I had to use a 144 turn to 12 turn ratio (12:1) to suit the 4 Ohm speaker.

Okay, let’s do some calculations. Impedance ratio is the turns rato squared, so 576 Ohms for a 4 Ohm load.

You will need a power of about 100 mW to get minimally audible sound out of a speaker.

The formula for power is

V^2 / R, so, setting that to 0.1,

V^2 / 576 = 0.1

V^2 = 57.6

V = 7.6

Now that is the peak voltage, so you need twice that or 15.2 V p-p. A typical amp can get only with about 2V of the ground and power so, say 19V. But the peak to peak is not the correct measure so you still need sqrt(2) times that or about 27V. That would be the supply voltage you would need to get minimal power output using your transformer.

Another point: it is not just turns ratio that matter. You cannot make a 2 to 1 transformer with 2 turns and 1 turn. I believe you need way more turns than you used.

Bob

 

Just an update, the circuit is working but I received no output from the speaker. The problem is I wound my own transformer and the transformer does not respond at the desired frequency the antenna tuner is operating on. I believe it is the ferrite material I have used? SP1 it is, please see ferrite brochure.

Under test with a frequency generator, the transformer seems to operate around the 3 MHz range. I'll be doing my research to get this transformer working i.e. sourcing the correct core material.

Note I had to use a 144 turn to 12 turn ratio (12:1) to suit the 4 Ohm speaker.

I am not aware of any ferrite cores that are suitable for audio frequencies. Audio transformers use laminated silicon steel cores similar to mains power transformers. The laminations are usually in the shapes of "E" and "I".
There are two ways of stacking them. If the transformer is to be used in a single ended circuit where DC current will be flowing in the primary, the "E"s and "I"s are assembled in two separate stacks with a very small insulated gap between them to avoid saturation.
If the transformer is for use in a push-pull circuit where there is very little DC current flowing in the primary winding, the core laminations are interleaved to form a single stack.
Regards,
Keith

 

In addition to using many tuned LC circuits for good selectivity, a REAL AM radio has an RF amplifier, an AF amplifier and Automatic-Gain-Control (AGC) so that weak distant stations have the same loudness as strong local stations.

AM means Amplitude modulation where audio frequencies modulate the level of the radio frequency. The AM detector diode rectifies the radio signal so that the audio os detected. Then the audio is amplified by an audio amplifier.

It seems that you are wrongly trying to feed the 1MHz radio signal to the speaker through an RF transformer.

 

The subject about a nice audio transformer with a 4 Ohm speaker is less common so that makes this more difficult but is more interesting. There used to be so many transistor radios to open up and pull a transformer and speaker from.
I had to work hard as a paper boy to pay for my first transistor radio so doing surgery to remove the transformer is not the best option. The second hand store has a few behind the glass but that also might have been someones prize possession.
The core of the EE14 appears to be used ok on 8 Ohm speakers. 4 ohm the right laminated steel core is bigger ?
I am not sure about powdered core for audio as a replacement for the steel core. What we do know is that the 16 Ohm works and possibly there is room for improvement finding a similar transformer. I think the transformer suited for the speaker will have an appropriate shape for it's hysteresis curve. It is easy to forget that the old radios had brilliant audio engineers. By sweeping the transformer's secondary and speaker we should find a peak somewhere in the audio range. Working our way back to the primary the data for inductance begins to coordinate with the turns ratio. Some have an audio impedance matching network and they can dial in the perfect match.

 

Audio output transformers and 16 ohm speakers were used with vacuum tube amplifiers 70 years ago.
Audio has a range of frequencies from 20Hz to 20kHz and it is difficult for a transformer to produce that wide range.
But this thread is about a lousy sounding AM radio that will have a response from 100Hz to 2kHz if you are lucky.

An amplifler using transistors does not match the amplifier output impedance to the speaker impedance. A modern audio amplifier has an output impedance of 0.04 ohms or less so that it can damp the resonances of a speaker and produce almost no voltage loss.

