Audio amp oscillations due to Rf.

[dr pepper]

A receiver I threw together using an Lm380 for the audio amp is misbehaving, it goes unstable at anything other that quiet volume levels.
Without the speaker connected it shows no oscillations on the 'scope, and a little RF, the detector (AM) filter isnt removing all the Rf.
So I tried a larger cap on the detector filter, but had to go so high the audio is really muffled, same goes for restricting the bandwidth of the amp by using a cap across the i/p's as per the datasheet.
So I bodged it putting a choke as per a passivle crossover inline with the spk.
What would be a cleaner approach to this, preferably without resorting to an op amp.

 

[KlausST]

Hi,

What would be a cleaner approach to this,

* a clean schematic, according datasheet and AN-69
* a proper PCB layout

none of both we can see.

Klaus

 

[dick_freebird]

You do not mention anything about power supply source or decoupling. These stand to allow self-oscillation in the load/gnd/power loop, worse as output power increases.

 

[biff44]

LM308! An oldie but a goodie.
you would have to work pretty hard to get it to oscillate, as it has less than 2 MHz of gain/bandwidth.

so first, it is a dual supply op amp. that means you need + and - supply rails. are you trying to use it with only one supply rail?
there is a compensation capacitor recommended from pin 8 to ground...are you using that?

how are you coupling your signal into the op amp? AC coupling? DC Coupling? either might have some quirks that would make it unstable.

like already mentioned, you need power supply bypass capacitors to ground on pin 4 and pin 8, at least 1 uf each

 

[dick_freebird]

I thought this was about a LM380 audio amplifier?

 

[Audioguru]

I betcha it is built with wires all over the place on a solderless "dreadboard" where every amplifier oscillates.
Power supply and the bypass pin decoupling and the important RC on the output to ground are needed.

 

[vfone]

LM380 is an old problem-maker circuit. Most probably (if working hard on filtering and decoupling) you will make it work somehow.
But I would recommend replacing it with any other equivalent circuit. I would recomend TDA2030, which at the same output power have much less distortions (<0.1%).

 

[Audioguru]

The TDA2030 and its replacement the TDA2030A are obsolete, not made anymore and not available at many places except fake or rejects might be sold on ebay.
The LM380 works fine if it is made with the complete circuit shown on its datasheet, on a compact pcb layout (not on a solderless breadboard).

 

[biff44]

you know...when we routinely make amplifiers that have high gain from DC to 40 GHz work without oscillating, it is not easy for us to remember the days when we had trouble getting an LM741 to work!
Let's not be dinging a newbie who is trying to learn.