Audio feedback problem

[Space Varmint]

I has to go back and re-bias some audio amps so I could maintain a ball park level for AGC in a radio receiver. Well, now I got more audio than I know what to do with. I tried bypassing the speaker pre-driver stage and didn't have quite enough so I'm leaving it in. If schematic is needed I can have done shortly as to what it looks like at the moment.
I have tried heavy bypass caps across the power buss but it just changed the frequency of the feedback. Any suggestions?

Thanks

 

[Mikebits]

Not sure what your saying. Are you getting audio into your power or modulating your RF? Can you decrease audio gain? Feedback could be layout, is this a PCB or BB?
If the latter, you might isolate grounds and tie at common point on chassis. Seperate supplies maybe?

 

[Space Varmint]

Not sure what your saying. Are you getting audio into your power or modulating your RF? Can you decrease audio gain? Feedback could be layout, is this a PCB or BB?
If the latter, you might isolate grounds and tie at common point on chassis. Seperate supplies maybe?

Oh, you think it's a ground loop? You could be right. It's on breadboard. I cannot decrease the audio up to where the AGC taps in. Then I have a pot and two amps. One is chip audioguru suggested. Great chip! A TDA 2822.

 

[audioguru]

You might have feedback through the power supply.
The audio amp is supposed to have a volume control at its input. Then you can decrease the audio to zero.

Aren't you making a SSB shortwave radio? Please post a schematic.

 

[Mikebits]

Oh, you think it's a ground loop? You could be right. It's on breadboard. I cannot decrease the audio up to where the AGC taps in. Then I have a pot and two amps. One is chip audioguru suggested. Great chip! A TDA 2822.

Well, AG knows his stuff so I would think his recomendation is a good one. I would run you audio on a seperate supply, and use a star ground configuration. Or perhaps try one then the other and maybe both. You are now working in the realm of voodootronics

Experimentation is the best approach...

Also sheilding the power audio from sensative areas might help. Anywhere that is high Z and gain will be easy targets for interference.

 

[Space Varmint]

You might have feedback through the power supply.
The audio amp is supposed to have a volume control at its input. Then you can decrease the audio to zero.

Aren't you making a SSB shortwave radio? Please post a schematic.

At the input of the chip? I have it at the input of the chip's driver. A transistor stage.

 

[Mikebits]

The TDA is the driver, right? Well there's your problem... Max gain on the TDA? I guess this is AG's area, so I will shut up now

 

[Space Varmint]

The TDA is the driver, right? Well there's your problem... Max gain on the TDA? I guess this is AG's area, so I will shut up now

Yeah the speaker driver. It's a hot chip!

 

[Space Varmint]

OK, I moved the pot and it did help. I will go after power bus to improve it.

 

[audioguru]

The TDA2822 power amplifier has a voltage gain of 100 which is plenty for a radio. It has a supply current drain of up to 1.2A. The 1.2A pulses cause the power supply voltage to sag and that is where the collector resistor of the transistor preamp (that is not needed) is connected.
I have seen high gain amplifiers vibrating ceramic coupling capacitors (that should not be used for coupling audio) which caused feedback.

The schematic will show exactly where the feedback is coming from.

 

[Space Varmint]

The TDA2822 power amplifier has a voltage gain of 100 which is plenty for a radio. It has a supply current drain of up to 1.2A. The 1.2A pulses cause the power supply voltage to sag and that is where the collector resistor of the transistor preamp (that is not needed) is connected.
I have seen high gain amplifiers vibrating ceramic coupling capacitors (that should not be used for coupling audio) which caused feedback.



Hey thanks audioguru. No, I did need a driver for the chip. Some signals are very weak in short-wave. I about got it. The chip seems to be prone toward low frequency oscillation. And...whenever a signal comes in above a certain level it starts picking up hum....probably the power wires hanging out everywhere. What I did was put some decoupling caps in the audio and a great big one (value) across the power source.
I tried using low value (.01uf coupling caps to emphasize the higher audio frequencies, but that chip make it's own low frequency. I swear. That's one reason I don't depend to heavily on IC's because you don't have the level of control you might like. But as far as putting out extreme audio power at now 5.6 volts. This guy does the job and more. I sould be able to button this thing up and cut off all the extra power leads and see how it actually performs.

