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.
It's trivial to make a receiver that's greatly higher quality than the transmitter and reception path, in fact it would be difficult NOT to make a receiver of much higher quality (although you do seem to have managed it?).

OK, whatever. Now you told me about this POT. Can you help me instead of cussing me?

Your right it is horrible. It has to be in line at the point of the schematic up to the speaker.

Read my explanation.
 
OK, how about a bit of simplification on the way.

Try connecting it like this.

JimB
 

Attachments

  • Audio Modified.GIF
    Audio Modified.GIF
    41.3 KB · Views: 144
Connect it as a potentiometer, as volume controls always are - one end to chassis/ground, slider feeds the poweramp, and the top end is the input.

Input impedance of the poweramp wants to be at least five times the pot resistance, and the source feeding the pot wants to be at least 1/5th of the pot resistance.
 
I am glad that Q19 and the many resistors that reduce the level of the audio are gone.

The datasheet for the TDA2822 shows a 10k resistor from the input to ground. If a 100k resistor is used then the input bias current in it (typically 100nA) might cause a high DC offset voltage at the output of the amp.
 
No wonder I was getting distortion....lol

What I did is I could not see under the pot in the can so I had two connections to the amp.

One through the pot and one straight to the amp....lol.

I was wondering why I couldn't turn it down unless I used that other configuration.

Oh lord. I took the pot out and it was amazing what I saw from changes I had made.

:D
 
OK, I fixed the pot. I did try to set it up like Bill said and this is how it is now with the resistor working.

This is the final product and any cracks you hear are from the camera audio. I backed up and turned the audio down and still could not avoid the audio cracks from the camera. That's what I tried to explain about SSB is that you will have sudden audio peak then back to nothing. You cannot just amplify away to your hearts content. You must distribute it evenly through out your audio section. I learned this the hardware by building. The AGC totally prevents any cracks in the audio.

This is video with audio

Free File Hosting Made Simple - MediaFire

The one called "Finished-Product.mov"
 
Last edited:
The voices are unintelligible because there is a huge peak at 600Hz with lower and upper audio frequencies very much attenuated.

The voices sound like the echo sound from a submarine's sonar.
 
I take it back. It ain't final.

I don't like the low tone and I changed out the 100K resistor with a 10K. That;s the tone I want. I got no problem driving a smaller speaker but I got a big ole Altec Lansing I want to use.

So guess what? Occasional feedback again. I know what it is now. Who said the driver stage wasn't needed for the Chip? This is where I get my revenge. See I build....OK? I'm a perfectionist. Let me tell you what's going on.....

The driver isn't enough. It needs a buffer stage. The load is so heavy at audio peaks that it distorts the signal through a quiescent compression (loading). What that means is the current demand is greater than what is being supplied. Sounds like a job for a Darlington Amp.

That was the feedback problem the whole time.
 
The voices are unintelligible because there is a huge peak at 600Hz with lower and upper audio frequencies very much attenuated.

The voices sound like the echo sound from a submarine's sonar.

Yep and I found out what causes those peaks. Over loading.

They might sound like they are in submarine....lol. But that's SSB for ya. That is the crystal filter. You've made audio filters. Well you never had a need to make real narrow ones. I have and others have done real good ones. I don't have enough experience at it. But they have made ones that are like 200 Hz wide. You can't even understand SSB. But it is awesome in CW where 100 guys are piled up on one frequency and you want to talk to only one guy. I've heard it. It's great but it sounds like it's in a tin can.
 
See, I will explain this. What happens is when you don't have enough current gain, all transistors have characteristics that they respond too. All transistors have high hfe at specific frequencies. When you over load an an amplifier it can only supply the current at those frequencies instead of the full range because the current was not there so it is able to create it's own current source at those frequencies. For full range you do not want to take advantage of this characteristic. Look at your specs. It will show you at what freqs they have high gain.
 
I take it back. It ain't final.

I don't like the low tone and I changed out the 100K resistor with a 10K. That;s the tone I want. I got no problem driving a smaller speaker but I got a big ole Altec Lansing I want to use.

So guess what? Occasional feedback again. I know what it is now. Who said the driver stage wasn't needed for the Chip? This is where I get my revenge. See I build....OK? I'm a perfectionist. Let me tell you what's going on.....

The driver isn't enough. It needs a buffer stage. The load is so heavy at audio peaks that it distorts the signal through a quiescent compression (loading). What that means is the current demand is greater than what is being supplied. Sounds like a job for a Darlington Amp.

That was the feedback problem the whole time.

Presumably you never followed the advice that every single poster has given you? - throw most of it away, design it properly, and decouple all stages (there should be far less of them), and it will work properly. As you didn't know how to wire a simple volume control I would also suggest dumping the audio AGC as well, it's not required for a comms radio, and may well be causing problems as well (as I presume it's not designed correctly either?).

But I don't expect you to take any advice, or follow any good design techniques - just keep happily bodging extra un-needed and wrongly designed stages on - as long as it makes you happy :p
 
I think you want a narrow IF bandwidth for Morse code and a much wider one for voices. Your radio is made for morse code but its noise level is too high.

Since none of your very high gain transistor have proper decoupling from the supply, maybe when the transistor that drove the power amplifier IC was removed then its coupling from the power supply cancelled coupling from the power supply of other transistors and causes the feedback now.
 
But I don't expect you to take any advice, or follow any good design techniques - just keep happily bodging extra un-needed and wrongly designed stages on - as long as it makes you happy :p

Where mstecha left off - Space Varmint takes over!:D

JimB
 
Could he really be mstecha?
 
I chatted with MStechCA on another forum about his modified super-regen radio a few months ago. He has been on two other forums in the last year with new names but I recognised him.

He was still trying to make his modified super-regen radio have the same sensitivity as a cheap clock radio. It was picking up interference maybe because it has an AM detector.

There are many SSB radio projects listed in Google. One said it has a gain of 100dB and its audio amp is so sensitive that it picks up buzzing from nearby mains wiring.
The RF gain is 20dB so the audio gain is "only" 80dB (10,000) which is not much more than a microphone preamp that does not pickup mains buzzing. Its audio bandwidth is 2500Hz (-6dB) for speech.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top