 

Forget about ferrites for audio transformer cores. They work best over a very narrow range of higher frequencies, You need a transformer that will work over a range of at least 100Hz to 10,000Hz. Designing an audio transformer is always a compromise. You need enough volume of iron in the core to handle the power at low frequencies and low capacitance in the windings to reduce losses at high frequencies. It's surprising what you can get away with though if you are prepared to sacrifice a little bit of quality.
The best audio amplifier I ever built was many years ago, using vacuum tubes with a pair of EL84 for push-pull output. With my limited funds I could not afford an output transformer for it so I used a second-hand power transformer with a tapped 120-0-120V primary and a secondary tapped at 6, 12 and 24 volts. I measured the frequency response at work with a B&K swept frequency recorder. It was remarkably flat up to 12KHz then it gradually tailed off.
Regards,
Keith

 

For most here this is just a review point.
After the diode (demodulator) there is one more process to remove the carrier.
The signal can look like this. The stripes (DC pulses) are removed with a low pass filter and what is left is the audio.
It does not look like this it is an audio and it forms it's own boundaries like a line along the top of these DC pulses.
The lower frequency remains as audio. It is then amplified and can be seen on an oscilloscope.

 

For most here this is just a review point.
After the diode (demodulator) there is one more process to remove the carrier.
The signal can look like this. The stripes (DC pulses) are removed with a low pass filter and what is left is the audio.
It does not look like this it is an audio and it forms it's own boundaries like a line along the top of these DC pulses.
The lower frequency remains as audio. It is then amplified and can be seen on an oscilloscope.
View attachment 217607

Hi Sparky,

Thank you for bringing this to my attention, I am aware that low pass filters are used to remove the carrier wave from the rectified signal.

If I could pick your brain with regard to a bigger problem. We humans have capacitance and when we touch metallic items we induce that capacitance in circuits. I have that issue at the moment were the radio is dead silent at the moment and once I touch the antenna the signals in the oscilloscope go up by a factor of 10 and you can hear the radio, its like the difference between night and day.

My understanding is this, human capacitance has a value of around 100pF. At my target frequency this gives a reactance of 1591 roughly 1600.

Why would my circuit be wanting to change that value? It would be sending the net reactance down to about 0.

 

I got a little success with this amplifier design:



Its Able to fill a room with audio although the cons are that selectivity is bad, stations are on top of each other.

I think I'll stick with the BD139 amplifier, with an audio transformer to match the impedance of the output to the speaker. I picked up a 500 Ohm to 8 Ohm audio transformer and i'll aim to gear the output impedance of the amplifier accordingly.

I just want to say a big thank you to everyone who contributed their time to this. Great advice you all gave.

 

When you touch the antenna terminal of the "radio" then it works because your body becomes the antenna and adds its capacitance to ground to the tuned LC.

Your new amplifier has 50 ohms plus 15 ohms in series with the 6 ohms speaker creating an attenuator reducing the maximum peak output to 12V/2 x 6/(50 + 15)= 0.55V which is a true power in the speaker of only 0.39W.
A properly designed 4 transistors amplifier powered from 12V produces a true power of 3W into a 6 ohms speaker.
But the input capacitor is only 100pF which passes no audio frequencies into the low input impedance of the amplifier.
All audio amplifiers have negative feedback to reduce distortion. Yours has none.

A 500 ohm transformer in a 12VDC circuit produces a peak power of (12V/2) squared / 500 ohms= 0.072W which is a true power of only 0.01W. If you use a 40V supply with a vacuum tube then the true power will be only 0.57W like a cheap clock radio.
An output transformer was used 70 years ago in a high voltage vacuum tubes amplifier.

The new "radio" has a few stations on top of each other because it has only one LC tuned circuit that is damped by the very low input impedance of the new amplifier.

 

Thank you to all that contributed. I see I need to learn more on amplifier construction before I can draw out the maximum power the speaker is capable. With a few changes to the amplifier design, I managed to extract 0.21W from the speaker, nothing like 3W rms that I hoped for.

I have picked up a book on transistor amplifiers by Dennis Feucht and will be going through it all throughout the next couple of months before I resurrect this thread with new improvements.

From there I will obtain my ham radio licence and start digesting "Solid State Design for the radio amateur" and start planning the band and configuration of my shack. (I am hoping for 100w 40m band output for Dxing).

Once again thank you to all who contributed and I have learned a great deal.

 

The new "radio" has a few stations on top of each other because it has only one LC tuned circuit that is damped by the very low input impedance of the new amplifier.

The channel crowding AM modulation will likely change to digital and move up higher frequency to sell more bandwidth.
The design change would be TRF AM all digital.
http://www.insideradio.com/free/fcc...cle_e3185dae-06dc-11eb-a24a-db9822043f62.html