built a 5w mini amp with LM384, need help

[HTH]

hi all, I've built a small practice amp into a stompbox enclosure using a LM684N and it's sort of working. When the volume is increased above 1/4, it becomes unstable. I've made a quick youtube video so you can see how it sounds with it working and also when it becomes unstable... YouTube - 5w LM384N amp (houston, we have a problem...)

The circuit is right out of the data sheet application - exactly the same with the exception of the smoothing cap on the supply rail. I used 100uF whereas the datasheet application schematic showed 0.1uF. Here's the datasheet, the schematic is on page 5... http://www.national.com/ds/LM/LM384.pdf

Any ideas why it's acting up?

[Mark Hammer]

See reply on DIY Stompbox forum

[Brad1]

Hey...you could stick the pot in a rocker pedal and have your very own foot-adjustable Tom Morello killswitch effect!

Brad1

[defaced]

You should use it as the LFO in a trem circuit.

[sean k]

If you look at the charts for power diss versus output you'll see that for a 6W output youve got about 5W dissapation which means you need to supply 11W at least( for 8 ohms and 28V supply). Rule of thumb for these chip amps is power supply is 3 times Watts that output power is.

Plus you'd be hard pressed to have a high impedance, 100k, drive that 10k pot. If you put a fet in front... ala ruby and a 2N5457 or suchlike, as a source follower, you might have more luck giving it the current it wants.

Now I'll go look at DIY.

[HTH]

Thanks for all the replies, will try those suggestions tomorrow. However, regarding the source follower idea - the Digitech pedal I'm feeding the LM384 with has a 2N5088 emitter-follower driving the output, so I'm not sure how much use an extra buffer is going to be.

I had another thought too - the voltage supply is currently obtained via three PP3's in series for 23v (they're not quite measuring 9v, hence I'm not getting 27v from them). Would this cause any problems? - might it be better just to get a 24vdc wall wart?

[sean k]

27V's fine but whats your current capability. Each 9V should be at least 300mA but it's not a great thing to have them in series as the seconds pulling it's own current plus the current from the previous and third ones pulling it's own plus the previous two.

What ohmage are you driving? 8 looks the best on the data sheets but sometimes what looks good on the data sheets ain't what happens to sound best in real life. I like 386's driving 16 ohms...though people say 4 ohms is best.

[HTH]

27V's fine but whats your current capability. Each 9V should be at least 300mA but it's not a great thing to have them in series as the seconds pulling it's own current plus the current from the previous and third ones pulling it's own plus the previous two.

What ohmage are you driving? 8 looks the best on the data sheets but sometimes what looks good on the data sheets ain't what happens to sound best in real life. I like 386's driving 16 ohms...though people say 4 ohms is best.

I'm driving an 8ohm speaker. Also, I think the power supply (or lack thereof) is the problem. I'm getting a 24vdc PS this afternoon to replace the batteries.

[Steve Conner]

Try an 0.1uF capacitor as well as the 100uF. Put it as close to the IC's supply pins as you can.

Then again, maybe it is just lack of power from the batteries. You're asking a lot to drive an 8 ohm speaker off three 9Vs, it could draw something like 1 amp cranked.

[HTH]

thanks for all the replies - the pedal/amp is currently dead. when I rigged up a 24vdc ps, the smoothing cap on the board blew (twice); 40v cap. any idea why this might happen?

one weird thing - the 'tip' on this PS is marked as -'ve, but when I use my DMM to measure the voltage, I get a negative voltage with the black probe to the pin and the red probe to the barrel of the PS connector plug.

once I get more LM384s and some non-polarised caps, I'm gonna build another.