Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

Audio feedback problem

Status
Not open for further replies.
The sound quality of space-Varmint's radio is horrible. No low frequencies nor high frequencies, just 1kHz to 2khz.
There are many people talking (or many stations received) all at the same time (poor selectivity).
The intellegibility was very poor but I heard a couple of numbers spoken.

My grandmother had a shortwave radio about 50 years ago and its sound was excellent.
With a short antenna it received very clear speech from around the world.

I made a Single-Sideband-Suppressed-Carrier transmitter and receiver for scrambling voices in a boardroom wireless tele-conference system. The scrambled signal was very scrambled without the demodulator (it sounded something like Space-Varmint's radio) and the output from the demodulator was perfect, exactly like the original sound.
 
Hi John,
I don't know enough about this radio stuff to fully understand what is happening with Frank's device, but I must learn from somewhere. I'm starting to get an interest in radio and all of its complexities.
 
Hi Johnny,
Frank's radio has many odd amplifier and attenuator stages. He mentioned that it picked up hum (probably from its power supply) so he cut off all the low audio frequencies.
 
The sound quality of space-Varmint's radio is horrible. No low frequencies nor high frequencies, just 1kHz to 2khz.
There are many people talking (or many stations received) all at the same time (poor selectivity).
The intellegibility was very poor but I heard a couple of numbers spoken.

My grandmother had a shortwave radio about 50 years ago and its sound was excellent.
With a short antenna it received very clear speech from around the world.

I made a Single-Sideband-Suppressed-Carrier transmitter and receiver for scrambling voices in a boardroom wireless tele-conference system. The scrambled signal was very scrambled without the demodulator (it sounded something like Space-Varmint's radio) and the output from the demodulator was perfect, exactly like the original sound.

Lordy lordy. You just told on yourself trying to be a smart ass again. How do you think you are going to squeeze full fidelity through 2 KHz????

Knuckle head.

Those were SSB recordings. They are about 2 KHz wide....lol

edit:

That's why when you receive SSB you have to add the carrier. When you transmit SSB your not even sending enough intelligence to make out the voice. It takes atleast 3.5 KHz to receive voice communications like narrow band FM. SSB is only about 2 KHz.
 
Last edited:
Oh, and your grand mother's radio sounds like a piece of ****. It's just some old dump SWL radio. This is communications quality baby! Wait until I put some front end filtering on it. I'll be able to hear you walking around...lol ;)
 
Does "communications quality" mean that speech is unintellegible and garbled?

You have the low audio frequencies feeding Q39 completely eliminated. Other parts cut the high audio frequencies. Then the response is just a narrow band filter at 2kHz. A whistle. Also your radio receives many stations at the same time.

An ordinary AM radio has an audio frequency response of 50Hz to 2kHz or 3kHz and sounds muffled but doesn't sound like a 2kHz whistle. It receives only one station at a time.

Here is your Q39 transistor being fed from the detector diode:

Edit: The schematic had the labels backwards.
 

Attachments

  • amplifier transistor.PNG
    amplifier transistor.PNG
    15.9 KB · Views: 156
Last edited:
Does "communications quality" mean that speech is unintellegible and garbled?

You have the low audio frequencies feeding Q39 completely eliminated. Other parts cut the high audio frequencies. Then the response is just a narrow band filter at 2kHz. A whistle. Also your radio receives many stations at the same time.

An ordinary AM radio has an audio frequency response of 50Hz to 2kHz or 3kHz and sounds muffled but doesn't sound like a 2kHz whistle. It receives only one station at a time.

Here is your Q39 transistor being fed from the detector diode:

Edit: The schematic had the labels backwards.

It aint garbled at all. That's the best your gonna get after the digital conversion to record. It is completely class A from rf front end through 2 IFs and through the audio. Class A. No clipping, no fudging power, well buffered class A. Perfectly intelligible SSB. Your not hearing what I hear because of the computer DSP. It aint no studio quality.

An ordinary AM radio has an audio frequency response of 50Hz to 2kHz or 3kHz and sounds muffled but doesn't sound like a 2kHz whistle. It receives only one station at a time.

Totally wrong partner. First of all it is not "any" AM it is SSB reception. And SSB will sound more than muffled, it will be unintelligible. If the the signal strength is real real strong you might make out a couple of words, but for must signals you won't ever hear them.

The whistling is the steep squirts in the crystal filter which is desirable. You will even here that with an audio OP filter or DSP. The narrower the band pass the more whistle you get. I can receive normal AM on this too. I got a switch to widen the filter. But really if you are going to receive AM you just want a ceramic filter at the most. You will achieve better fidelity with an LC filter. Of course you selectivity goes in the crapper but that is the trade off. That is where many mistakenly refer to it as sensitivity because the receiver will not be as sensitive but it is actually the selectivity, the blocking of noise the greatly enhances a good communications receiver.
 
It aint garbled at all. That's the best your gonna get after the digital conversion to record. It is completely class A from rf front end through 2 IFs and through the audio. Class A. No clipping, no fudging power, well buffered class A. Perfectly intelligible SSB. Your not hearing what I hear because of the computer DSP. It aint no studio quality.

Why do you have a fetish about class A?, and why do you think it makes any difference? - every radio in the world is class A - there's no other type of receiver amplifier stage, apart from power amps, where your's isn't class A anyway (and would make no difference if it was).