 

[audioguru]

An AM radio should have enough gain in its front end and IF to play weak signals loudly. Then AGC reduces the gain of the IF and then the gain of the front end for very strong signals so that weak and strong signals are the same volume.

The TDA2822 and all other IC power amplifiers have an absolutely flat frequency response. Low frequencies and high frequencies have exactly the same amount of gain.

Your problem is probably caused because the radio does not have enough RF gain so you are using an excessive amount of audio amplifier gain instead, then feedback from the power supply is causing hum, then feedback.

Without seeing your schematic then we are only guessing.

 

[Space Varmint]

An AM radio should have enough gain in its front end and IF to play weak signals loudly. Then AGC reduces the gain of the IF and then the gain of the front end for very strong signals so that weak and strong signals are the same volume.

The TDA2822 and all other IC power amplifiers have an absolutely flat frequency response. Low frequencies and high frequencies have exactly the same amount of gain.

Your problem is probably caused because the radio does not have enough RF gain so you are using an excessive amount of audio amplifier gain instead, then feedback from the power supply is causing hum, then feedback.

Without seeing your schematic then we are only guessing.

Well your right about the rf gain. It was done deliberately. I do not want much rf gain because of blocking & intermod. I prefer to use front end filtering for my rf gain which has not been done yet. I have two IF's where I did as much amplification as possible. I got the second IF up to about 1 volt P-P for weakest signal input. That was done to reduce the amount of audio amplification needed. Here is what my audio looks like:

I have the whole radio in like five cat food cans and so there are power wires coming out of each and are excessive in length. I know these are AC 60 cycle antennas. I should be able to cut them off today.

Anyway....grade me on my audio. you are the king there audioguru.

Correction: the chip driver is not biased like that. I accidentally pulled up the wrong file. Here are the corrections:

1. R98 = 10K

2. R95 =10K

3. R96 =920 ohms

4. R100 = 100K

There are a couple of changes to the first audio amp too, like that 1500pf cap was removed. Very minor changes. I just don't have time to cut, paste and re-size the whole dern thing.

 

[Nigel Goodwin]

What's R99? - assuming it's supposed to be the volume control it's connected completely wrongly, also decouple your split supply for the opamp.

 

[Space Varmint]

What's R99? - assuming it's supposed to be the volume control it's connected completely wrongly, also decouple your split supply for the opamp.

Oh hey Nigel. Didn't think anyone would be up. Forgot you are in the UK.

Anyway the OP Amp is decoupled. Thanks for reminding me. I need to put that in the schematic.

What do you mean by split supply? It is single supply.

Actually my volume control is different. I need to get the right file and post it.

Hang on...

 

[Space Varmint]

This is the correct file. This is precisely what I have now.

 

[Nigel Goodwin]

Oh hey Nigel. Didn't think anyone would be up. Forgot you are in the UK.

Anyway the OP Amp is decoupled. Thanks for reminding me. I need to put that in the schematic.

What do you mean by split supply? It is single supply.

You're generating a split supply with two resistors, it needs to be decoupled with a capacitor.

Actually my volume control is different. I need to get the right file and post it.

It's still completely wrong, how can you have never seen how to wire a volume control?.

 

[Roff]

I can't believe how much gain you are throwing away. Here is a list of the resistors that are at the beginning of each attenuator, along with the approximate attenuation of each divider:

R135 -9dB
R147 -24dB
R57 -33dB
R58 -12dB
R100 -22dB

You are throwing away approximately 100dB of gain! Throwing away gain means more noise and more distortion at the output.

 

[Roff]

All three of your transistor amplifiers almost certainly have gain from Vcc to the output. Any noise on Vcc will be amplified at the output of each stage. You need your audio amps to have good power supply rejection.

 

[Nigel Goodwin]

All in all it's a pretty bizarre circuit, most of which isn't required.