As Audioguru says, your poor design is drastically reducing the frequency response, but this isn't much of a problem for the extremely low quality of SSB. However, you wouldn't normally reduce the bass as much as your wrong components are doing, and you would normally limit the frequency response using a proper filter.
 
As for the butchered audio stage. Thanks. I saw where I did not have a cap's value labeled. I know it cuts the lows. That's what I was after. It is effectively a high pass filter but it does not mean the low audio is gone, it is just attenuated in that stage. It was necessary for the feedback problem I had. The lows are restored in later stages where I have a large restore value with a cap going to ground.
 
Why do you have a fetish about class A?, and why do you think it makes any difference? - every radio in the world is class A - there's no other type of receiver amplifier stage, apart from power amps, where your's isn't class A anyway (and would make no difference if it was).

As Audioguru says, your poor design is drastically reducing the frequency response, but this isn't much of a problem for the extremely low quality of SSB. However, you wouldn't normally reduce the bass as much as your wrong components are doing, and you would normally limit the frequency response using a proper filter.

BS. FM is mostly class C. CW can be class C. With SSB it is absolutely essential to be class A all the way through the receiver. If even one stage anywhere is even class B, believe me you will know it. It may still be intelligible but you might strain a little to understand every word.

See, you can get away with class C in FM because it is frequency modulated and there for much more forgiving when amplifier clip.
 
BS. FM is mostly class C. CW can be class C.

And you still claim to have passed your RAE? - you're talking about TRANSMITTER rf stages, NOT receivers, where everything is class A.

Transmitter wise FM uses class C multipliers and output stages, AM 'can' as long as you use high level modulation on the final stage - SSB must use a linear amplifier, doesn't need to be pure class A (which would make no difference), but it MUST be linear. If you use low level AM modulation, you require linear stages as with SSB, and for exactly the same reasons.

With SSB it is absolutely essential to be class A all the way through the receiver. If even one stage anywhere is even class B, believe me you will know it. It may still be intelligible but you might strain a little to understand every word.

Again, that's transmitter, not receiver - where your receiver has a class AB output stage, like probably every other one, and makes no difference whatsoever.

You claim to have the ARRL handbook?, try having a look at it!.
 
And you still claim to have passed your RAE? - you're talking about TRANSMITTER rf stages, NOT receivers, where everything is class A.

Transmitter wise FM uses class C multipliers and output stages, AM 'can' as long as you use high level modulation on the final stage - SSB must use a linear amplifier, doesn't need to be pure class A (which would make no difference), but it MUST be linear. If you use low level AM modulation, you require linear stages as with SSB, and for exactly the same reasons.



Again, that's transmitter, not receiver - where your receiver has a class AB output stage, like probably every other one, and makes no difference whatsoever.

You claim to have the ARRL handbook?, try having a look at it!.

No no. You can use class C in FM IFs. It's not until after the discriminator you need to worry about class A.

Again, that's transmitter, not receiver - where your receiver has a class AB output stage, like probably every other one, and makes no difference whatsoever.

That sounds like a matter of opinion. I'll stick with the class A. I have heard what cutting corners will do.
 
The lows are restored in later stages where I have a large restore value with a cap going to ground.
No.
Your radio used the extremely low value of C37 to cut the low frequencies below 33kHz. Nothing boosted them back. The later stages just passed along the very thin audio without any low frequencies and made the sound just a 2kHz whistle.

C37 would need to be 330 times higher in value (3.3uf) to pass frequencies as low as 100Hz. Try it (but then the audio will be much louder and might overload many transistor amplifier stages that can be removed).
 
I think SSB BW should be same as TELCO, ~3KHz otherwise everyone sounds like helium high chipmonks. Here is an audio sample of a SSB using 3KHz bandwidth. Still sounds crappy to me but it is understandable.
 
I measured my telco's round-trip frequency response. 3kHz was down -12dB!
I complained to Bell and they said it was typical and within their spec of -7.5dB one-way.
Then for every boardroom tele-conference system I demo'd my equalizer that boosted 3kHz +10db and sold every single one I demo'd. The sound was then crisp and clear instead of muffled.
 
I didn't have telco test equipment. I had high quality tele-conferencing equipment and many lines office telephone PBXs. Then later I got into multi-line telephone conferencing switches then some video conferencing.
 
You stated that you measured Telco freq response. How did you do it? Was just curious.
Don't mean to hijack SV's thread, just a sidetrack.
 
You stated that you measured Telco freq response. How did you do it? Was just curious.
Don't mean to hijack SV's thread, just a sidetrack.
I measured the frequency response of my tele-conferencing equipment and used it to send a signal through the telco. I used another one to receive the signal and I plotted the response and doubly added the deviation from flat of the equipment.
 
No no. You can use class C in FM IFs. It's not until after the discriminator you need to worry about class A.

You 'could' possibly in the very last stage, but if you've got enough signal to drive a class C amplifer, then you don't need an amplifier :D

IF and RF stages (and low signal audio) are all class A, end of story - you're just talking rubbish as usual.

That sounds like a matter of opinion. I'll stick with the class A. I have heard what cutting corners will do.

So where's your class A output stage?.
 
Status
Not open for further replies.

New Articles From Microcontroller Tips

Back
Top